[FRR] set the enable as default value for zebra_nexthop by lipxu · Pull Request #25117 · sonic-net/sonic-buildimage

lipxu · 2026-01-18T11:30:53Z

Why I did it

Set the default value to enabled

Work item tracking

Microsoft ADO (number only):
36367940

How I did it

Set the default value to enabled

How to verify it

Local

Which release branch to backport (provide reason below if selected)

Tested branch (Please provide the tested image version)

Description for the changelog

Link to config_db schema for YANG module changes

A picture of a cute animal (not mandatory but encouraged)

Signed-off-by: Liping Xu <[email protected]>

mssonicbld · 2026-01-18T11:31:00Z

/azp run Azure.sonic-buildimage

azure-pipelines · 2026-01-18T11:31:14Z

Azure Pipelines successfully started running 1 pipeline(s).

Signed-off-by: Liping Xu <[email protected]>

mssonicbld · 2026-01-18T23:38:25Z

/azp run Azure.sonic-buildimage

azure-pipelines · 2026-01-18T23:38:35Z

Azure Pipelines successfully started running 1 pipeline(s).

Copilot

Pull request overview

This PR changes the default value of the zebra_nexthop configuration from disabled to enabled in the YANG model and updates the virtual switch FRR zebra configuration accordingly.

Changes:

Changed the default value for zebra_nexthop from disabled to enabled in the YANG model
Updated the VS platform zebra.conf to enable zebra nexthop kernel by default

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated no comments.

File	Description
src/sonic-yang-models/yang-models/sonic-device_metadata.yang	Changed default value for zebra_nexthop leaf from "disabled" to "enabled"
platform/vs/docker-sonic-vs/frr/zebra.conf	Changed zebra nexthop kernel configuration from disabled to enabled

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

deepak-singhal0408 · 2026-01-20T05:36:37Z

@lipxu , We probably need to modify the logic here too..
https://github.com/sonic-net/sonic-buildimage/blob/master/dockers/docker-fpm-frr/frr/zebra/zebra.conf.j2#L10

{% if (('localhost' in DEVICE_METADATA) and ('zebra_nexthop' in DEVICE_METADATA['localhost']) and
       (DEVICE_METADATA['localhost']['zebra_nexthop'] == 'enabled')) %}
zebra nexthop kernel enable
{% else %}
! Force disable next hop group support
no zebra nexthop kernel enable
{% endif %}

Signed-off-by: Liping Xu <[email protected]>

mssonicbld · 2026-01-20T05:41:36Z

/azp run Azure.sonic-buildimage

azure-pipelines · 2026-01-20T05:41:47Z

Azure Pipelines failed to run 1 pipeline(s).

lipxu · 2026-01-20T11:18:33Z

/azpw run Azure.sonic-buildimage

mssonicbld · 2026-01-20T11:18:36Z

/AzurePipelines run Azure.sonic-buildimage

azure-pipelines · 2026-01-20T11:18:49Z

Azure Pipelines successfully started running 1 pipeline(s).

eddieruan-alibaba · 2026-01-20T12:59:57Z

Can you explain the motivation why you change the default behavior?

deepak-singhal0408 · 2026-01-20T17:16:18Z

Can you explain the motivation why you change the default behavior?

@eddieruan-alibaba, This is to ensure by default we keep nexthop group support enabled in kernel. Disabling this by default has side effects(like memory explosion etc)..
for MSFT issue seen on T1s, we can disable this feature internally for specific platforms as required.. But by default its better we keep this enabled, so the benefits of nethop group sharing in kernel is available..

deepak-singhal0408

LGTM

Signed-off-by: Liping Xu <[email protected]>

) What is the motivation for this PR? The PR sonic-net/sonic-buildimage#25117 would update the default value of zebra_nexthop in FRR from disable to enable. Enable is the default configuration in FRR, it would not show in ouput of "show running" in FRR, so it would cause the test case test_frr_config_check failed once the previous PR merged. How did you do it? Skip the FRR hidden config check in test_frr_config_check. How did you verify/test it? Run the case in elastic with fix https://elastictest.org/scheduler/testplan/6978a655c5f3d9049fbb9138 without fix https://elastictest.org/scheduler/testplan/6978a6671910fa9097bc5e58 Signed-off-by: Raghavendran Ramanathan <[email protected]>

) What is the motivation for this PR? The PR sonic-net/sonic-buildimage#25117 would update the default value of zebra_nexthop in FRR from disable to enable. Enable is the default configuration in FRR, it would not show in ouput of "show running" in FRR, so it would cause the test case test_frr_config_check failed once the previous PR merged. How did you do it? Skip the FRR hidden config check in test_frr_config_check. How did you verify/test it? Run the case in elastic with fix https://elastictest.org/scheduler/testplan/6978a655c5f3d9049fbb9138 without fix https://elastictest.org/scheduler/testplan/6978a6671910fa9097bc5e58 Signed-off-by: Zhuohui Tan <[email protected]>

deepak-singhal0408 · 2026-02-20T20:03:24Z

recommend to not merge this in 202511, as this will alter the existing default behavior.. Newer branches going forward can have this default behavior..

let me know if any concerns.. Thanks,

looks like having this enabled by default works better for the community as well. we will take care internally for the devices where we dont need this setting, by explictly setting device_metadata flag.

mssonicbld · 2026-02-20T20:41:04Z

Cherry-pick PR to 202511: #25604

mssonicbld · 2026-02-23T22:12:57Z

@lipxu cherry pick PR didn't pass PR checker. Please check!!!
#25604

mssonicbld · 2026-02-23T23:12:39Z

@lipxu cherry pick PR didn't pass PR checker. Please check!!!
#25604

mssonicbld · 2026-02-24T22:12:51Z

@lipxu cherry pick PR didn't pass PR checker. Please check!!!
#25604

) What is the motivation for this PR? The PR sonic-net/sonic-buildimage#25117 would update the default value of zebra_nexthop in FRR from disable to enable. Enable is the default configuration in FRR, it would not show in ouput of "show running" in FRR, so it would cause the test case test_frr_config_check failed once the previous PR merged. How did you do it? Skip the FRR hidden config check in test_frr_config_check. How did you verify/test it? Run the case in elastic with fix https://elastictest.org/scheduler/testplan/6978a655c5f3d9049fbb9138 without fix https://elastictest.org/scheduler/testplan/6978a6671910fa9097bc5e58 Signed-off-by: mssonicbld <[email protected]>

) What is the motivation for this PR? The PR sonic-net/sonic-buildimage#25117 would update the default value of zebra_nexthop in FRR from disable to enable. Enable is the default configuration in FRR, it would not show in ouput of "show running" in FRR, so it would cause the test case test_frr_config_check failed once the previous PR merged. How did you do it? Skip the FRR hidden config check in test_frr_config_check. How did you verify/test it? Run the case in elastic with fix https://elastictest.org/scheduler/testplan/6978a655c5f3d9049fbb9138 without fix https://elastictest.org/scheduler/testplan/6978a6671910fa9097bc5e58 Signed-off-by: mssonicbld <[email protected]> Co-authored-by: Liping Xu <[email protected]>

) Why I did it Set the default value to enabled Work item tracking Microsoft ADO (number only): 36367940 How I did it Set the default value to enabled How to verify it Local Signed-off-by: Feng Pan <[email protected]>

) What is the motivation for this PR? The PR sonic-net/sonic-buildimage#25117 would update the default value of zebra_nexthop in FRR from disable to enable. Enable is the default configuration in FRR, it would not show in ouput of "show running" in FRR, so it would cause the test case test_frr_config_check failed once the previous PR merged. How did you do it? Skip the FRR hidden config check in test_frr_config_check. How did you verify/test it? Run the case in elastic with fix https://elastictest.org/scheduler/testplan/6978a655c5f3d9049fbb9138 without fix https://elastictest.org/scheduler/testplan/6978a6671910fa9097bc5e58 Signed-off-by: Ravali Yeluri (WIPRO LIMITED) <[email protected]>

) What is the motivation for this PR? The PR sonic-net/sonic-buildimage#25117 would update the default value of zebra_nexthop in FRR from disable to enable. Enable is the default configuration in FRR, it would not show in ouput of "show running" in FRR, so it would cause the test case test_frr_config_check failed once the previous PR merged. How did you do it? Skip the FRR hidden config check in test_frr_config_check. How did you verify/test it? Run the case in elastic with fix https://elastictest.org/scheduler/testplan/6978a655c5f3d9049fbb9138 without fix https://elastictest.org/scheduler/testplan/6978a6671910fa9097bc5e58 Signed-off-by: Abhishek <[email protected]>

Why I did it Set the default value to enabled Work item tracking Microsoft ADO (number only): 36367940 How I did it Set the default value to enabled How to verify it Local Signed-off-by: dprital <[email protected]>

) What is the motivation for this PR? The PR sonic-net/sonic-buildimage#25117 would update the default value of zebra_nexthop in FRR from disable to enable. Enable is the default configuration in FRR, it would not show in ouput of "show running" in FRR, so it would cause the test case test_frr_config_check failed once the previous PR merged. How did you do it? Skip the FRR hidden config check in test_frr_config_check. How did you verify/test it? Run the case in elastic with fix https://elastictest.org/scheduler/testplan/6978a655c5f3d9049fbb9138 without fix https://elastictest.org/scheduler/testplan/6978a6671910fa9097bc5e58

) What is the motivation for this PR? The PR sonic-net/sonic-buildimage#25117 would update the default value of zebra_nexthop in FRR from disable to enable. Enable is the default configuration in FRR, it would not show in ouput of "show running" in FRR, so it would cause the test case test_frr_config_check failed once the previous PR merged. How did you do it? Skip the FRR hidden config check in test_frr_config_check. How did you verify/test it? Run the case in elastic with fix https://elastictest.org/scheduler/testplan/6978a655c5f3d9049fbb9138 without fix https://elastictest.org/scheduler/testplan/6978a6671910fa9097bc5e58 Signed-off-by: selldinesh <[email protected]>

) What is the motivation for this PR? The PR sonic-net/sonic-buildimage#25117 would update the default value of zebra_nexthop in FRR from disable to enable. Enable is the default configuration in FRR, it would not show in ouput of "show running" in FRR, so it would cause the test case test_frr_config_check failed once the previous PR merged. How did you do it? Skip the FRR hidden config check in test_frr_config_check. How did you verify/test it? Run the case in elastic with fix https://elastictest.org/scheduler/testplan/6978a655c5f3d9049fbb9138 without fix https://elastictest.org/scheduler/testplan/6978a6671910fa9097bc5e58

) What is the motivation for this PR? The PR sonic-net/sonic-buildimage#25117 would update the default value of zebra_nexthop in FRR from disable to enable. Enable is the default configuration in FRR, it would not show in ouput of "show running" in FRR, so it would cause the test case test_frr_config_check failed once the previous PR merged. How did you do it? Skip the FRR hidden config check in test_frr_config_check. How did you verify/test it? Run the case in elastic with fix https://elastictest.org/scheduler/testplan/6978a655c5f3d9049fbb9138 without fix https://elastictest.org/scheduler/testplan/6978a6671910fa9097bc5e58 Signed-off-by: opcoder0 <[email protected]>

set the default value for zebra_nexthop

c327f69

Signed-off-by: Liping Xu <[email protected]>

set enable as default for vs

72c0e1d

Signed-off-by: Liping Xu <[email protected]>

lipxu marked this pull request as ready for review January 19, 2026 22:38

lipxu requested review from lguohan and qiluo-msft as code owners January 19, 2026 22:38

lipxu requested review from StormLiangMS and Copilot January 19, 2026 22:38

Copilot started reviewing on behalf of lipxu January 19, 2026 22:38 View session

lipxu requested review from dgsudharsan, eddieruan-alibaba and r12f January 19, 2026 22:38

Copilot AI reviewed Jan 19, 2026

View reviewed changes

lipxu requested a review from deepak-singhal0408 January 20, 2026 03:46

deepak-singhal0408 mentioned this pull request Jan 20, 2026

FRR config "no zebra nexthop kernel enable" increases system mem usage in T2 downstream LC #24607

Closed

if no metadata value set the default value to enable

d6146f4

Signed-off-by: Liping Xu <[email protected]>

deepak-singhal0408 added this to SONiC Chassis Jan 20, 2026

deepak-singhal0408 approved these changes Jan 20, 2026

View reviewed changes

fix UT test

d534ef3

Signed-off-by: Liping Xu <[email protected]>

deepak-singhal0408 added the Request for 202511 Branch label Feb 20, 2026

vmittal-msft added the Approved for 202511 Branch label Feb 20, 2026

mssonicbld mentioned this pull request Feb 20, 2026

[action] [PR:25117] [FRR] set the enable as default value for zebra_nexthop #25604

Merged

8 tasks

mssonicbld added the Created PR to 202511 Branch label Feb 20, 2026

mssonicbld mentioned this pull request Feb 24, 2026

[action] [PR:22106] [test_frr_config_check] skip frr hidden config checking sonic-net/sonic-mgmt#22579

Merged

12 tasks

mssonicbld added Included in 202511 Branch and removed Created PR to 202511 Branch labels Feb 27, 2026

tirupatihemanth mentioned this pull request Mar 18, 2026

Memory regression: "no zebra nexthop kernel enable" workaround causes massive memory increase on scaled routes system, increase is seen in kernel percpu memory #24742

Closed

Conversation

lipxu commented Jan 18, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Why I did it

Work item tracking

How I did it

How to verify it

Which release branch to backport (provide reason below if selected)

Tested branch (Please provide the tested image version)

Description for the changelog

Link to config_db schema for YANG module changes

A picture of a cute animal (not mandatory but encouraged)

Uh oh!

mssonicbld commented Jan 18, 2026

Uh oh!

azure-pipelines bot commented Jan 18, 2026

Uh oh!

mssonicbld commented Jan 18, 2026

Uh oh!

azure-pipelines bot commented Jan 18, 2026

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

deepak-singhal0408 commented Jan 20, 2026

Uh oh!

mssonicbld commented Jan 20, 2026

Uh oh!

azure-pipelines bot commented Jan 20, 2026

Uh oh!

lipxu commented Jan 20, 2026

Uh oh!

mssonicbld commented Jan 20, 2026

Uh oh!

azure-pipelines bot commented Jan 20, 2026

Uh oh!

eddieruan-alibaba commented Jan 20, 2026

Uh oh!

deepak-singhal0408 commented Jan 20, 2026

Uh oh!

deepak-singhal0408 left a comment

Choose a reason for hiding this comment

Uh oh!

deepak-singhal0408 commented Feb 20, 2026

Uh oh!

mssonicbld commented Feb 20, 2026

Uh oh!

mssonicbld commented Feb 23, 2026

Uh oh!

mssonicbld commented Feb 23, 2026

Uh oh!

mssonicbld commented Feb 24, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

10 participants

lipxu commented Jan 18, 2026 •

edited

Loading