- icmp-echo
- icmp-jitter
- interval (LSP discovery)
- interval (params)
- ip-address (endpoint list)
- ip sla
- ip sla auto discovery
- ip sla auto endpoint-list
- ip sla auto group
- ip sla auto schedule
- ip sla auto template
- ip sla enable reaction-alerts
- ip sla ethernet-monitor
- ip sla ethernet-monitor reaction-configuration
- ip sla ethernet-monitor schedule
- ip sla group schedule
- ip sla key-chain
- ip sla logging traps
- ip sla low-memory
- ip sla monitor
- ip sla monitor group schedule
- ip sla monitor key-chain
- ip sla monitor logging traps
- ip sla monitor low-memory
- ip sla monitor reaction-configuration
- ip sla monitor reaction-trigger
- ip sla monitor reset
- ip sla monitor responder
- ip sla monitor responder type tcpConnect ipaddress
- ip sla monitor responder type udpEcho ipaddress
- ip sla monitor restart
- ip sla monitor schedule
- ip sla reaction-configuration
- ip sla reaction-trigger
- ip sla reset
- ip sla responder
- ip sla responder auto-register
- ip sla responder tcp-connect ipaddress
- ip sla responder udp-echo ipaddress
- ip sla restart
- ip sla schedule
- life
- lives-of-history-kept
- lsp-selector
- lsp-selector-base
- lsr-path
- maximum-sessions
- measurement-retry
- mpls discovery vpn interval
- mpls discovery vpn next-hop
- mpls lsp ping ipv4
- mpls lsp ping pseudowire
- mpls lsp trace ipv4
- num-packets
- operation-packet priority
- owner
- parameters
- path-discover
- path-echo
- path-jitter
- paths-of-statistics-kept
- precision
- probe-interval
- probe-packet priority
icmp-echo
To configure an IP Service Level Agreements (SLAs) Internet Control Message Protocol (ICMP) echo operation, use the icmp-echo command in IP SLA configuration mode.
icmp-echo {destination-ip-address | destination-hostname} [source-ip {ip-address | hostname} | source-interface interface-name]
Syntax Description
Defaults
No IP SLAs operation type is configured for the operation being configured.
Command Modes
IP SLA configuration (config-ip-sla)
Command History
Usage Guidelines
The default request packet data size for an ICMP echo operation is 28 bytes. Use the request-data-size command to modify this value. This data size is the payload portion of the ICMP packet, which makes a 64-byte IP packet.
You must configure the type of IP SLAs operation (such as User Datagram Protocol [UDP] jitter or ICMP echo) before you can configure any of the other parameters of the operation. To change the operation type of an existing IP SLAs operation, you must first delete the IP SLAs operation (using the no ip sla global configuration command) and then reconfigure the operation with the new operation type.
IP SLAs ICMP echo operations support both IPv4 and IPv6 addresses.
Examples
In the following example, IP SLAs operation 10 is created and configured as an echo operation using the ICMP protocol and the destination IPv4 address 172.16.1.175:
ip sla 10
icmp-echo 172.16.1.175
!
ip sla schedule 10 start-time now
In the following example, IP SLAs operation 11 is created and configured as an echo operation using the ICMP protocol and the destination IPv6 address 2001:DB8:100::1:
ip sla 11
icmp-echo 2001:DB8:100::1
!
ip sla schedule 11 start-time now
Related Commands
|
|
---|---|
ip sla monitor |
Begins configuration for an IP SLAs operation and enters IP SLA monitor configuration mode. |
icmp-jitter
To configure an IP Service Level Agreements (SLAs) Internet Control Message Protocol (ICMP) jitter operation, use the icmp-jitter command in IP SLA configuration mode.
icmp-jitter {destination-ip-address | destination-hostname} [interval milliseconds] [num-packets packet-number] [source-ip {ip-address | hostname}]
Syntax Description
Command Default
No IP SLAs operation type is configured for the operation being configured.
Command Modes
IP SLA configuration (config-ip-sla)
Command History
|
|
---|---|
12.4(6)T |
This command was introduced. |
12.0(32)SY |
This command was integrated into Cisco IOS Release 12.0(32)SY. |
Usage Guidelines
You must configure the type of IP SLAs operation (such as User Datagram Protocol [UDP] jitter or Internet Control Message Protocol [ICMP] echo) before you can configure any of the other parameters of the operation. To change the operation type of an existing IP SLAs operation, you must first delete the IP SLAs operation (using the no ip sla global configuration command) and then reconfigure the operation with the new operation type.
Examples
The following example shows how to configure an IP SLAs ICMP jitter operation:
ip sla 1
icmp-jitter 172.18.1.129 interval 40 num-packets 100 source-ip 10.1.2.34
frequency 50
!
ip sla reaction-configuration 1 react jitterAvg threshold-value 5 2 action-type trap threshold-type immediate
!
ip sla schedule 1 start-time now life forever
Related Commands
|
|
---|---|
ip sla |
Begins configuration for an IP SLAs operation and enters IP SLA configuration mode. |
interval (LSP discovery)
To specify the time interval between Multiprotocol Label Switching (MPLS) echo requests that are sent as part of the label switched path (LSP) discovery process for a Cisco IOS IP Service Level Agreements (SLAs) LSP Health Monitor operation, use the interval command in auto IP SLA MPLS LSP discovery parameters configuration mode. To return to the default value, use the no form of this command.
interval milliseconds
no interval
Syntax Description
milliseconds |
Number of milliseconds between each MPLS echo request. The default is 0. |
Command Default
0 milliseconds
Command Modes
Auto IP SLA MPLS LSP discovery parameters configuration (config-auto-ip-sla-mpls-lpd-params)
Command History
|
|
---|---|
12.2(31)SB2 |
This command was introduced. |
12.2(33)SRB |
This command was integrated into Cisco IOS Release 12.2(33)SRB. |
Usage Guidelines
Use the path-discover command to enable the LSP discovery option for an IP SLAs LSP Health Monitor operation and enter auto IP SLA MPLS LSP discovery parameters configuration mode.
Examples
The following example shows how to configure operation parameters, proactive threshold monitoring, and scheduling options using the LSP Health Monitor. In this example, the LSP discovery option is enabled for LSP Health Monitor operation 1. Operation 1 is configured to automatically create IP SLAs LSP ping operations for the equal-cost multipaths to all Border Gateway Protocol (BGP) next hop neighbors in use by all VPN routing and forwarding (VRF) instances associated with the source Provider Edge (PE) router. To discover the equal cost multipaths per BGP next hop neighbor, MPLS echo requests are sent every 2 milliseconds.
auto ip sla mpls-lsp-monitor 1
type echo ipsla-vrf-all
path-discover
!
maximum-sessions 2
session-timeout 60
interval 2
timeout 4
force-explicit-null
hours-of-statistics-kept 1
scan-period 30
!
auto ip sla mpls-lsp-monitor schedule 1 schedule-period 60 frequency 100 start-time now
!
auto ip sla mpls-lsp-monitor reaction-configuration 1 react lpd tree-trace action-type trapOnly
auto ip sla mpls-lsp-monitor reaction-configuration 1 react lpd lpd-group retry 3 action-type trapOnly
Related Commands
interval (params)
To specify the interval between packets for a jitter operation in an auto IP Service Level Agreements (SLAs) operation template, use the interval command in the appropriate submode of IP SLA template parameters configuration mode. To return to the default, use the no form of this command.
interval milliseconds
no interval
Syntax Description
milliseconds |
Interval between packets in milliseconds (ms). Range is from 4 to 60000. Default is 20. |
Command Default
The default interval between packets is 20 ms.
Command Modes
IP SLA Template Parameters Configuration
ICMP jitter configuration (config-icmp-jtr-params)
UDP jitter configuration (config-udp-jtr-params)
Command History
|
|
---|---|
15.1(1)T |
This command was introduced. |
Usage Guidelines
This command changes the interval between packets sent during a jitter operation from the default (20 ms) to the specified interval.
You must configure the type of IP SLAs operation, such as User Datagram Protocol (UDP) jitter or Internet Control Message Protocol (ICMP) jitter, before you can configure any other parameters of the operation.
Before you can use this command to configure auto IP SLAs operation templates, you must enter the parameters command in IP SLA template configuration mode.
Examples
The following example shows how to configure an auto IP SLAs operation template for an ICMP jitter operation with an interval of 30 ms between packets:
Router(config)#ip sla auto template type ip icmp-jitter 1
Router(config-tplt-icmp-jtr)#parameters
Router(config-icmp-jtr-params)#interval 30
Router(config-icmp-jtr-params)#end
Router# show ip sla auto template type ip icmp-jitter
IP SLAs Auto Template: 1
Measure Type: icmp-jitter
Description:
IP options:
Source IP: 0.0.0.0
VRF: TOS: 0x0
Operation Parameters:
Number of Packets: 10 Inter packet interval: 30
Timeout: 5000 Threshold: 5000
Statistics Aggregation option:
Hours of statistics kept: 2
Statistics Distributions options:
Distributions characteristics: RTT
Distributions bucket size: 20
Max number of distributions buckets: 1
Reaction Configuration: None
Related Commands
ip-address (endpoint list)
To specify destination IP addresses for routing devices or Cisco IOS IP Service Level Agreements (SLAs) Responders in Cisco devices and add them to an auto IP SLAs endpoint list, use the ip-address command in IP SLA endpoint-list configuration mode. To remove some or all IP addresses from the template, use the no form of this command.
ip-address address [-address | ,...,address] port port
no ip-address address [address-address | ,...,address] port port
Syntax Description
Command Default
The auto IP SLAs endpoint list is empty.
Command Modes
IP SLA endpoint-list configuration (config-epl)
Command History
|
|
---|---|
15.1(1)T |
This command was introduced. |
Usage Guidelines
This command adds IP addresses to the auto IP SLAs endpoint list being configured.
Destination IP addresses can either be manually configured by using this command or automatically discovered by using the discover command. If you use this command to configure an auto IP SLAs endpoint list, you cannot use the discover command to discover IP addresses for this endpoint list.
You cannot combine a list of individual IP addresses (address,address) and a range of IP addresses (address-address) in a single command.
The maximum number of IP addresses allowed in a list of individual addresses (address,address) per command is five.
To remove one or more IP addresses without reconfiguring the entire template, use the no form of this command. You can delete a range of IP addresses or a single IP addresses per command.
Modifications to auto IP SLAs endpoint lists, such as adding or removing IP addresses, take effect in the next schedule cycle.
Use the destination command in IP SLA auto-measure group configuration mode to specify an endpoint list for an IP SLAs auto-measure group.
Examples
The following example shows how to configure an IP SLAs endpoint list using this command:
Router(config)#ip sla auto endpoint-list type ip test
Router(config-epl)#ip-address 10.1.1.1-13 port 5000
Router(config-epl)#no ip-address 10.1.1.3-4 port 5000
Router(config-epl)#no ip-address 10.1.1.8 port 5000
Router(config-epl)#no ip-address 10.1.1.12 port 5000
Router(config-epl)#exit
Router#
The following output from the show ip sla auto endpoint command shows the results of the preceding configuration:
Router# show ip sla auto endpoint-list
Endpoint-list Name: test
Description:
ip-address 10.1.1.1-2 port 5000
ip-address 10.1.1.5-7 port 5000
ip-address 10.1.1.9-11 port 5000
ip-address 10.1.1.13 port 5000
Related Commands
ip sla
To begin configuring a Cisco IOS IP Service Level Agreements (SLAs) operation and enter IP SLA configuration mode, use the ip sla command in global configuration mode. To remove all configuration information for an operation, including the schedule of the operation, reaction configuration, and reaction triggers, use the no form of this command.
ip sla operation-number
no ip sla operation-number
Syntax Description
operation-number |
Operation number used for the identification of the IP SLAs operation you want to configure. |
Defaults
No IP SLAs operation is configured.
Command Modes
Global configuration
Command History
Usage Guidelines
The ip sla command is used to begin configuration for an IP SLAs operation. Use this command to specify an identification number for the operation you are about to configure. After you enter this command, the router will enter IP SLA configuration mode.
The ip sla command is supported in IPv4 networks. This command can also be used when configuring an IP SLAs operation that supports IPv6 addresses.
IP SLAs allows a maximum of 2000 operations.
Debugging is supported only on the first 32 operation numbers.
After you configure an operation, you must schedule the operation. For information on scheduling an operation, refer to the ip sla schedule and ip sla group schedule global configuration commands. You can also optionally set reaction triggers for the operation. For information on reaction triggers, refer to the ip sla reaction-configuration and ip sla reaction-trigger global configuration commands.
To change the operation type of an existing IP SLAs operation, you must first delete the IP SLAs operation (using the no ip sla) and then reconfigure the operation with the new operation type.
Note After you schedule an operation, you cannot modify the configuration of the operation. To modify the configuration of the operation after it is scheduled, you must first delete the IP SLAs operation (using the no ip sla command) and then reconfigure the operation with the new operation parameters.
To display the current configuration settings of the operation, use the show ip sla configuration command in user EXEC or privileged EXEC mode.
Examples
In the following example, operation 99 is configured as a UDP jitter operation in an IPv4 network and scheduled to start running in 5 hours. The example shows the ip sla command being used in an IPv4 network.
ip sla 99
udp-jitter 172.29.139.134 dest-port 5000 num-packets 20
!
ip sla schedule 99 life 300 start-time after 00:05:00
Note If operation 99 already exists and has not been scheduled, the command line interface will enter IP SLA configuration mode for operation 99. If the operation already exists and has been scheduled, this command will fail.
Related Commands
ip sla auto discovery
To enable auto discovery in Cisco IOS IP Service Level Agreements (SLAs) Engine 3.0, use the ip sla auto discovery command in global configuration mode. To disable auto discovery, use the no form of this command.
ip sla auto discovery
no ip sla auto discovery
Syntax Description
This command has no arguments or keywords.
Command Default
Auto discovery is disabled.
Command Modes
Global configuration (config)
Command History
|
|
---|---|
15.1(1)T |
This command was introduced. |
Usage Guidelines
This command enables the source for IP SLAs operations to auto-discover Cisco IP SLAs Responder endpoints.
Examples
The following example shows how to configure the ip sla auto discovery command:
Router>show ip sla auto discovery
IP SLAs auto-discovery status: Disabled
The following Endpoint-list are configured to auto-discovery:
Router>enable
Router#configure terminal
Enter configuration commands, one per line. End with CNTL/Z.
Router(config)#ip sla auto discovery
Router(config)#exit
Router#
Router# show ip sla auto discovery
IP SLAs auto-discovery status: Enabled
The following Endpoint-list are configured to auto-discovery:
.
.
.
Related Commands
|
|
---|---|
show ip sla auto discovery |
Displays the status of IP SLAs auto discovery and the configuration of auto IP SLAs endpoint lists configured using auto discovery. |
ip sla auto endpoint-list
To enter IP SLA endpoint-list configuration mode and begin configuring an auto IP Service Level Agreements (SLAs) endpoint list, use the ip sla auto endpoint-list command in global configuration mode. To remove an endpoint list, use the no form of this command.
ip sla auto endpoint-list type ip template-name
no ip sla auto endpoint-list template-name
Syntax Description
type ip |
Specifies that the operation type is Internet Protocol (IP). |
template-name |
Unique identifier of the endpoint list. Length of string is 1 to 64 ASCII characters. |
Command Default
No auto IP SLAs endpoint list is configured.
Command Modes
Global configuration (config)
Command History
|
|
---|---|
15.1(1)T |
This command was introduced. |
Usage Guidelines
This command assigns a name to an auto IP SLAs endpoint list and enters IP SLA endpoint-list configuration mode on the router.
Use the commands in IP SLA endpoint-list configuration mode to configure a template of destination IP addresses of routing devices or Cisco IOS IP SLAs Responders in Cisco devices to be referenced by one or more IP SLAs auto-measure groups. Destination addresses can be either manually configured by using the ip-address command or automatically discovered using the discover command.
Each auto IP SLAs endpoint list can be referenced by one or more IP SLAs auto-measure groups. Use the destination command in IP SLA auto-measure group configuration mode to specify an endpoint list for an IP SLAs auto-measure group.
Examples
The following example shows how to configure two auto IP SLAs endpoint lists of endpoints, one by manually configuring destination IP addresses and one using auto discovery:
Router(config)# ip sla auto endpoint-list type ip man1
Router(config-epl)# ip-address 10.1.1.1-10.1.1.12 port 23
Router(config-epl)# ip-address 10.1.1.15,10.1.1.23 port 23
Router(config-epl)# no ip-address 10.1.1.8,10.1.1.10 port 23
Router(config-epl)# description testing manual build
Router(config-epl)# exit
Router(config)#
Router(config)#ip sla auto discover
Router(config)#ip sla auto endpoint-list type ip autolist
Router(config-epl)#discover port 5000
Router(config-epl)#access-list 3
Router(config-epl)#exit
Router#
Router# show ip sla auto endpoint-list
Endpoint-list Name: man1
Description: testing manual build
ip-address 10.1.1.1-7 port 23
ip-address 10.1.1.9,10.1.1.15,10.1.1.23 port 23
Endpoint-list Name: autolist
Description:
Auto Discover Parameters
Destination Port: 5000
Access-list: 3
Ageout: 3600 Measurement-retry: 3
1 endpoints are discovered for autolist
Related Commands
ip sla auto group
To enter IP SLA auto-measure group configuration mode and begin configuring a Cisco IOS IP Service Level Agreements (SLAs) auto-measure group, use the ip sla auto group command in global configuration mode. To remove the auto-measure group configuration, use the no form of this command.
ip sla auto group type ip group-name
no ip sla auto group group-name
Syntax Description
type ip |
Specifies that the operation type for the group is Internet Protocol (IP). |
group-name |
Identifier of the group. String of 1 to 64 ASCII characters. |
Command Default
No IP SLAs auto-measure group is configured.
Command Modes
Global configuration (config)
Command History
|
|
---|---|
15.1(1)T |
This command was introduced. |
Usage Guidelines
This command assigns a name to an IP SLAs auto-measure group and enters IP SLA auto-measure group configuration mode.
Use the commands in IP SLA auto-measure group configuration mode to specify an auto IP SLAs operation template, endpoint list, and scheduler for the group.
Examples
The following example shows how to configure an IP SLAs auto-measure group:
Router(config)#ip sla auto group type ip 1
Router(config-am-grp)#destination 1
Router(config-am-grp)#schedule 1
Router(config-am-grp)#end
Router#
Router# show ip sla auto group
Group Name: 1
Description:
Activation Trigger: Immediate
Destination: 1
Schedule: 1
IP SLAs Auto Template: default
Measure Type: icmp-jitter
Description:
IP options:
Source IP: 0.0.0.0
VRF: TOS: 0x0
Operation Parameters:
Number of Packets: 10 Inter packet interval: 20
Timeout: 5000 Threshold: 5000
Statistics Aggregation option:
Hours of statistics kept: 2
Statistics Distributions options:
Distributions characteristics: RTT
Distributions bucket size: 20
Max number of distributions buckets: 1
Reaction Configuration: None
IP SLAs auto-generated operations of group 1
no operation created
Related Commands
|
|
---|---|
show ip sla auto group |
Displays configuration including default values of IP SLAs auto-measure groups. |
ip sla auto schedule
To enter IP SLA auto-measure schedule configuration mode and begin configuring an auto IP Service Level Agreements (SLAs) scheduler, use the ip sla auto schedule command in global configuration mode. To remove the configuration and stop all operations controlled by this scheduler, use the no form of this command.
ip sla auto schedule schedule-id
no ip sla auto schedule schedule-id
Syntax Description
schedule-id |
Unique identifier of scheduler. Range is 1 to 64 alphanumeric characters. |
Command Default
No auto IP SLAs scheduler is configured.
Command Modes
Global configuration (config)
Command History
|
|
---|---|
15.1(1)T |
This command was introduced. |
Usage Guidelines
This command assigns a unique identifier to an auto IP SLAs scheduler and enters IP SLA auto-measure schedule configuration mode on the router.
Use the commands in IP SLA auto-measure schedule configuration mode to modify the default configuration of an auto IP SLAs scheduler.
Each auto IP SLAs scheduler can be referenced by one or more IP SLAs auto-measure groups. Use the schedule command in IP SLA auto-measure group configuration mode to specify a scheduler for an IP SLAs auto-measure group.
Examples
The following example shows how to create the default configuration for an auto IP SLAs scheduler:
Router(config)#ip sla auto schedule 2
Router(config-am-schedule)#end
Router#
Router# show ip sla auto schedule 2
Group sched-id: 2
Probe Interval (ms) : 1000
Group operation frequency (sec): 60
Status of entry (SNMP RowStatus): Active
Next Scheduled Start Time: Pending trigger
Life (sec): 3600
Entry Ageout (sec): never
Related Commands
ip sla auto template
To enter IP SLA template configuration mode and begin configuring an auto IP Service Level Agreements (SLAs) operation template, use the ip sla auto template command in global configuration mode. To remove the operation template, use the no form of this command.
ip sla auto template type ip operation template-name
no ip sla auto template type ip operation template-name
Syntax Description
Command Default
No IP SLAs operation template is configured.
Command Modes
Global configuration (config)
Command History
|
|
---|---|
15.1(1)T |
This command was introduced. |
Usage Guidelines
This command assigns a name and operation to an auto IP SLAs operation template and enters a submode of the IP SLA template configuration mode based on the specified operation argument, such as IP SLA template icmp-echo configuration submode (config-tplt-icmp-ech).
Use the commands in IP SLA template configuration submode to modify the default configuration of an auto IP SLAs operation template.
Each auto IP SLAs operation template can be referenced by one or more IP SLAs auto-measure groups. Use the template command in IP SLA auto-measure group configuration mode to specify an operation template for an IP SLAs auto-measure group.
Examples
The following example shows how to create a default configuration for an auto IP SLAs operation template for ICMP echo:
Router(config)# ip sla auto template type ip icmp-echo
Router(config-tplt-icmp-ech)#end
Router# show ip sla auto template type ip icmp-echo
IP SLAs Auto Template: basic_icmp_echo
Measure Type: icmp-echo
Description:
IP options:
Source IP: 0.0.0.0
VRF: TOS: 0x0
Operation Parameters:
Request Data Size: 28 Verify Data: false
Timeout: 5000 Threshold: 5000
Statistics Aggregation option:
Hours of statistics kept: 2
History options:
History filter: none
Max number of history records kept: 15
Lives of history kept: 0
Statistics Distributions options:
Distributions characteristics: RTT
Distributions bucket size: 20
Max number of distributions buckets: 1
Reaction Configuration: None
Related Commands
ip sla enable reaction-alerts
To enable Cisco IP Service Level Agreements (SLAs) notifications to be sent to all registered applications, use the ip sla enable reaction-alerts command in global configuration mode. To disable IP SLAs notifications, use the no form of this command.
ip sla enable reaction-alerts
no ip sla enable reaction-alerts
Syntax Description
This command has no arguments or keywords.
Command Default
IP SLAs notifications are not sent to registered applications.
Command Modes
Global configuration (config)
Command History
|
|
---|---|
12.4(22)T |
This command was introduced. |
12.2(33)SRE |
This command was integrated into Cisco IOS Release 12.2(33)SRE. |
Usage Guidelines
The only applications that can register are Cisco IOS processes running on the router. Proactive threshold monitoring parameters for a Cisco IOS IP SLAs operation can be configured that will generate notifications when a threshold is crossed.
Examples
The following example shows how to enable IP SLAs notifications to be sent to all registered applications:
Router(config)# ip sla enable reaction-alerts
Related Commands
ip sla ethernet-monitor
To begin configuring a Cisco IOS IP Service Level Agreements (SLAs) auto Ethernet operation and enter IP SLA Ethernet monitor configuration mode, use the ip sla ethernet-monitor command in global configuration mode. To remove all configuration information for an auto Ethernet operation, including the schedule of the operation, reaction configuration, and reaction triggers, use the no form of this command.
ip sla ethernet-monitor operation-number
no ip sla ethernet-monitor operation-number
Syntax Description
operation-number |
Operation number used for the identification of the IP SLAs operation you want to configure. |
Command Default
No IP SLAs operation is configured.
Command Modes
Global configuration (config)
Command History
Usage Guidelines
The ip sla ethernet-monitor command is used to begin configuration for an IP SLAs auto Ethernet operation. Use this command to specify an identification number for the operation you are about to configure. After you enter this command, the router will enter IP SLA Ethernet monitor configuration mode.
After you configure an auto Ethernet operation, you must schedule the operation. To schedule an auto Ethernet operation, use the ip sla ethernet-monitor schedule command in global configuration mode. You can also optionally set reaction configuration for the operation (see the ip sla ethernet-monitor reaction-configuration command).
To display the current configuration settings of an auto Ethernet operation, use the show ip sla ethernet-monitor configuration command in user EXEC or privileged EXEC mode.
To change the operation type of an existing auto Ethernet operation, you must first delete the operation (using the no ip sla ethernet-monitor global configuration command) and then reconfigure the operation with the new operation type.
Examples
The following example shows how to configure operation parameters, proactive threshold monitoring, and scheduling options using an IP SLAs auto Ethernet operation. In this example, operation 10 is configured to automatically create IP SLAs Ethernet ping operations for all the discovered maintenance endpoints in the domain named testdomain and VLAN identification number 34. As specified by the proactive threshold monitoring configuration, when three consecutive connection loss events occur, a Simple Network Management Protocol (SNMP) trap notification should be sent. The schedule period for operation 10 is 60 seconds, and the operation is scheduled to start immediately.
ip sla ethernet-monitor 10
type echo domain testdomain vlan 34
!
ip sla ethernet-monitor reaction-configuration 10 react connectionLoss threshold-type consecutive 3 action-type trapOnly
!
ip sla ethernet-monitor schedule 10 schedule-period 60 start-time now
Related Commands
ip sla ethernet-monitor reaction-configuration
To configure proactive threshold monitoring parameters for a Cisco IOS IP Service Level Agreements (SLAs) auto Ethernet operation, use the ip sla ethernet-monitor reaction-configuration command in global configuration mode. To clear all threshold monitoring configuration for a specified auto Ethernet operation, use the no form of this command.
ip sla ethernet-monitor reaction-configuration operation-number [react monitored-element [action-type {none | trapOnly}] [threshold-type {average [number-of-measurements] | consecutive [occurrences] | immediate | never | xofy [x-value y-value]}] [threshold-value upper-threshold lower-threshold]]
no ip sla ethernet-monitor reaction-configuration operation-number [react monitored-element]
Syntax Description
operation-number |
Number of the IP SLAs operation for which reactions are to be configured. |
react monitored-element |
(Optional) Specifies the element to be monitored for threshold violations. Keyword options for the monitored-element argument are as follows: •connectionLoss—Specifies that a reaction should occur if there is a one-way connection loss for the monitored operation. •jitterAvg—Specifies that a reaction should occur if the average round-trip jitter value violates the upper threshold or lower threshold. •jitterDSAvg—Specifies that a reaction should occur if the average one-way destination-to-source jitter value violates the upper threshold or lower threshold. •jitterSDAvg—Specifies that a reaction should occur if the average one-way source-to-destination jitter value violates the upper threshold or lower threshold. •maxOfNegativeDS—Specifies that a reaction should occur if the one-way maximum negative jitter destination-to-source threshold is violated. •maxOfNegativeSD—Specifies that a reaction should occur if the one-way maximum negative jitter source-to-destination threshold is violated. •maxOfPositiveDS—Specifies that a reaction should occur if the one-way maximum positive jitter destination-to-source threshold is violated. •maxOfPositiveSD—Specifies that a reaction should occur if the one-way maximum positive jitter source-to-destination threshold is violated. |
react monitored-element (continued) |
•packetLateArrival—Specifies that a reaction should occur if the one-way number of late packets violates the upper threshold or lower threshold. •packetLossDS—Specifies that a reaction should occur if the one-way destination-to-source packet loss value violates the upper threshold or lower threshold. •packetLossSD—Specifies that a reaction should occur if the one-way source-to-destination packet loss value violates the upper threshold or lower threshold. •packetMIA—Specifies that a reaction should occur if the one-way number of missing packets violates the upper threshold or lower threshold. •packetOutOfSequence—Specifies that a reaction should occur if the one-way number of packets out of sequence violates the upper threshold or lower threshold. •rtt—Specifies that a reaction should occur if the round-trip time violates the upper threshold or lower threshold. •timeout—Specifies that a reaction should occur if there is a one-way timeout for the monitored operation. |
action-type none |
(Optional) Specifies that no action is taken when threshold events occur. The none keyword is the default value. Note If the threshold-type never keywords are configured, the action-type keyword is disabled. |
action-type trapOnly |
(Optional) Specifies that a Simple Network Management Protocol (SNMP) trap notification should be sent when threshold violation events occur. Note If the threshold-type never keywords are configured, the action-type keyword is disabled. |
threshold-type average [number-of-measurements] |
(Optional) Specifies that when the average of a specified number of measurements for the monitored element exceeds the upper threshold or when the average of a specified number of measurements for the monitored element drops below the lower threshold, the action defined by the action-type keyword should be performed. For example, if the upper threshold for react rtt threshold-type average 3 is configured as 5000 ms and the last three results of the operation are 6000, 6000, and 5000 ms, the average would be 6000 + 6000 + 5000 = 17000/3 = 5667. In this case, the average exceeds the upper threshold. The default number of 5 averaged measurements can be changed using the number-of-measurements argument. The valid range is from 1 to 16. This syntax is not available if the connectionLoss or timeout keyword is specified as the monitored element, because upper and lower thresholds do not apply to these options. |
threshold-type consecutive [occurrences] |
(Optional) Specifies that when a threshold violation for the monitored element is met consecutively for a specified number of times, the action defined by the action-type keyword should be performed. The default number of 5 consecutive occurrences can be changed using the occurrences argument. The valid range is from 1 to 16. |
threshold-type immediate |
(Optional) Specifies that when a threshold violation for the monitored element is met, the action defined by the action-type keyword should be performed immediately. |
threshold-type never |
(Optional) Specifies that threshold violations should not be monitored. This is the default threshold type. |
threshold-type xofy [x-value y-value] |
(Optional) Specifies that when a threshold violation for the monitored element is met x number of times within the last y number of measurements ("x of y"), action defined by the action-type keyword should be performed. The default is 5 for both the x and y values (xofy 5 5). The valid range for each value is from 1 to 16. |
threshold-value [upper-threshold lower-threshold] |
(Optional) Specifies the upper-threshold and lower-threshold values of the applicable monitored elements. See Table 10 in the "Usage Guidelines" section for a list of the default values. |
Command Default
IP SLAs proactive threshold monitoring is disabled.
Command Modes
Global configuration (config)
Command History
Usage Guidelines
You can configure the ip sla ethernet-monitor reaction-configuration command multiple times to enable proactive threshold monitoring for multiple elements (for example, configuring thresholds for round-trip time and destination-to-source packet loss) for the same operation. However, disabling of individual monitored elements is not supported. In other words, the no ip sla ethernet-monitor reaction-configuration command will disable all proactive threshold monitoring configuration for the specified IP SLAs operation.
SNMP traps for IP SLAs are supported by the CISCO-RTTMON-MIB and CISCO-SYSLOG-MIB. Use the ip sla logging traps command to enable the generation of SNMP system logging messages specific to IP SLAs trap notifications. Use the snmp-server enable traps rtr command to enable the sending of IP SLAs SNMP trap notifications.
To display the current threshold monitoring configuration settings for an auto Ethernet operation, use the show ip sla ethernet-monitor configuration command.
Table 10 lists the default upper and lower thresholds for specific monitored elements.
Examples
The following example shows how to configure operation parameters, proactive threshold monitoring, and scheduling options using an IP SLAs auto Ethernet operation. In this example, operation 10 is configured to automatically create IP SLAs Ethernet ping operations for all the discovered maintenance endpoints in the domain named testdomain and VLAN identification number 34. As specified by the proactive threshold monitoring configuration, when three consecutive connection loss events occur, an SNMP trap notification should be sent. The schedule period for operation 10 is 60 seconds, and the operation is scheduled to start immediately.
Router(config)# ip sla ethernet-monitor 10
Router(config-ip-sla-ethernet-monitor)# type echo domain testdomain vlan 34
!
Router(config)# ip sla ethernet-monitor reaction-configuration 10 react connectionLoss threshold-type consecutive 3 action-type trapOnly
!
Router(config)# ip sla ethernet-monitor schedule 10 schedule-period 60 start-time now
Related Commands
ip sla ethernet-monitor schedule
To configure the scheduling parameters for an IP Service Level Agreements (SLAs) auto Ethernet operation, use the ip sla ethernet-monitor schedule command in global configuration mode. To stop the operation and place it in the default state (pending), use the no form of this command.
ip sla ethernet-monitor schedule operation-number schedule-period seconds [frequency [seconds]] [start-time {after hh:mm:ss | hh:mm[:ss] [month day | day month] | now | pending}]
no ip sla ethernet-monitor schedule operation-number
Syntax Description
Command Default
The IP SLAs auto Ethernet operation is placed in a pending state (that is, the operation is enabled but is not actively collecting information).
Command Modes
Global configuration (config)
Command History
Usage Guidelines
After you schedule an IP SLAs auto Ethernet operation with the ip sla ethernet-monitor schedule command, you should not change the configuration of the operation until the operation has finished collecting information. To change the configuration of the operation, use the no ip sla ethernet-monitor schedule operation-number command in global configuration mode and then enter the new configuration information.
To display the current configuration settings of an IP SLAs auto Ethernet operation, use the show ip sla ethernet-monitor configuration command in user EXEC or privileged EXEC mode.
Examples
The following example shows how to configure operation parameters, proactive threshold monitoring, and scheduling options using an IP SLAs auto Ethernet operation. In this example, operation 10 is configured to automatically create IP SLAs Ethernet ping operations for all the discovered maintenance endpoints in the domain named testdomain and VLAN identification number 34. As specified by the proactive threshold monitoring configuration, when three consecutive connection loss events occur, a Simple Network Management Protocol (SNMP) trap notification should be sent. The schedule period for operation 10 is 60 seconds, and the operation is scheduled to start immediately.
ip sla ethernet-monitor 10
type echo domain testdomain vlan 34
!
ip sla ethernet-monitor reaction-configuration 10 react connectionLoss threshold-type consecutive 3 action-type trapOnly
!
ip sla ethernet-monitor schedule 10 schedule-period 60 start-time now
Related Commands
ip sla group schedule
To perform multioperation scheduling for Cisco IOS IP Service Level Agreements (SLAs) operations, use the ip sla group schedule command in global configuration mode. To cause all the IP SLAs operations belonging to a multioperation schedule to become inactive, use the no form of this command.
ip sla group schedule group-id {operation-ids | add operation-ids | delete operation-ids | reschedule} schedule-period seconds [ageout seconds] [frequency [seconds | range random-frequency-range]] [life {forever | seconds}] [start-time {hh:mm[:ss] [month day | day month] | pending | now | after hh:mm:ss}]
no ip sla group schedule group-id
Syntax Description
Defaults
The multioperation schedule is placed in a pending state (that is, the group of IP SLAs operations are enabled but are not actively collecting information).
Command Modes
Global configuration
Command History
Usage Guidelines
Though the IP SLAs multioperation scheduling functionality helps in scheduling thousands of operations, you should be cautious when specifying the number of operations, the schedule period, and the frequency to avoid any significant CPU impact.
For example, consider a scenario where you are scheduling 1 to 780 operations at a schedule period of 60 seconds. The command would be as follows:
ip sla group schedule 2 1-780 schedule-period 60 start-time now
IP SLAs calculates how many operations it should start in each 1-second interval by dividing the number of operations by the schedule period (780 operations divided by 60 seconds, which is 13 operations per second). Operations 1 to 13 in multioperation group 2 start after 0 seconds, operations 14 to 26 start after 1 second, operations 27 to 40 start after 2 seconds, and the iteration continues until operations 768 to 780 start after 59 seconds. This high value of operations starting at every 1-second interval (especially for jitter operations) can load the CPU to very high values.
On a Cisco 2600 router, the maximum recommended value of operations per second is 6 or 7 (approximately 350 to 400 operations per minute). Exceeding this value of 6 or 7 operations per second could cause major performance (CPU) impact. Note that the maximum recommended value of operations per second varies from platform to platform.
Note No warning messages will be displayed if IP SLAs multiple operations scheduling leads to a high number of operations starting per second.
When you reboot the router, the IP SLAs multiple operations scheduling functionality schedules the operations in the same order as was done before the reboot. For example, assume the following operation had been scheduled:
ip sla group schedule 2 1-20 schedule-period 40 start-time now
Over a range of 40 seconds, 20 operations have to be started (that is, one operation every 2 seconds). After the system reboot, operation 1 will start at t seconds and operation 2 starts at t+2 seconds, operation 3 starts at t+4 seconds, and so on.
The IP SLAs multiple operations scheduling functionality schedules the maximum number of operations possible without aborting. However, this functionality skips those IP SLAs operations that are already running or those that are not configured and hence do not exist. The total number of operations will be calculated based on the number of operations specified in the command, irrespective of the number of operations that are missing or already running. The IP SLAs multiple operations scheduling functionality displays a message showing the number of active and missing operations. However, these messages are displayed only if you schedule operations that are not configured or are already running.
IP SLAs Random Scheduler
The IP SLAs random scheduler option provides the capability to schedule multiple IP SLAs operations to begin at random intervals over a specified duration of time. The random scheduler option is disabled by default. To enable the random scheduler option, you must configure the frequency range random-frequency-range keywords and argument. The operations within the multioperation schedule restart at uniformly distributed random frequencies within the specified frequency range. The following guidelines apply for setting the frequency range:
•The starting value of the frequency range should be greater than the timeout values of all the operations in the multioperation schedule.
•The starting value of the frequency range should be greater than the schedule period (amount of time for which the group of operations is scheduled). This guideline ensures that the same operation does not get scheduled more than once within the schedule period.
The following guidelines apply if the random scheduler option is enabled:
•The individual operations in a multioperation schedule will be uniformly distributed to begin at random intervals over the schedule period.
•The operations within the multioperation schedule restart at uniformly distributed random frequencies within the specified frequency range.
•The minimum time interval between the start of each operation in a multioperation schedule is 100 milliseconds (0.1 seconds). If the random scheduler option is disabled, the minimum time interval is 1 second.
•Only one operation can be scheduled to begin at any given time. If the random scheduler option is disabled, multiple operations can begin at the same time.
•The first operation will always begin at 0 milliseconds of the schedule period.
•The order in which each operation in a multioperation schedule begins is random.
Adding or Deleting IP SLAs Operations
The following guidelines apply when an IP SLAs operation is added to or deleted from an existing multioperation schedule:
•If an operation is added that already belongs to the multioperation schedule, no action is taken.
•If two or more operations are added after the multioperation schedule has started, then the start times of the newly added operations will be uniformly distributed based on a time interval that was calculated prior to the addition of the new operations. If two or more operations are added before the multioperation schedule has started, then the time interval is recalculated based on both the existing and newly added operations.
•If an operation is added to a multioperation schedule in which the random scheduler option is enabled, then the start time and frequency of the newly added operation will be randomly chosen within the specified parameters.
•If an operation is added to a multioperation schedule in which the existing operations have aged out or the lifetimes of the existing operations have ended, the newly added operation will start and remain active for the amount of time specified by the multioperation schedule.
•If an active operation is deleted, then the operation will stop collecting information and become inactive.
•If the ip sla group schedule group-id reschedule command is entered after an operation is added or deleted, the time interval between the start times of the operations is recalculated based on the new number of operations belonging to the multioperation schedule.
Before Cisco IOS Release 15.1(1)T, this command could not be used to schedule a single operation because the only valid options for the operation-ids argument were a list (id,id,id) of IDs, a range (id-id) of IDs, or a combination of lists and ranges. If you attempted to use this command to schedule a single operation, the following messages were displayed:
Router(config)# sla group schedule 1 1 schedule-period 5 start-time now
%Group Scheduler: probe list wrong syntax
%Group schedule string of probe ID's incorrect
Router(config)#
In Cisco IOS Release 15.1(1)T and later releases, a single operation ID is a valid option for the operation-ids argument.
Examples
The following example shows how to schedule IP SLAs operations 3, 4, and 6 to 10 (identified as group 1) using multioperation scheduling. In this example, the operations are scheduled to begin at equal intervals over a schedule period of 20 seconds. The first operation (or set of operations) is scheduled to start immediately. Since the frequency is not specified, it is set to the value of the schedule period (20 seconds) by default.
ip sla group schedule 1 3, 4, 6-10 schedule-period 20 start-time now
The following example shows how to schedule IP SLAs operations 1 to 3 (identified as group 2) using the random scheduler option. In this example, the operations are scheduled to begin at random intervals over a schedule period of 50 seconds. The first operation is scheduled to start immediately. The frequency at which each operation will restart will be chosen randomly within the range of 80 to 100 seconds.
ip sla group schedule 2 1-3 schedule-period 50 frequency range 80-100 start-time now
Related Commands
ip sla key-chain
To enable Cisco IOS IP Service Level Agreements (SLAs) control message authentication and specify an MD5 key chain, use the ip sla key-chain command in global configuration mode. To remove control message authentication, use the no form of this command.
ip sla key-chain name
no ip sla key-chain
Syntax Description
name |
Name of MD5 key chain. |
Defaults
Control message authentication is disabled.
Command Modes
Global configuration
Command History
Usage Guidelines
The authentication configuration on the IP SLAs source and IP SLAs Responder devices must be the same. In other words, both devices must be configured with the same key chain or both devices must not use authentication.
If the ip sla key-chain command is entered, at least one key must be added to the specified MD5 key chain in order for MD5 authentication to occur.
Examples
In the following example, the IP SLAs control message uses MD5 authentication, and the key chain name is CSAA. The authentication string for key 1 is csaakey1.
ip sla key-chain csaa
key chain csaa
key 1
key-string csaakey1
Related Commands
ip sla logging traps
To enable the generation of Simple Network Management Protocol (SNMP) system logging messages specific to Cisco IOS IP Service Level Agreements (SLAs) trap notifications, use the ip sla logging traps command in global configuration mode. To disable IP SLAs system logging SNMP traps, use the no form of this command.
ip sla logging traps
no ip sla logging traps
Syntax Description
This command has no arguments or keywords.
Defaults
SNMP system logging messages specific to IP SLAs trap notifications are not generated.
Command Modes
Global configuration
Command History
Usage Guidelines
SNMP trap notifications for IP SLAs can be configured as a triggered action, to be sent when monitored values exceed an upper threshold or fall below a lower threshold, or when a set of defined conditions are met. For example, an SNMP trap can be triggered by five consecutive timeouts during an IP SLAs operation. The sending of SNMP traps is one of the options for triggered actions that can be configured for IP SLAs threshold violations. To configure proactive threshold monitoring parameters for an IP SLAs operation, use the ip sla reaction-configuration command in global configuration mode.
SNMP traps for IP SLAs are supported by the CISCO-RTTMON-MIB and CISCO-SYSLOG-MIB. Use the snmp-server enable traps rtr command to enable the sending of IP SLAs SNMP trap notifications.
Examples
The following example shows the configuration of IP SLAs traps to be triggered for round-trip time (RTT) violations and Voice over IP (VoIP) mean opinion score (MOS) violations, and the necessary SNMP configuration for enabling these SNMP logging traps:
ip sla 1
udp-jitter 209.165.200.225 dest-port 9234
!
ip sla schedule 1 start now life forever
ip sla reaction-configuration 1 react rtt threshold-type immediate threshold-value 3000 2000 action-type trapOnly
ip sla reaction-configuration 1 react MOS threshold-type consecutive 4 threshold-value 390 220 action-type trapOnly
!
ip sla logging traps
snmp-server enable traps rtr
Related Commands
ip sla low-memory
To specify how much unused memory must be available to allow Cisco IOS IP Service Level Agreements (SLAs) configuration, use the ip sla low-memory command in global configuration mode. To remove the type configuration for the operation, use the no form of this command.
ip sla low-memory bytes
no ip sla low-memory
Syntax Description
bytes |
Specifies amount of memory, in bytes, that must be available to configure IP SLA. The range is from 0 to the maximum amount of free memory bytes available. |
Defaults
The default amount of memory is 25 percent of the memory available on the system.
Command Modes
Global configuration
Command History
Usage Guidelines
The ip sla low-memory command allows you to specify the amount of memory that the IP SLAs can use. If the amount of available free memory falls below the value specified in the ip sla low-memory command, then the IP SLAs will not allow new operations to be configured. If this command is not used, the default low-memory value is 25 percent. This means that if 75 percent of system memory has been utilized you will not be able to configure any IP SLAs characteristics.
The value of the ip sla low-memory command should not exceed the amount of free memory available on the system. To determine the amount of free memory available on the system, use the show memory user EXEC or privileged EXEC command.
Examples
In the following example, the router is configured so that no less than 2 MB of memory will be free for IP SLAs configuration:
ip sla low-memory 2097152
Related Commands
|
|
---|---|
ip sla |
Begins configuration for an IP SLAs operation and enters IP SLA configuration mode. |
show memory |
Displays statistics about memory, including memory-free pool statistics. |
ip sla monitor
Note Effective with Cisco IOS Release 12.4(4)T, 12.2(33)SB, and 12.2(33)SXI, the ip sla monitor command is replaced by the ip sla command. See the ip sla command for more information.
To begin configuring a Cisco IOS IP Service Level Agreements (SLAs) operation and enter IP SLA monitor configuration mode, use the ip sla monitor command in global configuration mode. To remove all configuration information for an operation, including the schedule of the operation, reaction configuration, and reaction triggers, use the no form of this command.
ip sla monitor operation-number
no ip sla monitor operation-number
Syntax Description
operation-number |
Operation number used for the identification of the IP SLAs operation you want to configure. |
Defaults
No IP SLAs operation is configured.
Command Modes
Global configuration
Command History
Usage Guidelines
The ip sla monitor command is used to begin configuration for an IP SLAs operation. Use this command to specify an identification number for the operation you are about to configure. After you enter this command, the router will enter IP SLA monitor configuration mode.
IP SLAs allows a maximum of 2000 operations.
Debugging is supported only on the first 32 operation numbers.
After you configure an operation, you must schedule the operation. For information on scheduling an operation, refer to the ip sla monitor schedule and ip sla monitor group schedule global configuration commands. You can also optionally set reaction triggers for the operation. For information on reaction triggers, refer to the ip sla monitor reaction-configuration and ip sla monitor reaction-trigger global configuration commands.
To change the operation type of an existing IP SLAs operation, you must first delete the IP SLAs operation (using the no ip sla monitor global configuration command) and then reconfigure the operation with the new operation type.
Note After you schedule an operation, you cannot modify the configuration of the operation. To modify the configuration of the operation after it is scheduled, you must first delete the IP SLAs operation (using the no ip sla monitor command) and then reconfigure the operation with the new operation parameters.
To display the current configuration settings of the operation, use the show ip sla monitor configuration command in user EXEC or privileged EXEC mode.
Examples
In the following example, operation 99 is configured as a UDP jitter operation and scheduled to start running in 5 hours:
ip sla monitor 99
type jitter dest-ipaddr 172.29.139.134 dest-port 5000 num-packets 20
!
ip sla monitor schedule 99 life 300 start-time after 00:05:00
Note If operation 99 already exists and has not been scheduled, the command line interface will enter IP SLA monitor configuration mode for operation 99. If the operation already exists and has been scheduled, this command will fail.
Related Commands
ip sla monitor group schedule
Note Effective with Cisco IOS Release 12.4(4)T, 12.2(33)SB, and 12.2(33)SXI, the ip sla monitor group schedule command is replaced by the ip sla group schedule command. See the ip sla group schedule command for more information.
To perform group scheduling for Cisco IOS IP Service Level Agreements (SLAs) operations, use the ip sla monitor group schedule command in global configuration mode. To stop the operation and place it in the default state of normal scheduling, use the no form of this command.
ip sla monitor group schedule group-operation-number operation-id-numbers schedule-period seconds [ageout seconds] [frequency [seconds | range random-frequency-range]] [life {forever | seconds}] [start-time {hh:mm[:ss] [month day | day month] | pending | now | after hh:mm:ss}]
no ip sla monitor group schedule
Syntax Description
Defaults
The operation is placed in a pending state (that is, the operation is enabled but is not actively collecting information).
Command Modes
Global configuration
Command History
Usage Guidelines
Though IP SLAs multiple operations scheduling functionality helps in scheduling thousands of operations, you should be cautious while specifying the number of operations, the schedule period, and the operation group frequency to avoid any significant CPU impact.
For example, consider a scenario where you are scheduling 1 to 780 operations at a schedule period of 60 seconds. The command would be as follows:
ip sla monitor group schedule 2 1-780 schedule-period 60 start-time now
IP SLAs calculates how many operations it should start in each 1-second interval by dividing the number of operations by the schedule period (780 operations divided by 60 seconds, which is 13 operations per second). Operations 1 to 13 in operation group 2 start after 0 seconds, operations 14 to 26 start after 1 second, operations 27 to 40 start after 2 seconds, and the iteration continues until operations 768 to 780 start after 59 seconds. This high value of operations starting at every 1-second interval (especially for jitter operations) can load the CPU to very high values.
On a Cisco 2600 router, the maximum recommended value of operations per second is 6 or 7 (approximately 350 to 400 operations per minute). Exceeding this value of 6 or 7 operations per second could cause major performance (CPU) impact. Note that the maximum recommended value of operations per second varies from platform to platform.
Note No warning messages will be displayed if IP SLAs multiple operations scheduling leads to a high number of operations starting per second.
When you reboot the router, the IP SLAs multiple operations scheduling functionality schedules the operations in the same order as was done before the reboot. For example, assume the following operation had been scheduled:
ip sla monitor group schedule 2 1-20 schedule-period 40 start-time now
Over a range of 40 seconds, 20 operations have to be started (that is, one operation every 2 seconds). After the system reboot, operation 1 will start at t seconds and operation 2 starts at t+2 seconds, operation 3 starts at t+4 seconds, and so on.
The IP SLAs multiple operations scheduling functionality schedules the maximum number of operations possible without aborting. However, this functionality skips those IP SLAs operations that are already running or those that are not configured and hence do not exist. The total number of operations will be calculated based on the number of operations specified in the command, irrespective of the number of operations that are missing or already running. The IP SLAs multiple operations scheduling functionality displays a message showing the number of active and missing operations. However, these messages are displayed only if you schedule operations that are not configured or are already running.
IP SLAs Random Scheduler
The IP SLAs random scheduler option provides the capability to schedule multiple IP SLAs operations to begin at random intervals over a specified duration of time. The random scheduler option is disabled by default. To enable the random scheduler option, you must configure the frequency range random-frequency-range keywords and argument. The group of operations restarts at uniformly distributed random frequencies within the specified frequency range. The following guidelines apply for setting the frequency range:
•The starting value of the frequency range should be greater than the timeout values of all the operations in the group operation.
•The starting value of the frequency range should be greater than the schedule period (amount of time for which the group operation is scheduled). This guideline ensures that the same operation does not get scheduled more than once within the schedule period.
The following guidelines apply if the random scheduler option is enabled:
•The individual operations in a group operation will be uniformly distributed to begin at random intervals over the schedule period.
•The group of operations restarts at uniformly distributed random frequencies within the specified frequency range.
•The minimum time interval between the start of each operation in a group operation is 100 milliseconds (0.1 seconds). If the random scheduler option is disabled, the minimum time interval is 1 second.
•Only one operation can be scheduled to begin at any given time. If the random scheduler option is disabled, multiple operations can begin at the same time.
•The first operation will always begin at 0 milliseconds of the schedule period.
•The order in which each operation in a group operation begins is random.
Examples
The following example shows how to schedule IP SLAs operations 3, 4, and 6 to 10 as a group (identified as group 1). In this example, the operations are scheduled to begin at equal intervals over a schedule period of 20 seconds. The first operation (or set of operations) is scheduled to start immediately. Since the frequency is not specified, it is set to the value of the schedule period (20 seconds) by default.
ip sla monitor group schedule 1 3, 4, 6-10 schedule-period 20 start-time now
The following example shows how to schedule IP SLAs operations 1 to 3 as a group (identified as group 2). In this example, the operations are scheduled to begin at random intervals over a schedule period of 50 seconds. The first operation is scheduled to start immediately. The random scheduler option is enabled and the frequency at which the group of operations will restart will be chosen randomly within the range of 80-100 seconds.
ip sla monitor group schedule 2 1-3 schedule-period 50 frequency range 80-100 start-time now
Related Commands
ip sla monitor key-chain
Note Effective with Cisco IOS Release 12.4(4)T, 12.2(33)SB, and 12.2(33)SXI, the ip sla monitor key-chain command is replaced by the ip sla key-chain command. See the ip sla key-chain command for more information.
To enable Cisco IOS IP Service Level Agreements (SLAs) control message authentication and specify an MD5 key chain, use the ip sla monitor key-chain command in global configuration mode. To remove control message authentication, use the no form of this command.
ip sla monitor key-chain name
no ip sla monitor key-chain
Syntax Description
name |
Name of MD5 key chain. |
Defaults
Control message authentication is disabled.
Command Modes
Global configuration
Command History
Usage Guidelines
The authentication configuration on the IP SLAs source and IP SLAs Responder devices must be the same. In other words, both devices must be configured with the same key chain or both devices must not use authentication.
If the ip sla monitor key-chain command is entered, at least one key must be added to the specified MD5 key chain in order for MD5 authentication to occur.
Examples
In the following example, the IP SLAs control message uses MD5 authentication, and the key chain name is CSAA. The authentication string for key 1 is csaakey1.
ip sla monitor key-chain csaa
key chain csaa
key 1
key-string csaakey1
Related Commands
ip sla monitor logging traps
Note Effective with Cisco IOS Release 12.4(4)T, 12.2(33)SB, and 12.2(33)SXI, the ip sla monitor logging traps command is replaced by the ip sla logging traps command. See the ip sla logging traps command for more information.
To enable the generation of Simple Network Management Protocol (SNMP) system logging messages specific to Cisco IOS IP Service Level Agreements (SLAs) trap notifications, use the ip sla monitor logging traps command in global configuration mode. To disable IP SLAs system logging SNMP traps, use the no form of this command.
ip sla monitor logging traps
no ip sla monitor logging traps
Syntax Description
This command has no arguments or keywords.
Defaults
SNMP system logging messages specific to IP SLAs trap notifications are not generated.
Command Modes
Global configuration
Command History
Usage Guidelines
SNMP trap notifications for IP SLAs can be configured as a triggered action, to be sent when monitored values exceed an upper threshold or fall below a lower threshold, or when a set of defined conditions are met. For example, an SNMP trap can be triggered by five consecutive timeouts during an IP SLAs operation. The sending of SNMP traps is one of the options for triggered actions that can be configured for IP SLAs threshold violations. To configure proactive threshold monitoring parameters for an IP SLAs operation, use the ip sla monitor reaction-configuration command in global configuration mode.
SNMP traps for IP SLAs are supported by the CISCO-RTTMON-MIB and CISCO-SYSLOG-MIB. Use the snmp-server enable traps rtr command to enable the sending of IP SLAs SNMP trap notifications.
Examples
The following example shows the configuration of IP SLAs traps to be triggered for round-trip time (RTT) violations and Voice over IP (VoIP) mean opinion score (MOS) violations, and the necessary SNMP configuration for enabling these SNMP logging traps:
ip sla monitor 1
type jitter dest-ipaddr 209.165.200.225 dest-port 9234
!
ip sla monitor schedule 1 start now life forever
ip sla monitor reaction-configuration 1 react rtt threshold-type immediate threshold-value 3000 2000 action-type trapOnly
ip sla monitor reaction-configuration 1 react MOS threshold-type consecutive 4 threshold-value 390 220 action-type trapOnly
!
ip sla monitor logging traps
snmp-server enable traps rtr
Related Commands
ip sla monitor low-memory
Note Effective with Cisco IOS Release 12.4(4)T, 12.2(33)SB, and 12.2(33)SXI, the ip sla monitor low-memory command is replaced by the ip sla low-memory command. See the ip sla low-memory command for more information.
To specify how much unused memory must be available to allow Cisco IOS IP Service Level Agreements (SLAs) configuration, use the ip sla monitor low-memory command in global configuration mode. To remove the type configuration for the operation, use the no form of this command.
ip sla monitor low-memory bytes
no ip sla monitor low-memory
Syntax Description
bytes |
Specifies amount of memory, in bytes, that must be available to configure IP SLA. The range is from 0 to the maximum amount of free memory bytes available. |
Defaults
The default amount of memory is 25 percent of the memory available on the system.
Command Modes
Global configuration
Command History
Usage Guidelines
The ip sla monitor low-memory command allows you to specify the amount of memory that the IP SLAs can use. If the amount of available free memory falls below the value specified in the ip sla monitor low-memory command, then the IP SLAs will not allow new operations to be configured. If this command is not used, the default low-memory value is 25 percent. This means that if 75 percent of system memory has been utilized you will not be able to configure any IP SLAs characteristics.
The value of the ip sla monitor low-memory command should not exceed the amount of free memory available on the system. To determine the amount of free memory available on the system, use the show memory user EXEC or privileged EXEC command.
Examples
In the following example, the router is configured so that no less than 2 MB of memory will be free for IP SLAs configuration:
ip sla monitor low-memory 2097152
Related Commands
ip sla monitor reaction-configuration
Note Effective with Cisco IOS Release 12.4(4)T, 12.2(33)SB, and 12.2(33)SXI, the ip sla monitor reaction-configuration command is replaced by the ip sla reaction-configuration command. See the ip sla reaction-configuration command for more information.
To configure proactive threshold monitoring parameters for a Cisco IOS IP Service Level Agreements (SLAs) operation, use the ip sla monitor reaction-configuration command in global configuration mode. To clear all threshold monitoring configuration for a specified IP SLAs operation, use the no form of this command.
ip sla monitor reaction-configuration operation-number react monitored-element [action-type option] [threshold-type {average [number-of-measurements] | consecutive [occurrences] | immediate | never | xofy [x-value y-value]}] [threshold-value upper-threshold lower-threshold]
no ip sla monitor reaction-configuration operation-number
Syntax Description
operation-number |
Number of the IP SLAs operation for which reactions are to be configured. |
react monitored-element |
Specifies the element to be monitored for threshold violations. Note The elements available for monitoring will vary depending on the type of IP SLAs operation you are configuring. Keyword options for the monitored-element argument are as follows: •connectionLoss—Specifies that a reaction should occur if there is a one-way connection loss for the monitored operation. •icpif—Specifies that a reaction should occur if the one-way Calculated Planning Impairment Factor (ICPIF) value violates the upper threshold or lower threshold. •jitterAvg—Specifies that a reaction should occur if the average round-trip jitter value violates the upper threshold or lower threshold. •jitterDSAvg—Specifies that a reaction should occur if the average one-way destination-to-source jitter value violates the upper threshold or lower threshold. •jitterSDAvg—Specifies that a reaction should occur if the average one-way source-to-destination jitter value violates the upper threshold or lower threshold. |
react monitored-element (continued) |
•maxOfNegativeDS—Specifies that a reaction should occur if the one-way maximum negative jitter destination-to-source threshold is violated. •maxOfNegativeSD—Specifies that a reaction should occur if the one-way maximum negative jitter source-to-destination threshold is violated. •maxOfPositiveDS—Specifies that a reaction should occur if the one-way maximum positive jitter destination-to-source threshold is violated. •maxOfPositiveSD—Specifies that a reaction should occur if the one-way maximum positive jitter source-to-destination threshold is violated. •mos—Specifies that a reaction should occur if the one-way mean opinion score (MOS) value violates the upper threshold or lower threshold. •packetLateArrival—Specifies that a reaction should occur if the one-way number of late packets violates the upper threshold or lower threshold. •packetLossDS—Specifies that a reaction should occur if the one-way destination-to-source packet loss value violates the upper threshold or lower threshold. •packetLossSD—Specifies that a reaction should occur if the one-way source-to-destination packet loss value violates the upper threshold or lower threshold. •packetMIA—Specifies that a reaction should occur if the one-way number of missing packets violates the upper threshold or lower threshold. •packetOutOfSequence—Specifies that a reaction should occur if the one-way number of packets out of sequence violates the upper threshold or lower threshold. •rtt—Specifies that a reaction should occur if the round-trip time violates the upper threshold or lower threshold. •timeout—Specifies that a reaction should occur if there is a one-way timeout for the monitored operation. •verifyError—Specifies that a reaction should occur if there is a one-way error verification violation. |
action-type option |
(Optional) Specifies what action or combination of actions the operation performs when threshold events occur. If the threshold-type never keywords are defined, the action-type keyword is disabled. The option argument can be one of the following keywords: •none—No action is taken. This option is the default value. •trapAndTrigger—Trigger an Simple Network Management Protocol (SNMP) trap and start another IP SLAs operation when the violation conditions are met, as defined in the trapOnly and triggerOnly options. •trapOnly—Send an SNMP logging trap when the specified violation type occurs for the monitored element. •triggerOnly—Have one or more target operation's operational state make the transition from pending to active when the violation conditions are met. The target operations to be triggered are specified using the ip sla monitor reaction-trigger command. A target operation will continue until its life expires, as specified by the target operation's configured lifetime value. A triggered target operation must finish its life before it can be triggered again. |
threshold-type average [number-of-measurements] |
(Optional) When the average of a specified number of measurements for the monitored element exceeds the upper threshold or when the average of a specified number of measurements for the monitored element drops below the lower threshold, perform the action defined by the action-type keyword. For example, if the upper threshold for react rtt threshold-type average 3 is configured as 5000 ms and the last three results of the operation are 6000, 6000, and 5000 ms, the average would be 6000 + 6000 + 5000 = 17000/3 = 5667, thus violating the 5000 ms upper threshold. The default number of 5 averaged measurements can be changed using the number-of-measurements argument. The valid range is from 1 to 16. This syntax is not available if the connectionLoss, timeout, or verifyError keyword is specified as the monitored element, because upper and lower thresholds do not apply to these options. |
threshold-type consecutive [occurrences] |
(Optional) When a threshold violation for the monitored element is met consecutively for a specified number of times, perform the action defined by the action-type keyword. The default number of 5 consecutive occurrences can be changed using the occurrences argument. The valid range is from 1 to 16. The occurrences value will appear in the output of the show ip sla monitor reaction-configuration command as the "Threshold Count" value. |
threshold-type immediate |
(Optional) When a threshold violation for the monitored element is met, immediately perform the action defined by the action-type keyword. |
threshold-type never |
(Optional) Do not calculate threshold violations. This is the default threshold type. |
threshold-type xofy [x-value y-value] |
(Optional) When a threshold violations for the monitored element is met x number of times within the last y number of measurements ("x of y"), perform the action defined by the action-type keyword. The default is 5 for both the x and y values (xofy 5 5). The valid range for each value is from 1 to 16. The x-value will appear in the output of the show ip sla monitor reaction-configuration command as the "Threshold Count" value, and the y-value will appear as the "Threshold Count2" value. |
[threshold-value upper-threshold lower-threshold] |
(Optional) Specifies the upper-threshold and lower-threshold values of the applicable monitored elements. See Table 10 in the "Usage Guidelines" section for a list of the default values. Note For MOS threshold values (react mos), the number is expressed in three digits representing ones, tenths, and hundredths. For example, to express a MOS threshold of 3.20, enter 320. The valid range is from 100 (1.00) to 500 (5.00). |
Defaults
IP SLAs proactive threshold monitoring is disabled.
Note See Table 11 in the "Usage Guidelines" section for a list of the default upper and lower thresholds for specific monitored elements.
Command Modes
Global configuration
Command History
Usage Guidelines
You can configure the ip sla monitor reaction-configuration command multiple times to enable proactive threshold monitoring for multiple elements (for example, configuring thresholds for destination-to-source packet loss and MOS) for the same operation. However, disabling of individual monitored elements is not supported. In other words, the no ip sla monitor reaction-configuration command will disable all proactive threshold monitoring configuration for the specified IP SLAs operation.
SNMP traps for IP SLAs are supported by the CISCO-RTTMON-MIB and CISCO-SYSLOG-MIB. Use the ip sla monitor logging traps command to enable the generation of SNMP system logging messages specific to IP SLAs trap notifications. Use the snmp-server enable traps rtr command to enable the sending of IP SLAs SNMP trap notifications.
To display the current threshold monitoring configuration settings for an IP SLAs operation, use the show ip sla monitor configuration command.
Table 11 lists the default upper and lower thresholds for specific monitored elements.
Examples
In the following example, IP SLAs operation 10 (a UDP jitter operation) is configured to send an SNMP logging trap when the MOS value exceeds 4.9 (best quality) or falls below 2.5 (poor quality):
ip sla monitor reaction-configuration 10 react mos threshold-type immediate threshold-value 490 250 action-type trapOnly
Related Commands
ip sla monitor reaction-trigger
Note Effective with Cisco IOS Release 12.4(4)T, 12.2(33)SB, and 12.2(33)SXI, the ip sla monitor reaction-trigger command is replaced by the ip sla reaction-trigger command. See the ip sla reaction-trigger command for more information.
To define a second Cisco IOS IP Service Level Agreements (SLAs) operation to make the transition from a pending state to an active state when one of the trigger action type options are defined with the ip sla monitor reaction-configuration command, use the ip sla monitor reaction-trigger command in global configuration mode. To remove the trigger combination, use the no form of this command.
ip sla monitor reaction-trigger operation-number target-operation
no ip sla monitor reaction-trigger operation
Syntax Description
Defaults
No trigger combination is defined.
Command Modes
Global configuration
Command History
Usage Guidelines
Triggers are usually used for diagnostics purposes and are not intended for use during normal operation conditions.
Examples
In the following example, a trigger action type is defined for IP SLAs operation 2. When operation 2 experiences certain user-specified threshold violation events while it is actively collecting statistical information, the operation state of IP SLAs operation 1 will be triggered to change from pending to active.
ip sla monitor reaction-trigger 2 1
Related Commands
ip sla monitor reset
Note Effective with Cisco IOS Release 12.4(4)T, 12.2(33)SB, and 12.2(33)SXI, the ip sla monitor reset command is replaced by the ip sla reset command. See the ip sla reset command for more information.
To perform a shutdown and restart of the Cisco IOS IP Service Level Agreements (SLAs) engine, use the ip sla monitor reset command in global configuration mode.
ip sla monitor reset
Syntax Description
This command has no arguments or keywords.
Defaults
None
Command Modes
Global configuration
Command History
Usage Guidelines
The ip sla monitor reset command stops all operations, clears IP SLAs configuration information, and returns the IP SLAs feature to the startup condition. This command does not reread the IP SLAs configuration stored in the startup configuration in NVRAM. You must retype the configuration or load a previously saved configuration file.
Note The ip sla monitor reset command does not remove IP SLAs label switched path (LSP) Health Monitor configurations from the running configuration.
Note Use the ip sla monitor reset command only in extreme situations such as the incorrect configuration of a number of operations.
Examples
The following example shows how to reset the Cisco IOS IP SLAs engine, clearing all stored IP SLAs information and configuration:
ip sla monitor reset
Related Commands
|
|
---|---|
ip sla monitor restart |
Restarts a stopped IP SLAs operation. |
ip sla monitor responder
Note Effective with Cisco IOS Release 12.4(4)T, 12.2(33)SB, and 12.2(33)SXI, the ip sla monitor responder command is replaced by the ip sla responder command. See the ip sla responder command for more information.
To enable the Cisco IOS IP Service Level Agreements (SLAs) Responder for general IP SLAs operations, use the ip sla monitor responder command in global configuration mode. To disable the IP SLAs Responder, use the no form of this command.
ip sla monitor responder
no ip sla monitor responder
Syntax Description
This command has no arguments or keywords.
Defaults
The IP SLAs Responder is disabled.
Command Modes
Global configuration
Command History
Usage Guidelines
This command is used on the destination device for IP SLAs operations to enable the sending and receiving of IP SLAs control packets. Enabling the IP SLAs Responder allows the generation of packet loss statistics on the device sending IP SLAs operations.
Prior to sending an operation packet to the IP SLAs Responder, the IP SLAs operation sends a control message to the IP SLAs Responder to enable the destination port.
Examples
The following example shows how to enable the IP SLAs Responder:
ip sla monitor responder
Related Commands
ip sla monitor responder type tcpConnect ipaddress
Note Effective with Cisco IOS Release 12.4(4)T, 12.2(33)SB, and 12.2(33)SXI, the ip sla monitor responder type tcpConnect ipaddress command is replaced by the ip sla responder tcp-connect ipaddress command. See the ip sla responder tcp-connect ipaddress command for more information.
To enable the Cisco IOS IP Service Level Agreements (SLAs) Responder for TCP Connect operations, use the ip sla monitor responder type tcpConnect ipaddress command in global configuration mode. To disable the IP SLAs Responder, use the no form of this command.
ip sla monitor responder type tcpConnect ipaddress ip-address port port-number
no ip sla monitor responder type tcpConnect ipaddress ip-address port port-number
Syntax Description
ip-address |
Destination IP address. |
port port-number |
Specifies the destination port number. |
Defaults
The IP SLAs Responder is disabled.
Command Modes
Global configuration
Command History
Usage Guidelines
This command is used on the destination device for IP SLAs operations to enable the acceptance and return of TCP connection operation packets.
Examples
The following example shows how to enable the IP SLAs Responder for TCP connection operations:
ip sla monitor responder type tcpConnect ipaddress A.B.C.D port 1
Related Commands
ip sla monitor responder type udpEcho ipaddress
Note Effective with Cisco IOS Release 12.4(4)T, 12.2(33)SB, and 12.2(33)SXI, the ip sla monitor responder type udpEcho ipaddress command is replaced by the ip sla responder udp-echo ipaddress command. See the ip sla responder udp-echo ipaddress command for more information.
To enable the Cisco IOS IP Service Level Agreements (SLAs) Responder for User Datagram Protocol (UDP) echo or jitter operations, use the ip sla monitor responder type udpEcho ipaddress command in global configuration mode. To disable the IP SLAs Responder, use the no form of this command.
ip sla monitor responder type udpEcho ipaddress ip-address port port-number
no ip sla monitor responder type udpEcho ipaddress ip-address port port-number
Syntax Description
ip-address |
Destination IP address. |
port port-number |
Specifies the destination port number. |
Defaults
The IP SLAs Responder is disabled.
Command Modes
Global configuration
Command History
Usage Guidelines
This command is used on the destination device for IP SLAs operations to enable UDP echo and jitter (UDP+) operations with control disabled.
Examples
The following example shows how to enable the IP SLAs Responder for jitter operations:
ip sla monitor responder type udpEcho ipaddress A.B.C.D port 1
Related Commands
ip sla monitor restart
Note Effective with Cisco IOS Release 12.4(4)T, 12.2(33)SB, and 12.2(33)SXI, the ip sla monitor restart command is replaced by the ip sla restart command. See the ip sla restart command for more information.
To restart a Cisco IOS IP Service Level Agreements (SLAs) operation, use the ip sla monitor restart command in global configuration mode.
ip sla monitor restart operation-number
Syntax Description
operation-number |
Number of the IP SLAs operation to restart. IP SLAs allows a maximum of 2000 operations. |
Defaults
None
Command Modes
Global configuration
Command History
Usage Guidelines
To restart an operation, the operation should be in an active state.
IP SLAs allows a maximum of 2000 operations.
This command does not have a no form.
Examples
The following example shows how to restart operation 12:
ip sla monitor restart 12
Related Commands
|
|
---|---|
ip sla monitor reset |
Clears all current IP SLAs statistics and configuration information from the router and resets the IP SLAs engine. |
ip sla monitor schedule
Note Effective with Cisco IOS Release 12.4(4)T, 12.2(33)SB, and 12.2(33)SXI, the ip sla monitor schedule command is replaced by the ip sla schedule command. See the ip sla schedule command for more information.
To configure the scheduling parameters for a single Cisco IOS IP Service Level Agreements (SLAs) operation, use the ip sla monitor schedule command in global configuration mode. To stop the operation and place it in the default state (pending), use the no form of this command.
ip sla monitor schedule operation-number [life {forever | seconds}] [start-time {hh:mm[:ss] [month day | day month] | pending | now | after hh:mm:ss}] [ageout seconds] [recurring]
no ip sla monitor schedule operation-number
Syntax Description
Defaults
The operation is placed in a pending state (that is, the operation is enabled but not actively collecting information).
Command Modes
Global configuration
Command History
Usage Guidelines
After you schedule the operation with the ip sla monitor schedule command, you cannot change the configuration of the operation. To change the configuration of the operation, use the no form of the ip sla monitor global configuration command and reenter the configuration information.
If the operation is in a pending state, you can define the conditions under which the operation makes the transition from pending to active with the ip sla monitor reaction-trigger and ip sla monitor reaction-configuration global configuration commands. When the operation is in an active state, it immediately begins collecting information.
The following time line shows the age-out process of the operation:
W----------------------X----------------------Y----------------------Z
where:
•W is the time the operation was configured with the ip sla monitor global configuration command.
•X is the start time or start of life of the operation (that is, when the operation became "active").
•Y is the end of life as configured with the ip sla monitor schedule global configuration command (life seconds have counted down to zero).
•Z is the age out of the operation.
Age out starts counting down at W and Y, is suspended between X and Y, and is reset to its configured size at Y.
The operation to can age out before it executes (that is, Z can occur before X). To ensure that this does not happen, configure the difference between the operation's configuration time and start time (X and W) to be less than the age-out seconds.
Note The total RAM required to hold the history and statistics tables is allocated at the time of scheduling the IP SLAs operation. This prevents router memory problems when the router gets heavily loaded and lowers the amount of overhead an IP SLAs operation causes on a router when it is active.
The recurring keyword is supported only for scheduling single IP SLAs operations. You cannot schedule multiple IP SLAs operations using the ip sla monitor schedule command. The life value for a recurring IP SLAs operation should be less than one day. The ageout value for a recurring operation must be "never" (which is specified with the value 0), or the sum of the life and ageout values must be more than one day. If the recurring option is not specified, the operations are started in the existing normal scheduling mode.
Examples
In the following example, operation 25 begins actively collecting data at 3:00 p.m. on April 5. This operation will age out after 12 hours of inactivity, which can be before it starts or after it has finished with its life. When this operation ages out, all configuration information for the operation is removed (that is, the configuration information is no longer in the running configuration in RAM).
ip sla monitor schedule 25 life 43200 start-time 15:00 apr 5 ageout 43200
In the following example, operation 1 begins collecting data after a 5-minute delay:
ip sla monitor schedule 1 start-time after 00:05:00
In the following example, operation 3 begins collecting data immediately and is scheduled to run indefinitely:
ip sla monitor schedule 3 start-time now life forever
In the following example, operation 15 begins automatically collecting data every day at 1:30 a.m.:
ip sla monitor schedule 15 start-time 01:30:00 recurring
Related Commands
ip sla reaction-configuration
To configure proactive threshold monitoring parameters for a Cisco IOS IP Service Level Agreements (SLAs) operation, use the ip sla reaction-configuration command in global configuration mode. To disable all the threshold monitoring configuration for a specified IP SLAs operation, use the no form of this command.
ip sla reaction-configuration operation-number [react monitored-element [action-type option] [threshold-type {average [number-of-measurements] | consecutive [occurrences] | immediate | never | xofy [x-value y-value]}] [threshold-value upper-threshold lower-threshold]]
no ip sla reaction-configuration operation-number [react monitored-element]
Syntax Description
operation-number |
Number of the IP SLAs operation for which reactions are to be configured. |
react monitored-element |
(Optional) Specifies the element to be monitored for threshold violations. Note The elements supported for monitoring will vary depending on the type of IP SLAs operation you are running. See the Usage Guidelines for information. Keyword options for the monitored-element argument are as follows: •connectionLoss—Specifies that a reaction should occur if there is a one-way connection loss for the monitored operation. The threshold-value keyword does not apply to this monitored element. •frameLossDS—Specifies that a reaction should occur if the one-way destination-to-source digital signal processor (DSP) frame loss value violates the upper threshold or lower threshold. •iaJitterDS—Specifies that a reaction should occur if the one-way destination-to-source interarrival jitter value violates the upper threshold or lower threshold. •iaJitterSD—Specifies that a reaction should occur if the one-way source-to-destination interarrival jitter value violates the upper threshold or lower threshold. •icpif—Specifies that a reaction should occur if the one-way Calculated Planning Impairment Factor (ICPIF) value violates the upper threshold or lower threshold. •jitterAvg—Specifies that a reaction should occur if the average round-trip jitter value violates the upper threshold or lower threshold. •jitterDSAvg—Specifies that a reaction should occur if the average one-way destination-to-source jitter value violates the upper threshold or lower threshold. •jitterSDAvg—Specifies that a reaction should occur if the average one-way source-to-destination jitter value violates the upper threshold or lower threshold. |
react monitored-element (continued) |
•latencyDSAvg—Specifies that a reaction should occur if the average one-way destination-to-source latency value violates the upper threshold or lower threshold. •latencySDAvg—Specifies that a reaction should occur if the average one-way source-to-destination latency value violates the upper threshold or lower threshold. •maxOflatencyDS—Specifies that a reaction should occur if the one-way maximum latency destination-to-source threshold is violated. •maxOflatencySD—Specifies that a reaction should occur if the one-way maximum latency source-to-destination threshold is violated. •maxOfNegativeDS—Specifies that a reaction should occur if the one-way maximum negative jitter destination-to-source threshold is violated. •maxOfNegativeSD—Specifies that a reaction should occur if the one-way maximum negative jitter source-to-destination threshold is violated. •maxOfPositiveDS—Specifies that a reaction should occur if the one-way maximum positive jitter destination-to-source threshold is violated. •maxOfPositiveSD—Specifies that a reaction should occur if the one-way maximum positive jitter source-to-destination threshold is violated. •mos—Specifies that a reaction should occur if the one-way Mean Opinion Score (MOS) value violates the upper threshold or lower threshold. •moscqds—Specifies that a reaction should occur if the one-way destination-to-source Mean Opinion Score for Conversational Quality (MOS-CQ) value violates the upper threshold or lower threshold. •moscqsd—Specifies that a reaction should occur if the one-way source-to-destination Mean Opinion Score for Conversational Quality (MOS-CQ) value violates the upper threshold or lower threshold. •moslqds—Specifies that a reaction should occur if the one-way destination-to-source Mean Opinion Score for Listening Quality (MOS-LQ) value violates the upper threshold or lower threshold. •packetLateArrival—Specifies that a reaction should occur if the one-way number of late packets violates the upper threshold or lower threshold. |
react monitored-element (continued) |
•packetLoss—Specifies that a reaction should occur if the packet loss value violates the upper threshold or lower threshold. The path of the packets is unknown. •packetLossDS—Specifies that a reaction should occur if the one-way destination-to-source packet loss value violates the upper threshold or lower threshold. •packetLossSD—Specifies that a reaction should occur if the one-way source-to-destination packet loss value violates the upper threshold or lower threshold. •packetMIA—Specifies that a reaction should occur if the one-way number of missing packets violates the upper threshold or lower threshold. •packetOutOfSequence—Specifies that a reaction should occur if the one-way number of packets out of sequence violates the upper threshold or lower threshold. •rFactorDS—Specifies that a reaction should occur if the one-way destination-to-source estimated transmission rating factor R violates the upper threshold or lower threshold. •rFactorSD—Specifies that a reaction should occur if the one-way source-to-destination estimated transmission rating factor R violates the upper threshold or lower threshold. •rtt—Specifies that a reaction should occur if the round-trip time violates the upper threshold or lower threshold. •successivePacketLoss—Specifies that a reaction should occur if the one-way number of successively dropped packets violates the upper threshold or lower threshold. •timeout—Specifies that a reaction should occur if there is a one-way timeout for the monitored operation. The threshold-value keyword does not apply to this monitored element. •verifyError—Specifies that a reaction should occur if there is a one-way error verification violation. The threshold-value keyword does not apply to this monitored element. |
action-type option |
(Optional) Specifies what action or combination of actions the operation performs when threshold events occur. If the threshold-type never keywords are defined, the action-type keyword is disabled. The option argument can be one of the following keywords: •none—No action is taken. This option is the default value. •trapAndTrigger—Trigger a Simple Network Management Protocol (SNMP) trap and start another IP SLAs operation when the violation conditions are met, as defined in the trapOnly and triggerOnly options. •trapOnly—Send an SNMP logging trap when the specified violation type occurs for the monitored element. •triggerOnly—Have one or more target operation's operational state make the transition from pending to active when the violation conditions are met. The target operations to be triggered are specified using the ip sla reaction-trigger command. A target operation will continue until its life expires, as specified by the target operation's configured lifetime value. A triggered target operation must finish its life before it can be triggered again. |
threshold-type average [number-of-measurements] |
(Optional) When the average of a specified number of measurements for the monitored element exceeds the upper threshold or when the average of a specified number of measurements for the monitored element drops below the lower threshold, perform the action defined by the action-type keyword. For example, if the upper threshold for react rtt threshold-type average 3 is configured as 5000 ms and the last three results of the operation are 6000, 6000, and 5000 ms, the average would be 6000 + 6000 + 5000 = 17000/3 = 5667, thus violating the 5000 ms upper threshold. The default number of 5 averaged measurements can be changed using the number-of-measurements argument. The valid range is from 1 to 16. This syntax is not available if the connectionLoss, timeout, or verifyError keyword is specified as the monitored element, because upper and lower thresholds do not apply to these options. |
threshold-type consecutive [occurrences] |
(Optional) When a threshold violation for the monitored element is met consecutively for a specified number of times, perform the action defined by the action-type keyword. The default number of 5 consecutive occurrences can be changed using the occurrences argument. The valid range is from 1 to 16. The occurrences value will appear in the output of the show ip sla reaction-configuration command as the "Threshold Count" value. |
threshold-type immediate |
(Optional) When a threshold violation for the monitored element is met, immediately perform the action defined by the action-type keyword. |
threshold-type never |
(Optional) Do not calculate threshold violations. This is the default threshold type. |
threshold-type xofy [x-value y-value] |
(Optional) When a threshold violation for the monitored element is met x number of times within the last y number of measurements ("x of y"), perform the action defined by the action-type keyword. The default is 5 for both the x and y values (xofy 5 5). The valid range for each value is from 1 to 16. The x-value will appear in the output of the show ip sla reaction-configuration command as the "Threshold Count" value, and the y-value will appear as the "Threshold Count2" value. |
threshold-value upper-threshold lower-threshold |
(Optional) Specifies the upper-threshold and lower-threshold values of the applicable monitored elements. See Table 15 in the "Usage Guidelines" section for a list of the default values. Note For MOS threshold values (react mos), the number is expressed in three digits representing ones, tenths, and hundredths. For example, to express a MOS threshold of 3.20, enter 320. The valid range is from 100 (1.00) to 500 (5.00). |
Command Default
IP SLAs proactive threshold monitoring is disabled.
Command Modes
Global configuration (config)
Command History
Usage Guidelines
You can configure the ip sla reaction-configuration command multiple times to enable proactive threshold monitoring for multiple elements, such as configuring thresholds for both destination-to-source packet loss and MOS for the same operation. However, disabling individual monitored elements is not supported. The no ip sla reaction-configuration command disables all proactive threshold monitoring configuration for the specified IP SLAs operation.
The keyword options for this command are not case sensitive. The keywords in online help for the action-type option and react monitored-element keyword and argument combinations contain uppercase letters to enhance readability only.
Not all elements can be monitored by all IP SLAs operations. If you attempt to configure an unsupported monitored-element, such as MOS for a UDP echo operation, the following message displays:
Invalid react option for the Probe type configured
Before Cisco IOS Release 15.1(1)T, valid online help was not available for this command. See Table 12 and Table 13 for a list of elements that are supported for each IP SLA operation.
In Cisco IOS Release 15.1(1)T and later releases, type shift + ? to display a list of supported elements for the IP SLAs operation being configured.
Return-trip time (RTT) reactions for jitter operations are triggered only at the end of the operation and use the latest value for the return-trip time (LatestRTT). SNMP traps for RTT for jitter operations are based on the average value for the return-trip time (RTTAvg) for the whole operation only and do not include return-trip time values for individual packets sent during the operation.
The connectionLoss trap is sent if the control connection is established and the operation is running, then the IP SLAs responder process stops, for example, if the no ip sla responder command is issued. This trap is supported only by operations that use the IPSLA control protocol to establish a control connection, such as udp-jitter and udp-echo. ICMP operations do not support connectionLoss traps.
lists the action or combination of actions that are supported when a threshold event for a monitored element occurs.
Use the snmp-server enable traps rtr or snmp-server enable traps syslog command to enable the sending of IP SLAs SNMP trap notifications.
Use the ip sla logging traps command to enable the generation of SNMP system logging messages specific to IP SLAs trap notifications.
Table 15 lists the default upper and lower thresholds for specific monitored elements.
To display the current threshold monitoring configuration settings for an IP SLAs operation, use the show ip sla configuration command.
Examples
The following example shows how to configure IP SLAs operation 10 (a UDP jitter operation) to send an SNMP logging trap when the MOS value exceeds 4.9 (best quality) or falls below 2.5 (poor quality):
Router(config)# ip sla reaction-configuration 10 react mos threshold-type immediate threshold-value 490 250 action-type trapOnly
Related Commands
ip sla reaction-trigger
To define a second Cisco IOS IP Service Level Agreements (SLAs) operation to make the transition from a pending state to an active state when one of the trigger action type options are defined with the ip sla reaction-configuration command, use the ip sla reaction-trigger command in global configuration mode. To remove the trigger combination, use the no form of this command.
ip sla reaction-trigger operation-number target-operation
no ip sla reaction-trigger operation
Syntax Description
Defaults
No trigger combination is defined.
Command Modes
Global configuration
Command History
Usage Guidelines
Triggers are usually used for diagnostics purposes and are not intended for use during normal operation conditions.
Examples
In the following example, a trigger action type is defined for IP SLAs operation 2. When operation 2 experiences certain user-specified threshold violation events while it is actively collecting statistical information, the operation state of IP SLAs operation 1 will be triggered to change from pending to active.
ip sla reaction-trigger 2 1
Related Commands
ip sla reset
To perform a shutdown and restart of the Cisco IOS IP Service Level Agreements (SLAs) engine, use the ip sla reset command in global configuration mode.
ip sla reset
Syntax Description
This command has no arguments or keywords.
Defaults
None
Command Modes
Global configuration
Command History
Usage Guidelines
The ip sla reset command stops all IP SLAs operations, clears IP SLAs configuration information, and returns the IP SLAs feature to the startup condition. This command does not reread the IP SLAs configuration stored in the startup configuration in NVRAM. You must retype the configuration or load a previously saved configuration file.
Note The ip sla reset command does not remove IP SLAs label switched path (LSP) Health Monitor configurations from the running configuration. Use the auto ip sla mpls-lsp-monitor reset command to remove LSP Health Monitor configurations from the running configuration.
Note Use the ip sla reset command only in extreme situations such as the incorrect configuration of a number of operations.
Examples
The following example shows how to reset the Cisco IOS IP SLAs engine, clearing all stored IP SLAs information and configuration:
ip sla reset
Related Commands
|
|
---|---|
ip sla restart |
Restarts a stopped IP SLAs operation. |
ip sla responder
To enable the Cisco IOS IP Service Level Agreements (SLAs) Responder for general IP SLAs operations, use the ip sla responder command in global configuration mode. To disable the IP SLAs Responder, use the no form of this command.
ip sla responder
no ip sla responder
Syntax Description
This command has no arguments or keywords.
Defaults
The IP SLAs Responder is disabled.
Command Modes
Global configuration
Command History
Usage Guidelines
This command is used on the destination device for IP SLAs operations to enable the sending and receiving of IP SLAs control packets. Enabling the IP SLAs Responder allows the generation of packet loss statistics on the device sending IP SLAs operations.
Prior to sending an operation packet to the IP SLAs Responder, the IP SLAs operation sends a control message to the IP SLAs Responder to enable the destination port.
The ip sla responder command is supported in IPv4 networks. This command can also be used when configuring an IP SLAs operation that supports IPv6 addresses.
Examples
The following example shows how to enable the IP SLAs Responder:
ip sla responder
Related Commands
ip sla responder auto-register
To configure a destination Cisco routing device or Cisco IP Service Level Agreements (SLAs) Responder to automatically register with the source upon configuration, use the ip sla responder auto-register command in global configuration mode. To disable automatic registration, use the no form of this command.
ip sla responder auto-register {source-ipaddress | source-hostname} [client-id client-id] [endpoint-list template-name] [retry-timer minutes]
no ip sla responder auto-register {source-ipaddress | source-hostname} [client-id client-id] [endpoint-list template-name] [retry-timer minutes]
Syntax Description
Command Default
The Cisco IP SLAs Responder does not automatically register with source.
Command Modes
Global configuration (config)
Command History
|
|
---|---|
15.1(1)T |
This command was introduced. |
Usage Guidelines
This command is required to allow the Cisco destination routing device or Cisco IP SLAs Responder to automatically register with the source and enable the source to automatically discover the endpoint.
Examples
The following example shows how to configure this command to enable auto discovery for configuring an auto IP SLAs endpoint list:
Destination
Router(config)# ip sla responder auto-register 10.1.1.23 endpoint-list autolist
Router(config)# exit
Router#
Source
Router(config)#ip sla auto discover
Router(config)#ip sla auto endpoint-list type ip autolist
Router(config-epl)#discover port 5000
Router(config-epl)#access-list 3
Router(config-term)#exit
Router# show ip sla auto endpoint-list
Endpoint-list Name: autolist
Description:
Auto Discover Parameters
Destination Port: 5000
Access-list: 3
Ageout: 3600 Measurement-retry: 3
1 endpoints are discovered for autolist
Related Commands
ip sla responder tcp-connect ipaddress
To enable the Cisco IOS IP Service Level Agreements (SLAs) Responder for TCP Connect operations, use the ip sla responder tcp-connect ipaddress command in global configuration mode. To disable the IP SLAs Responder, use the no form of this command.
ip sla responder tcp-connect ipaddress ip-address port port-number
no ip sla responder tcp-connect ipaddress ip-address port port-number
Syntax Description
ip-address |
Destination IP address. |
port port-number |
Specifies the destination port number. |
Defaults
The IP SLAs Responder is disabled.
Command Modes
Global configuration
Command History
Usage Guidelines
This command is used on the destination device for IP SLAs operations to enable the acceptance and return of TCP connection operation packets.
Examples
The following example shows how to enable the IP SLAs Responder for TCP connection operations:
ip sla responder tcp-connect ipaddress A.B.C.D port 1
Related Commands
|
|
---|---|
ip sla |
Begins configuration for an IP SLAs operation and enters IP SLA configuration mode. |
ip sla responder |
Enables the IP SLAs Responder for nonspecific IP SLAs operations. |
ip sla responder udp-echo ipaddress
To enable the Cisco IOS IP Service Level Agreements (SLAs) Responder for User Datagram Protocol (UDP) echo or jitter operations, use the ip sla responder udp-echo ipaddress command in global configuration mode. To disable the IP SLAs Responder, use the no form of this command.
ip sla responder udp-echo ipaddress ip-address port port-number
no ip sla responder udp-echo ipaddress ip-address port port-number
Syntax Description
ip-address |
Destination IP address. |
port port-number |
Specifies the destination port number. |
Command Default
The IP SLAs Responder is disabled.
Command Modes
Global configuration (config)
Command History
Usage Guidelines
This command is used on the destination device for IP SLAs operations to enable UDP echo and jitter (UDP+) operations with control disabled.
Examples
The following example shows how to enable the IP SLAs Responder for jitter operations:
ip sla responder udp-echo ipaddress A.B.C.D port 1
Related Commands
|
|
---|---|
ip sla |
Begins configuration for an IP SLAs operation and enters IP SLA configuration mode. |
ip sla responder |
Enables the IP SLAs Responder for nonspecific IP SLAs operations. |
ip sla restart
To restart a Cisco IOS IP Service Level Agreements (SLAs) operation, use the ip sla restart command in global configuration mode.
ip sla restart operation-number
Syntax Description
operation-number |
Number of the IP SLAs operation to restart. IP SLAs allows a maximum of 2000 operations. |
Defaults
None
Command Modes
Global configuration
Command History
Usage Guidelines
To restart an operation, the operation should be in an active state.
IP SLAs allows a maximum of 2000 operations.
This command does not have a no form.
Examples
The following example shows how to restart operation 12:
ip sla restart 12
Related Commands
|
|
---|---|
ip sla reset |
Clears all current IP SLAs statistics and configuration information from the router and resets the IP SLAs engine. |
ip sla schedule
To configure the scheduling parameters for a single Cisco IOS IP Service Level Agreements (SLAs) operation, use the ip sla schedule command in global configuration mode. To stop the operation and place it in the default state (pending), use the no form of this command.
ip sla schedule operation-number [life {forever | seconds}] [start-time {hh:mm[:ss] [month day | day month] | pending | now | after hh:mm:ss}] [ageout seconds] [recurring]
no ip sla schedule operation-number
Syntax Description
Defaults
The operation is placed in a pending state (that is, the operation is enabled but not actively collecting information).
Command Modes
Global configuration
Command History
Usage Guidelines
After you schedule the operation with the ip sla schedule command, you cannot change the configuration of the operation. To change the configuration of the operation, use the no form of the ip sla global configuration command and reenter the configuration information.
If the operation is in a pending state, you can define the conditions under which the operation makes the transition from pending to active with the ip sla reaction-trigger and ip sla reaction-configuration global configuration commands. When the operation is in an active state, it immediately begins collecting information.
The following time line shows the age-out process of the operation:
W----------------------X----------------------Y----------------------Z
where:
•W is the time the operation was configured with the ip sla global configuration command.
•X is the start time or start of life of the operation (that is, when the operation became "active").
•Y is the end of life as configured with the ip sla schedule global configuration command (life seconds have counted down to zero).
•Z is the age out of the operation.
Age out starts counting down at W and Y, is suspended between X and Y, and is reset to its configured size at Y.
The operation to can age out before it executes (that is, Z can occur before X). To ensure that this does not happen, configure the difference between the operation's configuration time and start time (X and W) to be less than the age-out seconds.
Note The total RAM required to hold the history and statistics tables is allocated at the time of scheduling the IP SLAs operation. This prevents router memory problems when the router gets heavily loaded and lowers the amount of overhead an IP SLAs operation causes on a router when it is active.
The recurring keyword is supported only for scheduling single IP SLAs operations. You cannot schedule multiple IP SLAs operations using the ip sla schedule command. The life value for a recurring IP SLAs operation should be less than one day. The ageout value for a recurring operation must be "never" (which is specified with the value 0), or the sum of the life and ageout values must be more than one day. If the recurring option is not specified, the operations are started in the existing normal scheduling mode.
The ip sla schedule command is supported in IPv4 networks. This command can also be used when configuring an IP SLAs operation that supports IPv6 addresses.
Examples
In the following example, operation 25 begins actively collecting data at 3:00 p.m. on April 5. This operation will age out after 12 hours of inactivity, which can be before it starts or after it has finished with its life. When this operation ages out, all configuration information for the operation is removed (that is, the configuration information is no longer in the running configuration in RAM).
ip sla schedule 25 life 43200 start-time 15:00 apr 5 ageout 43200
In the following example, operation 1 begins collecting data after a 5-minute delay:
ip sla schedule 1 start-time after 00:05:00
In the following example, operation 3 begins collecting data immediately and is scheduled to run indefinitely:
ip sla schedule 3 start-time now life forever
In the following example, operation 15 begins automatically collecting data every day at 1:30 a.m.:
ip sla schedule 15 start-time 01:30:00 recurring
Related Commands
life
To specify the lifetime characteristic in an auto IP Service Level Agreements (SLAs) scheduler, use the life command in IP SLA auto-measure schedule configuration mode. To return to the default, use the no form of this command.
life {forever | seconds}
no life
Syntax Description
forever |
Runs operation indefinitely. |
seconds |
Length of time the operation actively collects information, in seconds (sec). Range is from 1 to 2147483647. Default is 3600. |
Command Default
Auto IP SLAs operation actively collects information for 3600 sec.
Command Modes
IP SLA auto-measure schedule configuration (config-am-schedule)
Command History
|
|
---|---|
15.1(1)T |
This command was introduced. |
Usage Guidelines
This command changes the default configuration for life (3600 sec) in an auto IP SLA scheduler to the specified value.
Examples
The following example shows how to configure an auto IP SLAs scheduler that will cause an auto IP SLAs operation to actively collect data at 3:00 p.m. on April 5. The operation will age out after 12 hours of inactivity, which can be before it starts or after it has finished its life. When the operation ages out, all configuration information for the operation is removed from the running configuration in RAM.
Router(config)#ip sla auto schedule apr5
Router(config-am-schedule)#ageout 43200
Router(config-am-schedule)#frequency 70
Router(config-am-schedule)#life 43200
Router(config-am-schedule)#probe-interval 1500
Router(config-am-schedule)#start-time 15:00 apr 5
Router(config-am-schedule)#end
Router#
Router# show ip sla auto schedule apr5
Group sched-id: apr5
Probe Interval (ms) : 1500
Group operation frequency (sec): 70
Status of entry (SNMP RowStatus): Active
Next Scheduled Start Time: P15:00 apr 5
Life (sec): 43200
Entry Ageout (sec): 43200
Router#
Related Commands
lives-of-history-kept
Note Effective with Cisco IOS Release 12.4(4)T, 12.2(33)SRB, 12.2(33)SB, and 12.2(33)SXI, the lives-of-history-kept command is replaced by the history lives-kept command. See the history lives-kept command for more information.
To set the number of lives maintained in the history table for a Cisco IOS IP Service Level Agreements (SLAs) operation, use the lives-of-history-kept command in the appropriate submode of IP SLA monitor configuration mode. To return to the default value, use the no form of this command.
lives-of-history-kept lives
no lives-of-history-kept
Syntax Description
lives |
Number of lives maintained in the history table for the operation. If you specify 0 lives, history is not collected for the operation. |
Defaults
0 lives
Command Modes
DHCP configuration (config-sla-monitor-dhcp)
DLSw configuration (config-sla-monitor-dlsw)
DNS configuration (config-sla-monitor-dns)
FTP configuration (config-sla-monitor-ftp)
HTTP configuration (config-sla-monitor-http)
ICMP echo configuration (config-sla-monitor-echo)
ICMP path echo configuration (config-sla-monitor-pathEcho)
ICMP path jitter configuration (config-sla-monitor-pathJitter)
TCP connect configuration (config-sla-monitor-tcp)
UDP echo configuration (config-sla-monitor-udp)
VoIP configuration (config-sla-monitor-voip)
Command History
Usage Guidelines
The following rules apply to the lives-of-history-kept command:
•The number of lives you can specify is dependent on the type of operation you are configuring.
•The default value of 0 lives means that history is not collected for the operation.
•When the number of lives exceeds the specified value, the history table wraps (that is, the oldest information is replaced by newer information).
•When an operation makes a transition from a pending to active state, a life starts. When the life of an operation ends, the operation makes a transition from an active to pending state.
Note The lives-of-history-kept command does not support the IP SLAs User Datagram Protocol (UDP) jitter operation.
An IP SLAs operation can collect history and capture statistics. By default, the history for an IP SLAs operation is not collected. If history is collected, each history bucket contains one or more history entries from the operation. When the operation type is ICMP path echo, an entry is created for each hop along the path that the operation takes to reach its destination. The type of entry stored in the history table is controlled by the filter-for-history command. The total number of entries stored in the history table is controlled by the combination of the samples-of-history-kept, buckets-of-history-kept, and lives-of-history-kept commands.
To disable history collection, use the no lives-of-history-kept command rather than the filter-for-history none command. The no lives-of-history-kept command disables history collection before an IP SLAs operation is attempted. The filter-for-history command checks for history inclusion after the operation attempt is made.
Note You must configure the type of IP SLAs operation (such as User Datagram Protocol [UDP] jitter or Internet Control Message Protocol [ICMP] echo) before you can configure any of the other parameters of the operation.
Examples
The following example shows how to maintain the history for five lives of IP SLAs ICMP echo operation 1.
ip sla monitor 1
type echo protocol ipIcmpEcho 172.16.1.176
lives-of-history-kept 5
!
ip sla monitor schedule 1 life forever start-time now
Related Commands
lsp-selector
To specify the local host IP address used to select the label switched path (LSP) for a Cisco IOS IP Service Level Agreements (SLAs) LSP Health Monitor operation, use the lsp-selector command in auto IP SLA MPLS parameters configuration mode. To return to the default value, use the no form of this command.
lsp-selector ip-address
no lsp-selector ip-address
Syntax Description
ip-address |
Specifies a local host IP address used to select the LSP. |
Command Default
The local host IP address used to select the LSP is 127.0.0.0.
Command Modes
Auto IP SLA MPLS parameters configuration (config-auto-ip-sla-mpls-params)
Command History
Usage Guidelines
This command is used to force an IP SLAs operation to use a specific LSP to obtain its response time measurement. This option is useful if there are equal-cost multipaths between the source Provider Edge (PE) router and the Border Gateway Protocol (BGP) next hop neighbor.
You must configure the type of LSP Health Monitor operation (such as LSP ping) before you can configure any of the other parameters of the operation.
Examples
The following example shows how to configure operation parameters, proactive threshold monitoring, and scheduling options using the LSP Health Monitor. In this example, LSP Health Monitor operation 1 is configured to automatically create IP SLAs LSP ping operations for all Border Gateway Protocol (BGP) next hop neighbors in use by all VPN routing and forwarding (VRF) instances associated with the source PE router. As specified in the example configuration, IP address 127.0.0.1 is the local host IP address chosen to select the LSP for obtaining response time measurements.
mpls discovery vpn interval 60
mpls discovery vpn next-hop
!
auto ip sla mpls-lsp-monitor 1
type echo ipsla-vrf-all
timeout 1000
scan-interval 1
secondary-frequency connection-loss 10
secondary-frequency timeout 10
delete-scan-factor 2
lsp-selector 127.0.0.1
!
auto ip sla mpls-lsp-monitor reaction-configuration 1 react connectionLoss threshold-type consecutive 3 action-type trapOnly
auto ip sla mpls-lsp-monitor reaction-configuration 1 react timeout threshold-type consecutive 3 action-type trapOnly
ip sla logging traps
!
auto ip sla mpls-lsp-monitor schedule 1 schedule-period 60 start-time now
Related Commands
|
|
---|---|
auto ip sla mpls-lsp-monitor |
Begins configuration for an IP SLAs LSP Health Monitor operation and enters auto IP SLA MPLS configuration mode. |
lsp-selector-base
To specify the base IP address used to select the label switched paths (LSPs) belonging to the LSP discovery groups of a Cisco IOS IP Service Level Agreements (SLAs) LSP Health Monitor operation, use the lsp-selector-base command in auto IP SLA MPLS LSP discovery parameters configuration mode. To return to the default value, use the no form of this command.
lsp-selector-base ip-address
no lsp-selector-base
Syntax Description
ip-address |
Base IP address used to select the LSPs within an LSP discovery group. The default IP address is 127.0.0.0. |
Command Default
The default base IP address is 127.0.0.0.
Command Modes
Auto IP SLA MPLS LSP discovery parameters configuration (config-auto-ip-sla-mpls-lpd-params)
Command History
|
|
---|---|
12.2(31)SB2 |
This command was introduced. |
12.2(33)SRB |
This command was integrated into Cisco IOS Release 12.2(33)SRB. |
Usage Guidelines
Each equal-cost multipath belonging to an LSP discovery group is uniquely identified by the following three parameters:
•Local host IP address of the LSP selector
•Outgoing interface
•Downstream MPLS label stack number
Use the path-discover command to enable the LSP discovery option for an IP SLAs LSP Health Monitor operation and enter auto IP SLA MPLS LSP discovery parameters configuration mode.
Examples
The following example shows how to configure operation parameters, proactive threshold monitoring, and scheduling options using the LSP Health Monitor. In this example, the LSP discovery option is enabled for LSP Health Monitor operation 1. Operation 1 is configured to automatically create IP SLAs LSP ping operations for the equal-cost multipaths to all Border Gateway Protocol (BGP) next hop neighbors in use by all VPN routing and forwarding (VRF) instances associated with the source Provider Edge (PE) router. The base IP address used to select the LSPs within the LSP discovery groups is set to 127.0.0.2.
auto ip sla mpls-lsp-monitor 1
type echo ipsla-vrf-all
path-discover
!
maximum-sessions 2
session-timeout 60
lsp-selector-base 127.0.0.2
interval 2
timeout 4
force-explicit-null
hours-of-statistics-kept 1
scan-period 30
!
auto ip sla mpls-lsp-monitor schedule 1 schedule-period 60 frequency 100 start-time now
auto ip sla mpls-lsp-monitor reaction-configuration 1 react lpd tree-trace action-type trapOnly
auto ip sla mpls-lsp-monitor reaction-configuration 1 react lpd lpd-group retry 3 action-type trapOnly
Related Commands
lsr-path
To define a loose source routing (LSR) path for a Cisco IOS IP Service Level Agreements (SLAs) operation, use the lsr-path command in the appropriate submode of IP SLA configuration or IP SLA configuration mode. To remove the definition, use the no form of this command.
lsr-path {hostname1 | ip-address1} [[hostname2 | ip-address2]...[hostname8 | ip-address8]]
no lsr-path
Syntax Description
Defaults
LSR path is disabled.
Command Modes
IP SLA Configuration
ICMP path echo configuration (config-ip-sla-pathEcho)
ICMP path jitter configuration (config-ip-sla-pathJitter)
IP SLA Monitor Configuration
ICMP path echo configuration (config-sla-monitor-pathEcho)
ICMP path jitter configuration (config-sla-monitor-pathJitter)
Note The configuration mode varies depending on the Cisco IOS release you are running and the operation type configured. See the "Usage Guidelines" section for more information.
Command History
Usage Guidelines
The maximum number of hops available is eight when an LSR path is configured.
Note This command is supported by the IP SLAs Internet Control Message Protocol (ICMP) path echo and path jitter operations only.
IP SLAs Operation Configuration Dependence on Cisco IOS Release
The Cisco IOS command used to begin configuration for an IP SLAs operation varies depending on the Cisco IOS release you are running (see Table 16). You must configure the type of IP SLAs operation (such as User Datagram Protocol [UDP] jitter or Internet Control Message Protocol [ICMP] echo) before you can configure any of the other parameters of the operation.
The configuration mode for the lsr-path command varies depending on the Cisco IOS release you are running (see Table 16) and the operation type configured. For example, if you are running Cisco IOS Release 12.4 and the ICMP path echo operation type is configured, you would enter the lsr-path command in ICMP path echo configuration mode (config-sla-monitor-pathEcho) within IP SLA monitor configuration mode.
Examples
In the following examples, the LSR path is defined for IP SLAs ICMP path echo operation 1. The target destination for the operation is at 172.16.1.176. The first hop on the LSR path is 172.18.4.149. The second hop on the LSR path is 172.18.16.155. Note that the Cisco IOS command used to begin configuration for an IP SLAs operation varies depending on the Cisco IOS release you are running (see Table 16).
IP SLA Configuration
ip sla 1
path-echo 172.16.1.176
lsr-path 172.18.4.149 172.18.26.155
!
ip sla schedule 1 life forever start-time now
IP SLA Monitor Configuration
ip sla monitor 1
type pathEcho protocol ipIcmpEcho 172.16.1.176
lsr-path 172.18.4.149 172.18.26.155
!
ip sla monitor schedule 1 life forever start-time now
Related Commands
maximum-sessions
To specify the maximum number of Border Gateway Protocol (BGP) next hop neighbors that can be concurrently undergoing label switched path (LSP) discovery for a single Cisco IOS IP Service Level Agreements (SLAs) LSP Health Monitor operation, use the maximum-sessions command in auto IP SLA MPLS LSP discovery parameters configuration mode. To return to the default value, use the no form of this command.
maximum-sessions number
no maximum-sessions
Syntax Description
number |
Maximum number of BGP next hop neighbors that can be concurrently undergoing LSP discovery. The default is 1. |
Command Default
By default, the number argument is set to 1.
Command Modes
Auto IP SLA MPLS LSP discovery parameters configuration (config-auto-ip-sla-mpls-lpd-params)
Command History
|
|
---|---|
12.2(31)SB2 |
This command was introduced. |
12.2(33)SRB |
This command was integrated into Cisco IOS Release 12.2(33)SRB. |
Usage Guidelines
Use the path-discover command to enable the LSP discovery option for an IP SLAs LSP Health Monitor operation and enter auto IP SLA MPLS LSP discovery parameters configuration mode.
Examples
The following example shows how to configure operation parameters, proactive threshold monitoring, and scheduling options using the LSP Health Monitor. In this example, the LSP discovery option is enabled for LSP Health Monitor operation 1. Operation 1 is configured to automatically create IP SLAs LSP ping operations for the equal-cost multipaths to all Border Gateway Protocol (BGP) next hop neighbors in use by all VPN routing and forwarding (VRF) instances associated with the source Provider Edge (PE) router. The maximum number of LSP discovery processes allowed to run concurrently is set to 2.
auto ip sla mpls-lsp-monitor 1
type echo ipsla-vrf-all
path-discover
!
maximum-sessions 2
session-timeout 60
interval 2
timeout 4
force-explicit-null
hours-of-statistics-kept 1
scan-period 30
!
auto ip sla mpls-lsp-monitor schedule 1 schedule-period 60 frequency 100 start-time now
!
auto ip sla mpls-lsp-monitor reaction-configuration 1 react lpd tree-trace action-type trapOnly
auto ip sla mpls-lsp-monitor reaction-configuration 1 react lpd lpd-group retry 3 action-type trapOnly
Related Commands
measurement-retry
To specify the number of times the endpoints belonging to an auto IP SLAs endpoint list are retested when an operation fails, use the measurement-retry command in IP SLAs endpoint-list auto-discovery configuration mode. To return to the default, use the no form of this command.
measurement-retry number-of-retries
no measurement-retry
Syntax Description
number-of-retries |
Range is from 0 to 65535. Default is 0. |
Command Default
No attempt to retry a failed operation is made.
Command Modes
IP SLA endpoint-list auto-discovery configuration (config-epl-disc)
Command History
|
|
---|---|
15.1(1)T |
This command was introduced. |
Usage Guidelines
This command specifies the number of times an operation associated with an auto IP SLAs endpoint list is retried when a failure is detected.
This option is supported only by auto IP SLAs endpoint lists that are configured using auto discovery in Cisco IOS IP SLAs Engine 3.0.
Examples
The following example shows how to configure an auto IP SLAs endpoint lists of endpoints using auto discovery:
Router(config)#ip sla auto discover
Router(config)#ip sla auto endpoint-list type ip autolist
Router(config-epl)#discover port 5000
Router(config-epl)#measurement-retry 3
Router(config-epl)#access-list 3
Router(config-epl)#exit
Router# show ip sla auto endpoint-list
Endpoint-list Name: man1
Description: testing manual build
ip-address 10.1.1.1-7 port 23
ip-address 10.1.1.9,10.1.1.15,10.1.1.23 port 23
Endpoint-list Name: autolist
Description:
Auto Discover Parameters
Destination Port: 5000
Access-list: 3
Ageout: 3600 Measurement-retry: 3
0 endpoints are discovered for autolist
Related Commands
|
|
---|---|
show ip sla auto endpoint-list |
Displays configuration including default values of auto IP SLAs endpoint lists. |
mpls discovery vpn interval
To specify the time interval at which routing entries that are no longer valid are removed from the Border Gateway Protocol (BGP) next hop neighbor discovery database of a Multiprotocol Label Switching (MPLS) Virtual Private Network (VPN), use the mpls discovery vpn interval command in global configuration mode. To return to the default scan interval, use the no form of this command.
mpls discovery vpn interval seconds
no mpls discovery vpn interval
Syntax Description
seconds |
Specifies the time interval (in seconds) at which routing entries that are no longer valid are removed from the BGP next hop neighbor discovery database of an MPLS VPN. The default is 300. |
Command Default
The default time interval is 300 seconds.
Command Modes
Global configuration
Command History
Usage Guidelines
When the BGP next hop neighbor discovery process is enabled (using the mpls discovery vpn next-hop command), a database of BGP next hop neighbors in use by any VPN routing or forwarding instance (VRF) associated with the source Provider Edge (PE) router is generated based on information from the local VRF and global routing tables. As routing updates are received, new BGP next hop neighbors are added immediately to the database. However, BGP next hop neighbors (that are no longer valid) are only removed from the database periodically as defined by the user (using the mpls discovery vpn interval command).
The BGP next hop neighbor discovery process is used by the Cisco IOS IP Service Level Agreements (SLAs) LSP Health Monitor feature.
Note The default interval of time that BGP neighbor statistics are updated is different for the IP SLAs LSP Health Monitor database and the BGP next hop neighbor discovery database. Use the scan-interval command to set the timer for the IP SLAs LSP Health Monitor database. Use the mpls discovery vpn interval command to set the timer for the BGP next hop neighbor discovery database.
Examples
The following example shows how to enable the MPLS VPN BGP next hop neighbor discovery process and specify 60 seconds as the time interval at which routing entries that are no longer valid are removed from the BGP next hop neighbor discovery database of an MPLS VPN:
mpls discovery vpn interval 60
mpls discovery vpn next-hop
Related Commands
mpls discovery vpn next-hop
To enable the Multiprotocol Label Switching (MPLS) Virtual Private Network (VPN) Border Gateway Protocol (BGP) next hop neighbor discovery process, use the mpls discovery vpn next-hop command in global configuration mode. To disable the discovery process, use the no form of this command.
mpls discovery vpn next-hop
no mpls discovery vpn next-hop
Syntax Description
This command has no arguments or keywords.
Command Default
The BGP next hop neighbor discovery process is disabled.
Command Modes
Global configuration
Command History
Usage Guidelines
When the BGP next hop neighbor discovery process is enabled, a database of BGP next hop neighbors in use by any VPN routing or forwarding instance (VRF) associated with the source Provider Edge (PE) router is generated based on information from the local VRF and global routing tables. As routing updates are received, new BGP next hop neighbors are added immediately to the database. However, BGP next hop neighbors (that are no longer valid) are only removed from the database periodically as defined by the user (using the mpls discovery vpn interval command in global configuration mode).
The mpls discovery vpn next-hop command is automatically enabled when an IP Service Level Agreements (SLAs) LSP Health Monitor operation is enabled. However, to disable the BGP next hop neighbor discovery process, you must use the no form of this command.
Examples
The following example shows how to enable the MPLS VPN BGP next hop neighbor discovery process and specify 60 seconds as the time interval at which routing entries that are no longer valid are removed from the BGP next hop neighbor discovery database of an MPLS VPN:
mpls discovery vpn interval 60
mpls discovery vpn next-hop
Related Commands
mpls lsp ping ipv4
To manually configure an individual Cisco IOS IP Service Level Agreements (SLAs) label switched path (LSP) ping IPv4 operation, use the mpls lsp ping ipv4 command in IP SLA configuration mode.
mpls lsp ping ipv4 destination-address destination-mask [force-explicit-null] [lsp-selector ip-address] [src-ip-addr source-address] [reply {dscp dscp-value | mode {ipv4 | router-alert}}]
Syntax Description
Command Default
No IP SLAs operation type is configured for the operation being configured.
Command Modes
IP SLA configuration
Command History
Usage Guidelines
You must configure the type of IP SLAs operation (such as LSP ping) before you can configure any of the other parameters of the operation. To change the operation type of an existing IP SLAs operation, you must first delete the IP SLAs operation (using the no ip sla global configuration command) and then reconfigure the operation with the new operation type.
The lsp-selector keyword is used to force an IP SLAs operation to use a specific LSP to obtain its response time measurement. This option is useful if there are multiple equal cost paths between Provider Edge (PE) routers.
Examples
The following example shows how to manually configure operation parameters, reaction conditions, and scheduling options for IP SLAs LSP ping operation 1:
ip sla 1
mpls lsp ping ipv4 192.168.1.4 255.255.255.255 lsp-selector 127.1.1.1
frequency 120
secondary-frequency timeout 30
exit
!
ip sla reaction-configuration 1 react connectionLoss threshold-type consecutive 3 action-type trapOnly
ip sla reaction-configuration 1 react timeout threshold-type consecutive 3 action-type trapOnly
ip sla logging traps
!
ip sla schedule 1 start-time now life forever
Related Commands
|
|
---|---|
ip sla |
Begins configuration for an IP SLAs operation and enters IP SLA configuration mode. |
mpls lsp ping pseudowire
To configure an IP Service Level Agreements (SLAs) Multiprotocol Label Switching (MPLS) Pseudo-Wire Emulation Edge-to-Edge (PWE3) services via Virtual Circuit Connectivity Verification (VCCV) operation and enter VCCV configuration mode, use the mpls lsp ping pseudowire command in IP SLA configuration mode.
mpls lsp ping pseudowire peer-ipaddr vc-id [source-ipaddr source-ipaddr]
Syntax Description
Command Default
No IP SLAs operation type is configured for the operation being configured.
Command Modes
IP SLA configuration (config-ip-sla)
Command History
|
|
---|---|
12.2(33)SRC |
This command was introduced. |
12.2(33)SB |
This command was integrated into Cisco IOS Release 12.2(33)SB. |
Usage Guidelines
Use the mpls lsp ping pseudowire command to configure a single IP SLAs VCCV operation, which checks MPLS label switched path (LSP) connectivity across an Any Transport over MPLS (AToM) VC by sending a series of pseudo-wire ping operations to the specified peer PE router. The IP SLA maintains pseudo-wire ping statistics for the operation, such as Round Trip Time (RTT). The optional source-ipaddr keyword is used to specify the source-ipaddr argument as the source IP address for the request originator.
To configure a faster measurement frequency (secondary frequency) to which an IP SLAs VCCV operation should change when a connection-loss or timeout condition occurs, use the secondary-frequency command in VCCV configuration mode.
To configure proactive threshold monitoring of an IP SLAs VCCV operation, configure actions to occur based on events under the control of that operation and enable Simple Network Management Protocol (SNMP) logging traps for that operation:
•To configure actions to occur based on events under the control of an IP SLAs operation, including the sending of SNMP logging trap when a specified violation type occurs for the monitored operation, use the ip sla reaction-configuration command in global configuration mode.
•To enable the generation of SNMP system logging messages specific to IP SLAs trap notifications, use the ip sla logging traps command in global configuration mode.
When these commands are used to configure continuous monitoring of PWE3 services, an IP SLAs VCCV operation can send out an SNMP trap if RTT threshold violations occur, if the connection is lost, or if a response times out.
To schedule an IP SLAs VCCV operation, use the ip sla schedule command in global configuration mode.
To display configuration values including all defaults for all IP SLAs operations or a specified operation, use the show ip sla configuration command. To display the current operational status and statistics for all IP SLAs operations or a specified operation, use the show ip sla statistics command. To display the aggregated statistical errors and distribution information for all IP SLAs operations or a specified operation, use the show ip sla statistics aggregated command. To display the reaction settings for all IP SLAs operations or a specified operation, use the show ip sla reaction-configuration command.
Examples
The following example shows how to manually configure operation parameters, reaction conditions, and scheduling options for IP SLAs VCCV operation 777.
Note In this example, a VC with the identifier 123 has already been established between the PE router and its peer at IP address 192.168.1.103.
ip sla 777
mpls lsp ping pseudowire 192.168.1.103 123
exp 5
frequency 120
secondary-frequency timeout 30
tag testgroup
threshold 6000
timeout 7000
exit
!
ip sla reaction-configuration 777 react rtt threshold-value 6000 3000 threshold-type immediate 3 action-type traponly
ip sla reaction-configuration 777 react connectionLoss threshold-type immediate action-type traponly
ip sla reaction-configuration 777 react timeout threshold-type consecutive 3 action-type traponly
ip sla logging traps
!
ip sla schedule 777 life forever start-time now
Related Commands exit
mpls lsp trace ipv4
To manually configure an individual Cisco IOS IP Service Level Agreements (SLAs) label switched path (LSP) traceroute IPv4 operation, use the mpls lsp trace ipv4 command in IP SLA configuration mode.
mpls lsp trace ipv4 destination-address destination-mask [force-explicit-null] [lsp-selector ip-address] [src-ip-addr source-address] [reply {dscp dscp-value | mode {ipv4 | router-alert}}]
Syntax Description
Command Default
No IP SLAs operation type is configured for the operation being configured.
Command Modes
IP SLA configuration
Command History
Usage Guidelines
You must configure the type of IP SLAs operation (such as LSP trace) before you can configure any of the other parameters of the operation. To change the operation type of an existing IP SLAs operation, you must first delete the IP SLAs operation (using the no ip sla global configuration command) and then reconfigure the operation with the new operation type.
Note This command supports only single path connectivity measurements between the source PE router and associated BGP next hop neighbors.
The lsp-selector keyword is used to force an IP SLAs operation to use a specific LSP to obtain its response time measurement. This option is useful if there are multiple equal cost paths between Provider Edge (PE) routers.
Examples
The following example shows how to manually configure operation parameters, reaction conditions, and scheduling options for IP SLAs LSP traceroute operation 1:
ip sla 1
mpls lsp trace ipv4 192.168.1.4 255.255.255.255 lsp-selector 127.1.1.1
frequency 120
exit
!
ip sla reaction-configuration 1 react connectionLoss threshold-type consecutive 3 action-type trapOnly
ip sla reaction-configuration 1 react timeout threshold-type consecutive 3 action-type trapOnly
ip sla logging traps
!
ip sla schedule 1 start-time now life forever
Related Commands
|
|
---|---|
ip sla |
Begins configuration for an IP SLAs operation and enters IP SLA configuration mode. |
num-packets
To specify the number of packets for a jitter operation in an auto IP Service Level Agreements (SLAs) operation template, use the num-packets command in the appropriate submode of the IP SLA template parameters configuration mode. To return to the default, use the no form of this command.
num-packets packet-number
no num-packets
Syntax Description
packet-number |
Number of packets to be sent in each operation. Range is 1 to 60000. Default is 10 per operation. |
Command Default
Default is 10 packets.
Command Modes
IP SLA Template Parameters Configuration
ICMP jitter configuration (config-icmp-jtr-params)
UDP jitter configuration (config-udp-jtr-params)
Command History
|
|
---|---|
15.1(1)T |
This command was introduced. |
Usage Guidelines
This command changes the number of packets sent during a jitter operation from the default (10) to the specified number of packets.
You must configure the type of IP SLAs operation, such as User Datagram Protocol (UDP) jitter or ICMP jitter, before you can configure any other parameters of the operation.
Before you can use this command to configure auto IP SLAs operation templates, you must enter the parameters command in IP SLA template configuration mode.
Examples
The following example shows how to configure an auto IP SLAs operation template for an ICMP jitter operation to change the number of packets from the default to 20 packets:
Router(config)#ip sla auto template type ip icmp-jitter 1
Router(config-tplt-icmp-jtr)#parameters
Router(config-icmp-jtr-params)#num-packets 20
Router(config-icmp-jtr-params)#end
Router# show ip sla auto template type ip icmp-jitter
IP SLAs Auto Template: 1
Measure Type: icmp-jitter
Description:
IP options:
Source IP: 0.0.0.0
VRF: TOS: 0x0
Operation Parameters:
Number of Packets: 20 Inter packet interval: 20
Timeout: 5000 Threshold: 5000
Statistics Aggregation option:
Hours of statistics kept: 2
Statistics Distributions options:
Distributions characteristics: RTT
Distributions bucket size: 20
Max number of distributions buckets: 1
Reaction Configuration: None
Related Commands
operation-packet priority
To specify the packet priority in a Cisco IOS IP Service Level Agreements (SLAs) operation template, use the operation-packet priority command in the appropriate submode of IP SLA configuration or IP SLA template parameters configuration mode. To return to the default value, use the no form of this command.
operation-packet priority {normal | high}
no operation-packet priority
Syntax Description
normal |
Specifies that the packet priority is normal. Default is normal. |
high |
Specifies that the packet priority is high. |
Command Default
Packet priority is normal.
Command Modes
IP SLA Configuration
UDP jitter configuration (config-ip-sla-jitter)
IP SLA Template Parameters Configuration
UDP jitter configuration (config-udp-ech-params)
Command History
Usage Guidelines
Increasing the packet priority of an IP SLAs operation can reduce the delay time for the packets in the queue.
This command is supported by the IP SLAs User Datagram Protocol (UDP) jitter operation only.
Before you can use this command to configure auto IP SLAs operation templates, you must enter the parameters command in IP SLA template configuration mode.
Examples
The following examples show how to enable microsecond precision, configure the Network Time Protocol (NTP) synchronization offset tolerance to 10 percent, and set the packet priority to high for an IP SLAs UDP jitter operation:
IP SLA Configuration
ip sla 1
udp-jitter 205.199.199.2 dest-port 9006
precision microseconds
clock-tolerance ntp oneway percent 10
operation-packet priority high
frequency 300
!
ip sla schedule 1 life forever start-time after 00:00:06
IP SLA Template Parameters Configuration
Router(config)# ip sla auto template type ip udp-jitter 1
Router(config-udp-jtr-tplt)# parameters
Router(config-udp-jtr-params)# precision microseconds
Router(config-udp-jtr-params)# clock-tolerance ntp oneway percent 10
Router(config-udp-jtr-params)# operation-packet priority high
Router(config-udp-jtr-params)# end
Router#
Router# show ip sla auto template type ip udp-jitter
IP SLAs Auto Template: 1
Measure Type: udp-jitter (control enabled)
Description:
IP options:
Source IP: 0.0.0.0 Source Port: 0
VRF: TOS: 0x0
Operation Parameters:
Request Data Size: 32 Verify Data: false
Number of Packets: 10 Inter packet interval: 20
Timeout: 5000 Threshold: 5000
Granularity: usec Operation packet priority: high
NTP Sync Tolerance: 10 percent
Statistics Aggregation option:
Hours of statistics kept: 2
Statistics Distributions options:
Distributions characteristics: RTT
Distributions bucket size: 20
Max number of distributions buckets: 1
Reaction Configuration: None
Related Commands
owner
To configure the Simple Network Management Protocol (SNMP) owner of a Cisco IOS IP Service Level Agreements (SLAs) operation, use the owner command in the appropriate submode of IP SLA configuration, IP SLA auto Ethernet configuration, or IP SLA monitor configuration mode. To return to the default value, use the no form of this command.
owner text
no owner
Syntax Description
text |
Name of the SNMP owner from 0 to 255 ASCII characters. |
Defaults
No owner is specified.
Command Modes
IP SLA Configuration
DHCP configuration (config-ip-sla-dhcp)
DLSw configuration (config-ip-sla-dlsw)
DNS configuration (config-ip-sla-dns)
Ethernet echo (config-ip-sla-ethernet-echo)
Ethernet jitter (config-ip-sla-ethernet-jitter)
FTP configuration (config-ip-sla-ftp)
HTTP configuration (config-ip-sla-http)
ICMP echo configuration (config-ip-sla-echo)
ICMP jitter configuration (config-ip-sla-icmpjitter)
ICMP path echo configuration (config-ip-sla-pathEcho)
ICMP path jitter configuration (config-ip-sla-pathJitter)
TCP connect configuration (config-ip-sla-tcp)
UDP echo configuration (config-ip-sla-udp)
UDP jitter configuration (config-ip-sla-jitter)
VCCV configuration (config-sla-vccv)
VoIP configuration (config-ip-sla-voip)
IP SLA Auto Ethernet Configuration
Ethernet parameters configuration (config-ip-sla-ethernet-params)
IP SLA Monitor Configuration
DHCP configuration (config-sla-monitor-dhcp)
DLSw configuration (config-sla-monitor-dlsw)
DNS configuration (config-sla-monitor-dns)
FTP configuration (config-sla-monitor-ftp)
HTTP configuration (config-sla-monitor-http)
ICMP echo configuration (config-sla-monitor-echo)
ICMP path echo configuration (config-sla-monitor-pathEcho)
ICMP path jitter configuration (config-sla-monitor-pathJitter)
TCP connect configuration (config-sla-monitor-tcp)
UDP echo configuration (config-sla-monitor-udp)
UDP jitter configuration (config-sla-monitor-jitter)
VoIP configuration (config-sla-monitor-voip)
Note The configuration mode varies depending on the Cisco IOS release you are running and the operation type configured. See the "Usage Guidelines" section for more information.
Command History
Usage Guidelines
The owner name contains one or more of the following: ASCII form of the network management station's transport address, network management station name (that is, the domain name), and network management personnel's name, location, or phone number. In some cases, the agent itself will be the owner of the operation. In these cases, the name can begin with "agent."
The owner command is supported in IPv4 networks. This command is also supported in IPv6 networks when configuring an IP SLAs operation that supports IPv6 addresses.
IP SLAs Operation Configuration Dependence on Cisco IOS Release
The Cisco IOS command used to begin configuration for an IP SLAs operation varies depending on the Cisco IOS release you are running (see Table 17). You must configure the type of IP SLAs operation (such as User Datagram Protocol [UDP] jitter or Internet Control Message Protocol [ICMP] echo) before you can configure any of the other parameters of the operation.
The configuration mode for the owner command varies depending on the Cisco IOS release you are running (see Table 17) and the operation type configured. For example, if you are running Cisco IOS Release 12.4 and the ICMP echo operation type is configured, you would enter the owner command in ICMP echo configuration mode (config-sla-monitor-echo) within IP SLA monitor configuration mode.
Examples
The following examples set the owner of IP SLAs ICMP echo operation 1 to 172.16.1.189 cwb.cisco.com User1 RTP 555-0100. Note that the Cisco IOS command used to begin configuration for an IP SLAs operation varies depending on the Cisco IOS release you are running (see Table 17).
IP SLA Configuration
This example shows the owner command being used in an IPv4 network in ICMP echo configuration mode within IP SLA configuration mode:
ip sla 1
icmp-echo 172.16.1.176
owner 172.16.1.189 cwb.cisco.com User1 RTP 555-0100
!
ip sla schedule 1 life forever start-time now
IP SLA Monitor Configuration
This example shows the owner command being used in an IPv4 network in ICMP echo configuration mode within IP SLA monitor configuration mode:
ip sla monitor 1
type echo protocol ipIcmpEcho 172.16.1.176
owner 172.16.1.189 cwb.cisco.com User1 RTP 555-0100
!
ip sla monitor schedule 1 life forever start-time now
Related Commands
parameters
To enter IP SLA template parameters configuration mode and begin configuring operation-specific parameters in an auto IP Service Level Agreements (SLAs) operation template, use the parameters command in the appropriate submode of IP SLA template configuration mode. To return the configuration for all operation parameters to default values, use the no form of this command.
parameters
no parameters
Syntax Description
This command has no arguments or keywords.
Command Default
All operation parameters are configured with default values.
Command Modes
IP SLA Template Configuration
ICMP echo configuration (config-tplt-icmp-ech)
ICMP jitter configuration (config-tplt-icmp-jtr)
TCP connect configuration (config-tplt-tcp-conn)
UDP echo configuration (config-tplt-udp-ech)
UDP jitter configuration (config-tplt-udp-jtr)
Command History
|
|
---|---|
15.1(1)T |
This command was introduced. |
Usage Guidelines
This command enters IP SLA template parameters configuration mode for configuring operation-specific parameters in an auto IP SLAs operation template.
You must configure the type of IP SLAs operation, such as User Datagram Protocol Internet Control Message Protocol (ICMP) jitter or Internet Control Message Protocol (ICMP) jitter, before you can configure any other parameters of the operation.
The commands available in IP SLA template parameters configuration mode differ depending on the operation being configured. Type ? in IP SLA template-parameters configuration mode to see the operation-specific parameters that can be configured.
Examples
The following example shows how to modify certain operation-specific parameters in an auto IP SLAs operation template for a UDP jitter operation:
Router(config)# ip sla auto template type ip udp-jitter 1
Router(config-tplt-udp-jtr)# parameters
Router(config-udp-jtr-params)# precision microseconds
Router(config-udp-jtr-params)# clock-tolerance ntp oneway percent 10
Router(config-udp-jtr-params)# operation-packet high
Router(config-udp-jtr-params)# end
Router#
Router# show ip sla auto template type ip udp-jitter
IP SLAs Auto Template: 1
Measure Type: udp-jitter (control enabled)
Description:
IP options:
Source IP: 0.0.0.0 Source Port: 0
VRF: TOS: 0x0
Operation Parameters:
Request Data Size: 32 Verify Data: false
Number of Packets: 10 Inter packet interval: 20
Timeout: 5000 Threshold: 5000
Granularity: usec Operation packet priority: high
NTP Sync Tolerance: 10 percent
Statistics Aggregation option:
Hours of statistics kept: 2
Statistics Distributions options:
Distributions characteristics: RTT
Distributions bucket size: 20
Max number of distributions buckets: 1
Reaction Configuration: None
Related Commands
|
|
ip sla auto template |
Begins configuration for an IP SLAs operation template and enters IP SLA template configuration mode. |
path-discover
To enable the label switched path (LSP) discovery option for a Cisco IOS IP Service Level Agreements (SLAs) LSP Health Monitor operation and enter auto IP SLA MPLS LSP discovery parameters configuration mode, use the path-discover command in auto IP SLA MPLS parameters configuration mode. To disable the LSP discovery option, use the no form of this command.
path-discover
no path-discover
Syntax Description
This command has no arguments or keywords.
Command Default
The LSP discovery option is disabled.
Command Modes
Auto IP SLA MPLS parameters configuration (config-auto-ip-sla-mpls-params)
Command History
|
|
---|---|
12.2(31)SB2 |
This command was introduced. |
12.2(33)SRB |
This command was integrated into Cisco IOS Release 12.2(33)SRB. |
Examples
The following example shows how to enable the LSP discovery option of IP SLAs LSP Health Monitor operation 1:
auto ip sla mpls-lsp-monitor 1
type echo ipsla-vrf-all
path-discover
Related Commands
|
|
---|---|
auto ip sla mpls-lsp-monitor |
Begins configuration for an IP SLAs LSP Health Monitor operation and enters auto IP SLA MPLS configuration mode. |
path-echo
To configure a Cisco IOS IP Service Level Agreements (SLAs) Internet Control Message Protocol (ICMP) path echo operation, use the path-echo command in IP SLA configuration mode.
path-echo {destination-ip-address | destination-hostname} [source-ip {ip-address | hostname}]
Syntax Description
Defaults
No IP SLAs operation type is configured for the operation being configured.
Command Modes
IP SLA configuration (config-ip-sla)
Command History
Usage Guidelines
You must configure the type of IP SLAs operation (such as User Datagram Protocol [UDP] jitter or Internet Control Message Protocol [ICMP] echo) before you can configure any of the other parameters of the operation. To change the operation type of an existing IP SLAs operation, you must first delete the IP SLAs operation (using the no ip sla global configuration command) and then reconfigure the operation with the new operation type.
Examples
In the following example, IP SLAs operation 10 is configured as an ICMP path echo operation using the IP/ICMP protocol and the destination IP address 172.16.1.175:
ip sla 10
path-echo 172.16.1.175
!
ip sla schedule 10 start-time now
Related Commands
|
|
---|---|
ip sla |
Begins configuration for an IP SLAs operation and enters IP SLA configuration mode. |
path-jitter
To configure a Cisco IOS IP Service Level Agreements (SLAs) Internet Control Message Protocol (ICMP) path jitter operation, use the path-jitter command in IP SLA configuration mode.
path-jitter {destination-ip-address | destination-hostname} [source-ip {ip-address | hostname}] [num-packets packet-number] [interval milliseconds] [targetOnly]
Syntax Description
Defaults
No IP SLAs operation type is configured for the operation number being configured.
Command Modes
IP SLA configuration (config-ip-sla)
Command History
Usage Guidelines
If the targetOnly keyword is used, the ICMP path jitter operation will send echoes to the destination only (the path from the source to the destination is not traced).
If the targetOnly keyword is not used, the IP SLAs ICMP path jitter operation will trace a "hop-by-hop" IP path from the source to the destination and then send a user-specified number of test packets to each hop along the traced path at user-specified time intervals.
You must configure the type of IP SLAs operation (such as User Datagram Protocol [UDP] jitter or Internet Control Message Protocol [ICMP] echo) before you can configure any of the other parameters of the operation. To change the operation type of an existing IP SLAs operation, you must first delete the IP SLAs operation (using the no ip sla global configuration command) and then reconfigure the operation with the new operation type.
Examples
The following example show how to enable the ICMP path jitter operation to trace the IP path to the destination 172.69.5.6 and send 50 test packets to each hop with an interval of 30 ms between each test packet:
ip sla 2
path-jitter 172.69.5.6 num-packets 50 interval 30
!
ip sla schedule 2 start-time now
Related Commands
|
|
---|---|
ip sla |
Begins configuration for an IP SLAs operation and enters IP SLA configuration mode. |
paths-of-statistics-kept
To set the number of paths for which statistics are maintained per hour for a Cisco IOS IP Service Level Agreements (SLAs) operation, use the paths-of-statistics-kept command in the appropriate submode of IP SLA configuration or IP SLA monitor configuration mode. To return to the default value, use the no form of this command.
paths-of-statistics-kept size
no paths-of-statistics-kept
Syntax Description
size |
Number of paths for which statistics are maintained per hour. The default is 5. |
Defaults
5 paths
Command Modes
IP SLA Configuration
ICMP path echo configuration (config-ip-sla-pathEcho)
IP SLA Monitor Configuration
ICMP path echo configuration (config-sla-monitor-pathEcho)
Note The configuration mode varies depending on the Cisco IOS release you are running and the operation type configured. See the "Usage Guidelines" section for more information.
Command History
Usage Guidelines
A path is the route the request packet of the operation traverses through the network to get to its destination. The packet may take a different path to reach the same destination for each IP SLAs operation.
When the number of paths reaches the size specified, no further path-based information is stored.
Note This command is supported by the IP SLAs Internet Control Message Protocol (ICMP) path echo operation only.
For the IP SLAs ICMP path echo operation, the amount of router memory required to maintain the distribution statistics table is based on multiplying all of the values set by the following four commands:
•distributions-of-statistics-kept
•hops-of-statistics-kept
•paths-of-statistics-kept
•hours-of-statistics-kept
The general equation used to calculate the memory requirement to maintain the distribution statistics table for an ICMP path echo operation is as follows:
Memory allocation = (160 bytes) * (distributions-of-statistics-kept size) * (hops-of-statistics-kept size) * (paths-of-statistics-kept size) * (hours-of-statistics-kept hours)
Note To avoid significant impact on router memory, careful consideration should be used when configuring the distributions-of-statistics-kept, hops-of-statistics-kept, paths-of-statistics-kept, and hours-of-statistics-kept commands.
IP SLAs Operation Configuration Dependence on Cisco IOS Release
The Cisco IOS command used to begin configuration for an IP SLAs operation varies depending on the Cisco IOS release you are running (see Table 18). You must configure the type of IP SLAs operation (such as User Datagram Protocol [UDP] jitter or Internet Control Message Protocol [ICMP] echo) before you can configure any of the other parameters of the operation.
The configuration mode for the paths-of-statistics-kept command varies depending on the Cisco IOS release you are running (see Table 18) and the operation type configured. For example, if you are running Cisco IOS Release 12.4 and the ICMP path echo operation type is configured, you would enter the paths-of-statistics-kept command in ICMP path echo configuration mode (config-sla-monitor-pathEcho) within IP SLA monitor configuration mode.
Examples
The following examples show how to maintain statistics for only three paths for IP SLAs ICMP path echo operation 2. Note that the Cisco IOS command used to begin configuration for an IP SLAs operation varies depending on the Cisco IOS release you are running (see Table 18).
IP SLA Configuration
ip sla 2
path-echo 172.16.1.177
paths-of-statistics-kept 3
!
ip sla schedule 2 life forever start-time now
IP SLA Monitor Configuration
ip sla monitor 2
type pathEcho protocol ipIcmpEcho 172.16.1.177
paths-of-statistics-kept 3
!
ip sla monitor schedule 2 life forever start-time now
Related Commands
precision
To set the level of precision at which the statistics for a Cisco IOS IP Service Level Agreements (SLAs) operation are measured, use the precision command in the UDP jitter submode of IP SLA configuration, IP SLA monitor configuration, or IP SLA template parameters configuration mode. To return to the default value, use the no form of this command.
precision {milliseconds | microseconds}
no precision
Syntax Description
Command Default
Measurements for the IP SLAs operation are displayed in milliseconds
Command Modes
IP SLA Configuration
UDP jitter configuration (config-ip-sla-jitter)
IP SLA Monitor Configuration
UDP jitter configuration (config-sla-monitor-jitter)
IP SLA Template Parameters Configuration
UDP jitter configuration (config-udp-jtr-params)
Command History
Usage Guidelines
This command changes value of the precision command from the default (milliseconds) to the specified value. If the milliseconds keyword is configured (default), the measurements for an IP SLAs operation will be displayed with the granularity of 1 ms. For example, a value of 22 equals 22 ms. If the microseconds keyword is configured, the measurements for an IP SLAs operation will be displayed with the granularity of 1 microsecond. For example, a value of 202 equals 202 microseconds.
This command is supported by the IP SLAs User Datagram Protocol (UDP) jitter operation only.
The precision microseconds command requires that both the source and IP SLAs Responder devices are running a version of Cisco IOS software that supports the precision microseconds command. See the "Command History" table for information about the supported Cisco IOS software releases.
Microsecond granularity for precision measurements is not supported on Cisco Catalyst 3000 and 2000 series switches that support IP SLAs. Do not configure the microseconds keyword with this command when you configure UDP jitter operations on devices running Cisco IOS Release 12.2SE and to which this limitation applies, such as Cisco Catalyst 3650 series switches. Use the Cisco Feature Navigator to find information about platform support for the Cisco IOS IP SLAs feature.
IP SLAs Operation Configuration Dependence on Cisco IOS Release
The Cisco IOS command used to begin configuration for an IP SLAs operation varies depending on the Cisco IOS release you are running (see Table 19). You must configure the type of IP SLAs operation, such as User Datagram Protocol (UDP) jitter or Internet Control Message Protocol (ICMP) jitter, before you can configure any of the other parameters of the operation.
The configuration mode for the precision command varies depending on the Cisco IOS release you are running (see Table 19) and the operation type configured.
If you are using auto IP SLAs in Cisco IOS IP SLAs Engine 3.0, you must enter the parameters command in IP SLA template configuration mode before you can use the precision command.
Examples
The following examples show how to enable microsecond precision, configure the Network Time Protocol (NTP) synchronization offset tolerance to 10 percent, and set the packet priority to high for an IP SLAs UDP jitter operation. Note that the Cisco IOS command used to begin configuration for an IP SLAs operation varies depending on the Cisco IOS release you are running (see Table 19).
IP SLA Configuration
ip sla 1
udp-jitter 192.168.202.169 9006
precision microseconds
clock-tolerance ntp oneway percent 10
probe-packet priority high
frequency 300
!
ip sla schedule 1 life forever start-time after 00:00:06
IP SLA Monitor Configuration
ip sla monitor 1
type jitter dest-ipaddr 192.168.202.169 dest-port 9006
precision microseconds
clock-tolerance ntp oneway percent 10
probe-packet priority high
frequency 300
!
ip sla monitor schedule 1 life forever start-time after 00:00:06
IP SLA Template Parameters Configuration
Router(config)# ip sla auto template type ip udp-jitter 1
Router(config-udp-jtr-tplt)# parameters
Router(config-udp-jtr-params)# precision microseconds
Router(config-udp-jtr-params)# clock-tolerance ntp oneway percent 10
Router(config-udp-jtr-params)# operation-packet high
Router(config-udp-jtr-params)# end
Router#
Router# show ip sla auto template type ip udp-jitter
IP SLAs Auto Template: 1
Measure Type: udp-jitter (control enabled)
Description:
IP options:
Source IP: 0.0.0.0 Source Port: 0
VRF: TOS: 0x0
Operation Parameters:
Request Data Size: 32 Verify Data: false
Number of Packets: 10 Inter packet interval: 20
Timeout: 5000 Threshold: 5000
Granularity: usec Operation packet priority: high
NTP Sync Tolerance: 10 percent
Statistics Aggregation option:
Hours of statistics kept: 2
Statistics Distributions options:
Distributions characteristics: RTT
Distributions bucket size: 20
Max number of distributions buckets: 1
Reaction Configuration: None
Related Commands
probe-interval
To configure the interval in an auto IP Service Level Agreements (SLAs) scheduler for staggering the start times of operations in Cisco IOS IP SLAs auto-measure groups that share the same schedule, use the probe-interval command in IP SLA auto-measure schedule configuration mode. To remove the interval configuration, use the no form of this command.
probe-interval milliseconds
no probe-interval
Syntax Description
milliseconds |
Length of time, in milliseconds (ms). Range is from 100 to 99000. Default is 1000. |
Command Default
There is a 1000 ms interval between the start time of one auto IP SLAs operation and the start time of the next auto IP SLAs operation being controlled by the same schedule.
Command Modes
IP SLAs auto-measure schedule configuration (config-am-schedule)
Command History
|
|
---|---|
15.1(1)T |
This command was introduced. |
Usage Guidelines
This command changes the default interval configuration (1000 ms) in an auto IP SLAs scheduler to the specified value.
An operation is created for each destination in an auto IP SLAs endpoint list specified for an IP SLAs auto-measure group.
Once the operations start, they continue operating based on the frequency specified by the frequency command.
Examples
The following example shows how to configure an auto IP SLAs scheduler that will cause an auto IP SLAs operation to actively collect data at 3:00 p.m. on April 5. The operation will age out after 12 hours of inactivity, which can be before it starts or after it has finished its life. When the operation ages out, all configuration information for the operation is removed from the running configuration in RAM:
Router(config)#ip sla auto schedule apr5
Router(config-am-schedule)#ageout 43200
Router(config-am-schedule)#frequency 70
Router(config-am-schedule)#life 43200
Router(config-am-schedule)#probe-interval 1500
Router(config-am-schedule)#start-time 15:00 apr 5
Router(config-am-schedule)#end
Router#
Router# show ip sla auto schedule apr5
Group sched-id: apr5
Probe Interval (ms) : 1500
Group operation frequency (sec): 70
Status of entry (SNMP RowStatus): Active
Next Scheduled Start Time: P15:00 apr 5
Life (sec): 43200
Entry Ageout (sec): 43200
Router#
Related Commands
probe-packet priority
Note Effective with Cisco IOS Release 12.4(6)T, the probe-packet priority command is replaced by the operation-packet-priority command. See the operation-packet priority command for more information.
To specify the packet priority of a Cisco IOS IP Service Level Agreements (SLAs) operation, use the probe-packet priority command in the appropriate submode of IP SLA configuration or IP SLA monitor configuration mode. To return to the default value, use the no form of this command.
probe-packet priority {normal | high}
no probe-packet priority
Syntax Description
probe-packet priority normal |
Sets the packet priority to normal. Packet priority is normal by default. |
probe-packet priority high |
Sets the packet priority to high. |
Command Default
Packet priority is normal.
Command Modes
IP SLA Configuration
UDP jitter configuration (config-ip-sla-jitter)
IP SLA Monitor Configuration
UDP jitter configuration (config-sla-monitor-jitter)
Note The configuration mode varies depending on the Cisco IOS release you are running and the operation type configured. See the "Usage Guidelines" section for more information.
Command History
Usage Guidelines
Increasing the packet priority of an IP SLAs operation can reduce the delay time for the packets in the queue.
Note This command is supported by the IP SLAs User Datagram Protocol (UDP) jitter operation only.
IP SLAs Operation Configuration Dependence on Cisco IOS Release
The Cisco IOS command used to begin configuration for an IP SLAs operation varies depending on the Cisco IOS release you are running (see Table 19). You must configure the type of IP SLAs operation (such as User Datagram Protocol [UDP] jitter or Internet Control Message Protocol [ICMP] echo) before you can configure any of the other parameters of the operation.
The configuration mode for the probe-packet priority command varies depending on the Cisco IOS release you are running (see Table 19) and the operation type configured. For example, if you are running Cisco IOS Release 12.4 and the UDP jitter operation type is configured, you would enter the probe-packet priority command in UDP jitter configuration mode (config-sla-monitor-jitter) within IP SLA monitor configuration mode.
Examples
The following examples show how to enable microsecond precision, configure the Network-Time Protocol (NTP) synchronization offset tolerance to 10 percent, and set the packet priority to high for IP SLAs UDP jitter operation 1. Note that the Cisco IOS command used to begin configuration for an IP SLAs operation varies depending on the Cisco IOS release you are running (see Table 19).
IP SLA Configuration
ip sla 1
udp-jitter 205.199.199.2 dest-port 9006
precision microseconds
clock-tolerance ntp oneway percent 10
probe-packet priority high
frequency 300
!
ip sla schedule 1 life forever start-time after 00:00:06
IP SLA Monitor Configuration
ip sla monitor 1
type jitter dest-ipaddr 205.199.199.2 dest-port 9006
precision microseconds
clock-tolerance ntp oneway percent 10
probe-packet priority high
frequency 300
!
ip sla monitor schedule 1 life forever start-time after 00:00:06