Classic CLI overview

Use the CLI to access, configure, and manage Nokia routers. CLI commands are entered at the command line prompt. Access to specific CLI commands is controlled by the permissions set by your system administrator. Entering a CLI command makes navigation possible from one command context (or level) to another.

When the user enters a CLI session, the user is in the operational root context. Navigate to another level by entering the name of successively lower contexts. For example, at the command prompt (#), enter configure or config. See Command completion for alternative syntax for the same command. The active context displays in the command prompt.

A:cses-E11# config
A:cses-E11>config#

In a CLI context, enter commands at that context level by entering the text. Press <Enter> to move to a lower context. The user can also include commands from lower context at one context level as long as the command and parameter syntax is correct.

The following example shows two methods to navigate to a service SDP ingress level.

Method 1:

A:cses-E11# configure service epipe 6 spoke-sdp 2:6 ingress 
*A:cses-E11>config>service>epipe>spoke-sdp>ingress#

Method 2:

A:cses-E11>config# service
A:cses-E11>config>service# epipe 6
*A:cses-E11>config>service>epipe# spoke-sdp 2:6
*A:cses-E11>config>service>epipe>spoke-sdp# ingress
*A:cses-E11>config>service>epipe>spoke-sdp>ingress#

The CLI returns an error message if the syntax is incorrect. For example, if the user enters rooter for the root command, it would result in an error.

*A:cses-E11>config# rooter
Error: Bad command.

The command parameter is the means by which a value is passed to the command processing program. The user must enter the value according to the syntax rules and, where applicable, the defined range. In the previous example, ‟6” and ‟2:6” are the parameter values that the user must enter to execute the command. The value ‟6” is the value for the Epipe service identifier parameter. For the value ‟2:6”, ‟2” is the value for the SDP identifier parameter and ‟6” is the value for the virtual circuit identifier.

The commands in the following table are available at the operational root level of the CLI hierarchy. For the command descriptions, see the respective command sections in this guide. For descriptions of the clear, monitor, show, and tools commands, see the 7450 ESS, 7750 SR, 7950 XRS, and VSR Clear, Monitor, Show, and Tools Command Reference Guide.

Global commands are commands that can be entered at any level in the CLI hierarchy. To display the list of all system global commands, enter help globals in the CLI.

The global commands are listed in the following table. For the command descriptions, see the respective command sections in this guide.

The CLI environment commands listed in the following table are found in the environment context of the operational root of the CLI tree. These commands control session preferences for a single CLI session. For more information on the commands, see the respective command sections in this guide.

The CLI supports both command abbreviation and command completion. If the keystrokes entered are enough to match a valid command, the CLI displays the remainder of the command syntax when Tab or Spacebar is pressed. When typing a command, Tab or Spacebar invokes auto-completion. If the keystrokes entered are sufficient to identity a specific command, auto-completion completes the command. If the letters are not sufficient to identify a specific command, pressing Tab or Spacebar displays commands matching the letters entered.

System commands are available in all CLI contexts.

In a command context, the CLI accepts command parameters in any order as long as the command is formatted in the proper command keyword and parameter syntax. Command completion works as long as enough recognizable characters of the command are entered.

The following output shows the command syntax for static-route-entry.

*A:cses-E11>config>router# static-route-entry ?
 - no static-route-entry <ip-prefix/prefix-length> [mcast] 
 - static-route-entry <ip-prefix/prefix-length> [mcast] 

<ip-prefix/prefix-*> : ipv4-prefix    - a.b.c.d (host bits must be 0)
                       ipv4-prefix-le - [0..32]  
                       ipv6-prefix    - x:x:x:x:x:x:x:x   (eight 16-bit pieces)
                                        x:x:x:x:x:x:d.d.d.d 
                                        x - [0..FFFF]H
                                        d - [0..255]D
                       ipv6-prefix-le - [0..128]  
<mcast>              : keyword - Indicates that static-route being configured 
                       is used for mcast table only

 [no] backup-tag      - Create/Configure or Delete/Deconfigure backup tag for
                        static-route-entry
 [no] black-hole      + Create/Configure or Delete/Deconfigure blackhole
                        nexthop for static-route-entry
 [no] community       - Create/Configure or Delete/Deconfigure community for
                        static-route-entry
 [no] indirect        + Create/Configure or Delete/Deconfigure indirect
                        next-hop for static-route-entry
 [no] next-hop        + Create/Configure or Delete/Deconfigure next-hop for
                        static-route-entry
 [no] tag             - Create/Configure or Delete/Deconfigure tag for
                        static-route-entry

Some SR OS CLI commands have multiple unnamed parameters. For example, the subrate csu-mode rate-step command has both a csu-mode parameter and a rate-step parameter that do not have leading keywords. SR OS uses a best-match algorithm to select which parts of the user input are intended to be used for each unnamed parameter. This best-match algorithm depends on the specific command.

In some cases, it is not possible for the algorithm to be 100% accurate, and SR OS may assign an unintended value to a parameter when two unnamed parameters have similar constraints and syntax. For example, the environment alias alias-name alias-command-name command may reverse the alias-name and alias-command-name parameters if the first parameter entered is more than 80 characters.

When entering a command, special keystrokes allow for editing of the command. The following table lists the command editing keystrokes.

CLI commands can be executed in any context by specifying the full path from the CLI root. To execute an out-of-context command, enter a forward slash (/) or backward slash (\) at the beginning of the command line. The commands are interpreted as absolute paths. Spaces between the slash and the first command return an error.

*A:cses-E11# configure router 
*A:cses-E11>config>router# interface system address 10.2.3.4 
*A:cses-E11>config>router# /admin save 
*A:cses-E11>config>router# \clear router interface 
*A:cses-E11>config>router# 

The ‟/” or ‟\” cannot be used as an absolute path at the beginning of the command string of the environment alias command. The command may change the current context depending on whether it is a leaf command. This is the same behavior the CLI performs when CLI commands are entered individually; for example:

*A:cses-E11# admin 
*A:cses-E11>admin# save

or

*A:cses-E11# admin save
*A:cses-E11# 

An absolute path command behaves in the same way as manually entering a series of command line instructions and parameters.

For example, beginning in an IES context service ID 4 (IES 4):

config>service>ies> /clear card 1

behaves in the way same as the following series of commands:

config>service>ies>exit all
clear card 1
configure service ies 4 (returns you to your starting point) 
config>service>ies

If the command takes you to a different context, the following occurs:

config>service>ies>/configure service vpls 5 create

becomes:

config>service>ies>exit all
configure service vpls 5 create
config>service>vpls>

The CLI maintains a history of the most recently entered commands. The history command shows the most recently entered CLI commands.

*A:cses-E11# history 
   1 environment terminal length 48 
   2 environment no create 
   3 show version 
   4 configure port 1/1/1 
   5 info 
   6 \configure router isis 
   7 \port 1/1/2
   8 con port 1/1/2
   9 \con port 1/1/2
  10 \configure router bgp 
  11 info 
  12 \configure system login-control 
  13 info 
  14 history 
  15 show version 
  16 history 
*A:cses-E11# !3

A:cses-E11# show version
TiMOS-B-0.0.I2016 both/i386 Nokia 7450 ESS Copyright (c) 2000-2016 Nokia
All rights reserved. All use subject to applicable license agreements.
Built on Sun Oct 12 20:01:13 PDT 2008 by builder in /rel0.0/I2016/panos/main
A:cses-E11# 

The SR OS CLI allows the use of a single numerical range as an argument in the command line. This range can be a set or a sequence of numbers, or a combination of both.

A set is a range of numerical values, from a minimum to a maximum, incremented by 1. For example:

configure service vpls [1..10] create customer 1

A sequence is a list of discrete integer elements, in any order. For example:

configure service vpls [1,2,3] no shutdown

A sequence can contain sets as well as integer elements. For example:

configure service vpls [4..6,7,8..10] no shutdown

For example, it is possible to shut down ports 1 through 10 on an XMA/MDA 1 in chassis slot 1. A port can be denoted by ‟slot/mda/port”, where slot is the slot number, mda is the XMA/MDA number and port is the port number. To shut down ports 1 through 10 on an XMA/MDA 1 in slot 1, the command is entered as follows:

configure port 1/1/[1..10] shutdown

Ctrl-C can be used to abort the execution of a range command.

CLI commands can contain ranges of hexadecimal values. This allows ranges to be used when working with data normally expressed in hexadecimal instead of decimal, such as IPv6 or MAC addresses. For example:

#config>service>vpls>sap$ static-mac aa:bb:[0x19..0x21]:dd:ee:ff create
#config>service>vpls>sap$ info
----------------------------------------------
                static-mac aa:bb:19:dd:ee:ff create
                static-mac aa:bb:1a:dd:ee:ff create
                static-mac aa:bb:1b:dd:ee:ff create
                static-mac aa:bb:1c:dd:ee:ff create
                static-mac aa:bb:1d:dd:ee:ff create
                static-mac aa:bb:1e:dd:ee:ff create
                static-mac aa:bb:1f:dd:ee:ff create
                static-mac aa:bb:20:dd:ee:ff create
                static-mac aa:bb:21:dd:ee:ff create
----------------------------------------------

A range can also be a reference to a previous range in the same command. This reference takes the form [$x], where x is an integer between 0 and 5. For example:

configure service vprn [11..20] router-id 10.20.[$0].1

This gives vprn 11 the router-id "10.20.11.1", vprn 12 the router-id "10.20.12.1", and so on.

Specifying a range in the CLI does have limitations. These limitations are summarized in the following table.

The | match output modifier searches for a character string or pattern. When using the | match output modifier, the variables and attributes must be spelled correctly. The attributes follow the output modifier and must come before the expression or pattern. The following are examples of how to use the | match output modifier to complete different tasks:

• Task: Capture all the lines that include ‟echo” and redirect the output to a file on the compact flash:

  admin display-config | match ‟echo” > cf1:\test\echo_list.txt

• Task: Display all the lines that do not include ‟echo”:

  admin display-config | match invert-match ‟echo”

• Task: Display the first match of ‟vpls” in the configuration file:

  admin display-config | match max-count 1 ‟vpls”

• Task: Display everything in the configuration after finding the first instance of ‟interface”:

  admin display-config | match post-lines 999999 interface

• Task: Display a count of the total number of lines of output instead of displaying the output itself:

  admin display-config | match interface | count

Command syntax:

match pattern context {parents | children | all} [ignore-case] [max-count lines-count] [expression]

match pattern [ignore-case] [invert-match] [pre-lines pre-lines] [post-lines lines-count] [max-count lines-count] [expression]

where:

pattern         string or regular expression
context         keyword:  display context associated with the matching line
parents         keyword:  display parent context information
children        keyword:  display child context information
all             keyword:  display both parent and child context information
ignore-case     keyword
max-count       keyword:  display only a specific number of instances of matching 
                          lines
lines-count     1 — 2147483647
expression      keyword:  pattern is interpreted as a regular expression
invert-match    keyword
pre-lines       keyword:  display some lines prior to the matching line
pre-lines       0 — 100
post-lines      keyword:  display some lines after the matching line
lines-count     1 — 2147483647

For example:

A:Dut-C# show log log-id 98 | match ignore-case "sdp bind"
"Status of SDP Bind 101:1002 in service 1001 (customer 1) changed to admin=up oper=u
p flags="
"Processing of a SDP state change event is finished and the status of all affected S
DP Bindings on SDP 101 has been updated."


A:Dut-C# show log log-id 98 | match max-count 1 "service 1001"
"Status of service 1001 (customer 1) changed to administrative state: up, 
operational state: up"

A:Dut-C# admin display-config | match post-lines 5 max-
count 2 expression "OSPF.*Config"
echo "OSPFv2 Configuration"
#--------------------------------------------------
        ospf
            timers
                spf-wait 1000 1000 1000
            exit
echo "OSPFv2 (Inst: 1) Configuration"
#--------------------------------------------------
        ospf 1
            asbr
            router-id 10.0.0.1
            export "testall"
*A:Dut# admin display-config | match debug_mirror
            profile "debug_mirror"


*A:Dut# admin display-config | match context parent debug_mirror
#--------------------------------------------------
    system
        security
            profile "debug_mirror"


*A:Dut# admin display-config | match context all debug_mirror
#--------------------------------------------------
    system
        security
            profile "debug_mirror"
                default-action deny-all
                entry 10
                exit


*A:Dut# show log event-control | match ignore-case pre-lines 10 SyncStatus
L  2016 tmnxLogOnlyEventThrottled        MA  gen          0           0
MCPATH:
   2005 tmnxMcPathSrcGrpBlackHole        MI  thr          0           0
   2006 tmnxMcPathSrcGrpBlackHoleCleared MI  thr          0           0
   2007 tmnxMcPathAvailBwLimitExceeded   MI  thr          0           0
   2008 tmnxMcPathAvailBwLimitCleared    MI  thr          0           0
MC_REDUNDANCY:
   2001 tmnxMcRedundancyPeerStateChanged WA  gen          0           0
   2002 tmnxMcRedundancyMismatchDetected WA  gen          0           0
   2003 tmnxMcRedundancyMismatchResolved WA  gen          0           0
   2004 tmnxMcPeerSyncStatusChanged      WA  gen          0           0

The following table describes regular expression symbols and their interpretation (similar to what is used for route policy regexp matching). Character class expressions describes character class expressions.

Character class expressions must be enclosed within brackets. The expression [[:digit:]] is treated as a regular expression containing the character class ‟digit”, while [:digit:] is treated as a regular expression matching ‟:”, ‟d”, ‟i”, ‟g”, or ‟t”.

The | count output modifier displays a count of the number of lines that would have otherwise been displayed. The | count output modifier is particularly useful when used in conjunction with the | match output modifier in order to count the number of output lines that match a specified pattern.

The following example shows usage of the | count output modifier.

*A:dut-c# show service service-using vprn

===============================================================================
Services [vprn]
===============================================================================
ServiceId    Type      Adm  Opr  CustomerId Service Name
-------------------------------------------------------------------------------
1            VPRN      Down Down 1
44           VPRN      Up   Up   1
100          VPRN      Down Down 1
102          VPRN      Up   Up   1
235          VPRN      Down Down 1
1000         VPRN      Down Down 1000
-------------------------------------------------------------------------------
Matching Services : 6
-------------------------------------------------------------------------------
===============================================================================
*A:dut-c# show service service-using vprn | match Down | count
Count: 4 lines
*A:dut-c#

The | reverse-dns output modifier performs a reverse DNS lookup on any IPv4 or IPv6 address in the input. The result of the lookup is inserted as the next line in the output on each line where an IP address is identified. If no match is found, no additional output is printed. If the output line is more than 80 characters, the line is truncated.

The following example shows usage of the | reverse-dns output modifier.

A:node-2# ping 10.184.216.34 | reverse-dns
PING 10.184.216.34 56 data bytes
(10.184.216.34) www.example.com
64 bytes from 10.184.216.34: icmp_seq=1 ttl=61 time=82.4ms.
64 bytes from 10.184.216.34: icmp_seq=2 ttl=61 time=82.5ms.
64 bytes from 10.184.216.34: icmp_seq=3 ttl=61 time=82.4ms.
64 bytes from 10.184.216.34: icmp_seq=4 ttl=61 time=82.3ms.
64 bytes from 10.184.216.34: icmp_seq=5 ttl=61 time=82.2ms.
---- 10.184.216.34 PING Statistics ----
(10.184.216.34) www.example.com
5 packets transmitted, 5 packets received, 0.00% packet loss
round-trip min = 82.2ms, avg = 82.4ms, max = 82.5ms, stddev = 0.122ms

SR OS supports output redirection (>), which allows the operator to store the output of a CLI command as a local or remote file.

For example:

‛ping <customer_ip> > cf3cf1:/ping/result.txt’
‛ping <customer_ip> > ftp://ron@ftp.nokia.com/ping/result.txt’

In some cases, only part of the output might be applicable. The | match and output redirection commands can be combined. For example, the following command matches lines with the expression ‟time.[[:digit:]]+” and redirects the output to the file <file format>cf3:/ping/time.txt</ff>.

ping 10.0.0.1 | match expression ‟time.[[:digit:]]+” > cf3:/ping/time.txt

The following list describes detailed behavior of the rollback feature, including the applicable CLI commands.

• The user can create a rollback checkpoint and later revert to this checkpoint with minimal impact to services.

  admin>rollback# save [comment comment-string]

  comment-string: a comment associated with the checkpoint, maximum 255 characters

• Rollback checkpoints include all current operationally active configurations:

  • changes from direct CLI commands in the configuration branch

  • SNMP sets

• Rollback checkpoints do not include BOF configurations. The BOF file and BOF configuration are not part of a rollback save or rollback. A rollback does not change the BOF configuration. The BOF contains basic information for the node and does not change frequently; changes are mostly made during initial commissioning of the node.

• A rollback save feature can be automatically executed (for example, scheduled monthly) using the CRON facility of SR OS.

• The latest rollback checkpoint file uses the suffix ‟.rb”. The next latest rollback checkpoint file has the suffix ‟.rb.1”, the next oldest has the suffix ‟rb.2”, and so on.

  file-url.rb <--- latest rollback file

  file-url.rb.1

  …

  file-url.rb.9 <--- oldest rollback file

• When a rollback save is executed, the system shifts the file suffix of all the previous checkpoints by 1 (new ID = old ID + 1). If there are already as many checkpoint files as the maximum number supported, the last checkpoint file is deleted.

• The maximum number of rollback checkpoints is configurable and defaults to 10, which includes the latest file and files 1 through 9.

• The locations and names of the rollback checkpoint files are configurable to be local (on the compact flash) or remote. The file-url must contain a path/directory and filename with no suffix. The .rb suffix for rollback checkpoint files is automatically appended to the rollback checkpoint files.

  config>system>rollback# rollback-location file-url

• There is no default rollback location. If a rollback location is not specified, or it is cleared using no rollback-location, and a rollback save is attempted, the rollback save fails and returns an error message.

• The entire set of rollback checkpoint files can be copied from the active CPM CF to the standby CPM CF. This synchronization is done using the following command:

  admin>redundancy# rollback-sync

• The operator can enable an automatic synchronization of rollback checkpoint files between the active CPM and standby CPM. When this automatic synchronization is enabled, a rollback save causes the new checkpoint file to be saved to both the active and standby CPMs. The suffixes of the old checkpoint files on both active and standby CPMs are incremented.

  Automatic synchronization only causes the new checkpoint file to be copied to both CFs. The older checkpoint files are not automatically copied from active to standby but can be copied manually with admin redundancy rollback-sync.

• The config>redundancy>synchronize {boot-env | config} and admin>redundancy>synchronize {boot-env | config} commands do not apply to rollback checkpoint files. These commands do not manually or automatically synchronize rollback checkpoint files. The dedicated rollback-sync commands must be used to synchronize rollback checkpoint files.

• Rollback files can be deleted using a dedicated rollback checkpoint deletion command.

  admin>rollback# delete {latest-rb | checkpoint-id}

  • Deleting a rollback checkpoint causes the suffixes to be adjusted (decremented) for all checkpoints older that the one that was deleted in order to close the ‟hole” in the list of checkpoint files and create room to create another checkpoint.

  • If config>redundancy>rollback-sync is enabled, a rollback delete also deletes the equivalent checkpoint on the standby CF and shuffles the suffixes on the standby CF.

  • If an operator manually deletes a rollback checkpoint file using file delete, the suffixes of the checkpoint files are not shuffled, nor is the equivalent checkpoint file deleted from the standby CF. This manual deletion creates a ‟hole” in the checkpoint file list until enough new checkpoints have been created to roll the ‟hole” off the end of the list.

• The following figure shows how a configuration is rolled back to a previous configuration (a saved rollback checkpoint). The previous configuration is loaded and takes operational effect.

  admin>rollback# revert [latest-rb | checkpoint-id]

  

• A rollback revert does not affect the currently stored rollback checkpoint files; files are not deleted or renumbered. This means that if an operator issues the command rollback revert 3 and then issues the rollback save command, the resulting rollback checkpoint files file-url.rb and file-url.rb.4 contain the same rollback state/configuration.

• The boot-good-exec or boot-bad-exec command is not automatically executed after a rollback.

• Impacts to the running services are minimized during a rollback.

  • There is no impact in areas of the configuration that did not change.

  • Configuration parameters that are changed or items that the changed configuration has dependencies on are removed (revert to default) and the previous values are restored, which may briefly impact services. The following are examples of service impact.

    • If the currently active configuration contains configure port 5/1/1 dwdm tdcm dispersion -1000 and the rollback checkpoint contains configure port 5/1/1 dwdm tdcm dispersion -1010, the operational dispersion transitions from -1000 to 0 and then back to -1010 for port 5/1/1, which causes a traffic interruption.
    • If changing neighbor 1 of an MC-APS port, the port must be configured as no neighbor and then configured as neighbor 2. Moving through the no neighbor intermediate state requires the working and protection circuits to be torn down and rebuilt. This impacts the 7450 ESS and 7750 SR.

• A rollback undoes any SNMP sets or direct CLI configuration commands that occur after the creation of the last checkpoint.

• When a node is processing a rollback revert, both CLI commands from other users and SNMP commands continue to be processed. The only commands that are blocked during a rollback revert are other rollback commands, including revert, save, and compare. Only one rollback command can be executing at a time on a node.

• Commands are available to view and compare the various rollback checkpoints to current operating and candidate configurations.

• Rollback checkpoint files are not guaranteed to be in any particular format. They are not interchangeable with normal configuration files or executable scripts. A normal configuration file from an admin save cannot be renamed as a rollback checkpoint and then referenced for a rollback revert operation. Only rollback checkpoint files generated with rollback save can be used for rollback revert operations.

• If a hardware change is made after a rollback save, then:

  • a rollback can be executed as long as the hardware change was an addition of hardware to the node (for example, a new card or IOM was installed into a previously empty slot)

  • a rollback is not guaranteed to work if hardware was removed or changed (for example, an XCM/IOM was removed, or an XMA/MDA was swapped for a different XMA/MDA type)

• Rollback across a change to the following parameters is not supported:

  • chassis-mode
  • configure isa application-assurance-group minimum-isa-generation

• Rollback is supported even after an admin reboot is performed or the primary configuration in the BOF is changed and an admin reboot is performed. The admin reboot command does not ‟break the chain” for rollback.

• Lawful intercept configuration under the config>li branch is not affected by a rollback or rescue. LI configuration is not saved in the rollback checkpoint or rescue file, and a rollback revert does not affect any configuration under the config>li branch.

• Any configuration or state change performed under the debug branch of the CLI is not saved in the rollback checkpoint file or impacted by a rollback.

• Rollbacks to a checkpoint created in a more recent release are not supported (for example, a node running in 9.0r5 cannot roll back to a checkpoint created in 9.0r7).

• The following list captures some side effects and specific behaviors of a rollback revert. Some of these side effects are not related purely to configuration, that is, in the CLI configuration branch, and may have interactions with tools commands, RADIUS, and so on.

  • SAA jobs that are running when a rollback revert is initiated, and need configuration changes due to the rollback, are stopped. If the SAA job is a continuous type, then it is restarted as part of the rollback revert after the configuration changes have been applied (just as if the operator had typed no shutdown for the continuous SAA job). Non-continuous SAA jobs that were modified by the rollback would need to be manually restarted if they need to be run again.

  • If max-nbr-mac-addr is reduced as part of the revert and the number of MAC addresses in the forwarding database is greater than the max-nbr-mac-addr, the rollback is aborted before any actions are taken and an informative error message is provided. The operator must take actions to remove the MAC addresses if they wish to proceed with the rollback.

  • If active subscribers or subscriber hosts or DHCP lease states are present, some associated configuration changes may be blocked, just as those same changes would be blocked if an operator tried to make them using CLI.

  • When trying to delete an SLA profile being used by active subscriber hosts, or trying to change a NAT policy in a subscriber profile, if certain configuration changes associated with the hosts or lease states are required as part of the rollback but those changes are blocked, then for each blocked configuration item a warning is printed, that particular configuration item is not changed, and the rollback continues. This is supported on the 7450 ESS and 7750 SR.

  • After a multi-chassis peer shutdown or if configuration changes have been made that affect the contents of the distributed database (for example, synchronization tag creation or deletion), further configuration changes related to that peer may be temporarily refused. The duration of the temporary configuration freeze depends on the size of the distributed database. A rollback attempting to make those refused configuration changes fails and an error message is provided to the CLI user.

  • If a force-switchover command (for example, tools perform service id 1 endpoint "x" force-switchover spoke-sdp-fec 1) has been applied to a spoke-sdp-fec of a dynamic multi-segment pseudowire, and a rollback revert needs to change the admin state of the spoke-sdp-fec (for example, to modify spoke-sdp-fec parameters that may be dependent on the admin state), the rollback revert automatically removes the force-switchover and the node reverts to whatever is the best spoke SDP in the redundant set.

  • Rollback impacts the configuration state of the router, and as with normal operator CLI or SNMP configuration changes, additional actions or steps may need to occur before certain configuration changes take operational effect.

    Configuration changes that require a manual shutdown and then no shutdown in order to take operational effect also need this manual shutdown/no shutdown in order to take operational effect after a rollback if the rollback changes those configuration items. Some examples include the following.

    • Changes to autonomous system or confederation values require a BGP shutdown/no shutdown command.

    • Changes to VPRN max-routes require a shutdown/no shutdown command on the VPRN service.

    • Changes to OSPF or IS-IS export-limit require a shutdown/no shutdown command on OSPF or IS-IS.

    • For configuration changes to an MSAP policy that normally require a tools perform subscriber-mgmt eval-msap command to take operational effect on subscribers that are already active, if a rollback changes the MSAP policy configuration, the operator must run the eval-msap tools command to have the changes applied to the active subscribers.

  • Any uncommitted changes (for example, the begin command was entered and some changes were made, but the commit command was never entered) in the following areas are lost or cleared when a rollback revert is initiated:

    • config>app-assure>group policy

    • config>router>policy-options

    • config>system>sync-if-timing

• Some card and mda commands require a reboot, removal, or rebuild of an entire card or XMA/MDA. When these commands need to be executed as part of a rollback, the impacted cards and MDAs are listed in a warning and the operator is prompted with a single y/n prompt to proceed. This prompt will not occur for a rollback initiated via SNMP or if the operator uses the now keyword with the rollback revert command. Some examples of card and mda commands that may cause a prompt are:

  • config>card>card-type

  • config>card>mda

  • config>card>mda>mda-type

• Although the use of the Ctrl-C key combination is not recommended during a rollback revert, it is supported via CLI or SNMP. Interrupting a rollback revert may leave the router in a state that is not necessarily between the old active configuration and the rollback checkpoint, as the rollback processing may have been in the middle of tearing things down or rebuilding configurations. A strong warning is issued in this case to indicate that the operator must examine the configuration and potentially issue another rollback revert to return to a known and coherent configuration.

• An HA CPM switchover during a rollback revert causes the rollback operation to abort. The newly active CPM has an indeterminate configuration. When an HA switchover occurs during a rollback or within a few seconds of a rollback completing, the operator is advised to repeat the rollback revert operation to the same checkpoint.

• A rollback revert operation does not check authorization of each command that is applied during the revert. Permission to execute the revert operation, that is, authorization to execute the admin rollback revert command, should only be given to users who are allowed to initiate a rollback revert. It is generally recommended that only system administrators be allowed access to the file system where the rollback files are stored so that they cannot be manually edited.

The SR OS has SNMP support for rollback status and control. See the TIMETRA-SYSTEM-MIB for details (for example, items such as tmnxSysRollbackStarted).

When the router is doing a rollback revert, SNMP managers see a tmnxSysRollbackStarted trap, then a rapid set of ‟config change” traps, and then finally, the tmnxSysRollbackStatusChange trap.

During the period when a router is processing a rollback revert, both CLI commands from other users and SNMP commands continue to be processed.

A special rescue configuration checkpoint can be created that an operator can revert to at any time. The rescue configuration has its own keyword (rescue) and does not use the same rolling suffix indices as the normal rollback checkpoints. This allows the operator to easily return to the rescue configuration state without having to consider a checkpoint index, and ensures that the rescue checkpoint is always available and does not roll off the bottom of the list of checkpoints.

The operator should define a basic rescue configuration that is known to work and give appropriate management access to the node.

The location and filename of the rescue file are configurable. The SR OS appends the .rc suffix to the specified rescue filename.

The following points offer some operational guidance on the use of rollback.

• The admin save and admin rollback save commands should be performed periodically.

• The admin save command can be used to back up a complete configuration file that can be used during router reboot, with the following considerations:

  • used with a reboot as a last resort

  • performed after any major hardware changes or major service changes

  • performed after any software upgrade

• The admin rollback save command can be used to create a rollback checkpoint as follows:

  • to be used for intermediate checkpoints that can be recovered with minimal impact to services

  • to be performed each time a moderate number of configuration changes have been made

  • to be performed after any hardware changes

  • to be performed after any software upgrade

  • to be scheduled with CRON (for example, once every one or two weeks)

• A new admin rollback save rescue must be created when hardware is changed.

• Rollback checkpoint files are not editable, or compatible or interchangeable with configuration files generated with admin save.

• The repeated use of the admin rollback save, admin rollback delete, and admin rollback revert commands over the course of weeks or months is not recommended without also executing an occasional admin save. In a serious situation, use one of the saved configurations as the primary configuration for an admin reboot.

• For a software upgrade, it is recommended that a rollback checkpoint be created using admin rollback save in addition to saving the configuration with admin save, after an upgrade has been performed and the system is operating as expected. This ensures that a good checkpoint that is fully compatible with the new release is available at a point shortly after the upgrade.

• An operator can create a set of rollback checkpoints to support busy or quiet days or weekends or weekdays and use CRON to shift between them.

• It is beneficial to create a rollback checkpoint before a rollback revert is initiated, especially if significant configuration changes have been applied since the last checkpoint was created. If the rollback is especially significant, that is, there are a lot of major changes, it is also a good practice to perform an admin save in case a full reboot is required to recover from an issue.

• A rollback failure may occur in some limited cases where the node needs a long time to complete one of the resulting configuration changes. If a rollback (for example, rollback revert 5) fails during execution, it should be attempted again. The second attempt typically completes the remaining configuration changes required to fully revert to the desired checkpoint.

• When a new backup CPM is commissioned, the user executes the admin redundancy rollback-sync command to copy the entire set of rollback files from the active CPM CF to the new standby CPM CF. If the operator wants the system to automatically copy new rollback checkpoints to both CFs whenever a new checkpoint is created, the configure redundancy rollback-sync command should be enabled.

• An HA CPM switchover during a rollback revert causes the rollback operation to abort. The newly active CPM has an indeterminate configuration. A log event is created in this case to warn the operator. When an HA switchover occurs during a rollback or within a few seconds of a rollback completing, the operator is advised to repeat the rollback revert operation to the same checkpoint.

• A rollback checkpoint stores the rollback location and the local-max-checkpoint and remote-max-checkpoint values, and it is possible that a rollback revert operation may change those values. If an operator changes the local-max-checkpoint or remote-max-checkpoint values, it is recommended that all the existing checkpoints be deleted to prevent a subsequent rollback revert from changing the maximum values to any of the previous values.

• If a warning prompt (y/n) is displayed when a rollback revert is initiated, it is highly recommended that the operator respond no to the warning prompt the first time, save a rollback checkpoint before attempting this rollback revert, execute the revert again, and respond yes. If the rollback encounters problems, a revert to the saved checkpoint can be used to return to the initial configuration state.

In order to edit the candidate configuration, the operator must first enter the candidate edit mode (edit-cfg) with the candidate>edit command. The operator can enter and quit the configuration mode as many times as they wish before finally committing the candidate configuration.

In edit-cfg mode, the operator builds a set of candidate configuration changes using the same CLI tree as the standard line-by-line, non-transactional configuration. Tab completion and keyword syntax checking is available.

Just as there is a single operational active configuration that can be modified simultaneously by multiple users in the SR OS, there is also a single global candidate configuration instance. All users make changes in the same global candidate configuration, and a commit operation by any user commits the changes made by all users.

Users can exclusively create a candidate configuration by blocking other users and sessions of the same user from entering edit-cfg mode by specifying the exclusive parameter. The config>system>management-interface>cli>classic-cli>allow-immediate command can be used to enforce the use of candidate configuration, instead of allowing immediate line-by-line configuration changes.

If a commit operation is successful, all of the candidate changes take operational effect and the candidate is cleared. If there is an error in the processing of the commit, or a commit confirmed is not confirmed and an auto-revert occurs, the router returns to a configuration state with none of the candidate changes applied. The operator can then continue editing the candidate and try a commit later.

All commands in the candidate configuration must be in the correct order for a commit to be successful. Configuration that depends on other candidate objects must be placed after those objects in the candidate. A set of candidate editing commands (copy, insert, and so on) are available to correct and reorder the candidate configuration.

The edit-cfg mode is primarily intended for building a candidate configuration while navigating the configure branch of the CLI. Many CLI commands in branches other than configure are supported while in edit-cfg mode, but access to some CLI branches and commands are blocked, including:

• exec command

• enable-admin command

• enable-dynamic-services-config command

• admin branch

• bof branch

• debug branch

• tools branch

The candidate configuration can be saved to a file and subsequently loaded into a candidate configuration. A saved candidate is similar to, but not the same as, an SR OS configuration file generated with an admin save command. The saved candidate cannot be used in general as a configuration file and may not exec without failures.

There is no SNMP access to the candidate configuration and no SNMP management of candidates, although any configuration changes made using transactional configuration are reported via the standard SR OS SNMP change traps and basic candidate status information is available via SNMP.

A commit may fail for a number of reasons, including:

• misordering

  The candidate configuration has changes that are not in the correct order, that is, an object is referred to before it is actually created

• invalid options and combinations

  Although many syntax errors are eliminated during the candidate editing process, the candidate configuration may contain combinations of configurations and options that are not valid and are rejected when the SR OS attempts to have them take operational effect

• out of resources

  The application of the candidate may exhaust system resources, such as queue resources

Error messages are provided for commit failures to help the operator take the necessary actions to correct the candidate.

Standard line-by-line, non-transactional CLI and SNMP commands are not blocked during the creation or editing of a candidate or the processing of a commit. These commands take immediate operational effect when Enter is pressed.

By default, the SR OS automatically creates a new rollback checkpoint after a commit operation. The rollback checkpoint includes the new configuration changes made by the commit. An optional no-checkpoint keyword can be used to prevent the auto-creation of a rollback checkpoint after a commit. If the commit fails, no new rollback checkpoint is created.

When the commit confirmed option is used, a rollback checkpoint is created after the processing of the commit and exists whether the commit is automatically reverted or not.

Transactional configuration relies on the rollback mechanism to operate. Any commands and configurations that are not supported in a rollback revert are also not supported in edit-cfg mode; for example, changes to chassis-mode.

Authorization works transparently in edit-cfg mode and no unique or new local profile or TACACS+ permissions rules are required other than allowing access to the candidate branch. For example, if an operator has permission to access the configure filter context, they automatically also have access to the configure filter context when in edit-cfg mode.

If the operator’s profile allows access to the candidate load and save commands, the operations load and save only those items that the user is authorized to access.

The candidate view only displays the items that the user is authorized to access.

The candidate editing commands (such as adding lines, removing lines, and delete operations) only allow operations on items that the user is authorized to access.

The candidate commit and discard commands, along with admin rollback revert, operate on the entire candidate and impact all items; authorization does not apply.