Conversation
Signed-off-by: Pratik Dam <pdam@arista.com>
|
/azp run |
|
Azure Pipelines: There may be pipelines that require an authorized user to comment /azp run to run. |
|
Azure Pipelines: Successfully started running 1 pipeline(s). |
|
arp/test_neighbor_mac.py failed in pipeline . Its not directly related to my fix . We could address this subsequently |
|
This PR has backport request for branch(es): 202605. ---Powered by SONiC BuildBot
|
|
The Tested branch section has been ticked and Test result is provided for branch(es): 202605. Added label(s): Tested for 202605 Branch. ---Powered by SONiC BuildBot
|
|
/azp run |
|
Azure Pipelines: There may be pipelines that require an authorized user to comment /azp run to run. |
|
Azure Pipelines: Successfully started running 1 pipeline(s). |
|
@pdam-arista — this 202605 backport request is not approval-ready yet. Please update the PR description with:
This information needs to be in the backport/test sections so the request is traceable and BuildBot can validate it. |
Hi @StormLiangMS , updated the necessary information |
|
Cherry-pick PR to 202605: #27337 |
Description of PR
PR #26376 added dual-ToR golden config generation support for
prober_typeand
neighbor_mode. As part of that change, dualtor topologies startedgenerating
golden_config_db.jsonwith aPORTtable copied from minigraph.The deploy flow already runs
config load_minigraph -ybefore golden configgeneration for
lt2andft2, because those topologies modify/copyPORTand need the DUT config to reflect testbed-specific interface admin state
first. The new dualtor golden config path missed this preload step.
Without the preload, unused routed ports can be preserved as
admin upingolden_config_db.json. These ports have no link, remainoper down, andpretest interface sanity fails.
This PR extends the existing preload condition to include dualtor.
Summary:
Fixes #26673
Type of change
Back port request
Tracking issue/work item for backport/cherry-pick request (GitHub issue or Microsoft ADO):
Failure type: regression
Tested branch
Test result
202605
Failure Results
test_pretest.log
test_pretest.xml
Pass Results (after fix)
test_pretest.log
test_pretest.xml
sonic-mgmt change under test:
Note: Both before and after tests used the same SONiC image/build; the validated change is in sonic-mgmt.
Approach
What is the motivation for this PR?
Fix a regression introduced when dualtor was added to the golden config
generation path. dualtor golden config now copies
PORT, so it needs thesame minigraph preload step that already exists for
lt2andft2.How did you do it?
Updated
ansible/config_sonic_basedon_testbed.ymlso the existingconfig load_minigraph -ystep runs for dualtor beforegolden_config_db.jsonis generated.How did you verify/test it?
Inspected failed pretest logs from tst-esx-31.
All failed during pretest setup sanity with unused routed ports
left as admin up / oper down.
Applied the fix and sanitized the duts , ran deploy-mg and pret-tests
Any platform specific information?
N/A
Supported testbed topology if it's a new test case?
N/A
Documentation
N/A