[chassis][sfm][swss] swss.sh shows backtrace for all empty SFM slots when shutdown or remove a SFM on Supervisor #18384
Assignees
Labels
Triaged
this issue has been triaged
Comments
mlok-nokia
changed the title
11 tasks
lguohan
pushed a commit
that referenced
this issue
Mar 30, 2024
…wn a SFM. (#18393) On a Supervisor card of a VOQ chassis, when remove or shutdown a Fabric card, swss.sh shows Stacktrace for all related empty SFM slots in the syslog file. This PR fixes #18384 How I did it In the asic_status.py, all empty SFM slots related swss.sh is in the waiting state to wait for the presence event of SFM -- SET operation. The subscriber event handler also includes the "DEL" operation when a SFM is shutdown/removal. When a SFM is shutdown, all empty slot's swss.sh also get the "DEL" event although it is not for them. In the "DEL" operation, the current implementation doesn't check if this "DEL" operation for them, and then they exit the wait state and proceed to docker-wait-any with wrong operation in the wrong slot. docker-wait0any raise the backtarce. How to verify it In a chassis which has some empty SMF slot, remove or shutdown a SFM. There should not be related stacktrace shown in syslog Signed-off-by: mlok <marty.lok@nokia.com>
mlok-nokia
added a commit
to mlok-nokia/sonic-buildimage
that referenced
this issue
Jun 5, 2024
…wn a SFM. (sonic-net#18393) On a Supervisor card of a VOQ chassis, when remove or shutdown a Fabric card, swss.sh shows Stacktrace for all related empty SFM slots in the syslog file. This PR fixes sonic-net#18384 How I did it In the asic_status.py, all empty SFM slots related swss.sh is in the waiting state to wait for the presence event of SFM -- SET operation. The subscriber event handler also includes the "DEL" operation when a SFM is shutdown/removal. When a SFM is shutdown, all empty slot's swss.sh also get the "DEL" event although it is not for them. In the "DEL" operation, the current implementation doesn't check if this "DEL" operation for them, and then they exit the wait state and proceed to docker-wait-any with wrong operation in the wrong slot. docker-wait0any raise the backtarce. How to verify it In a chassis which has some empty SMF slot, remove or shutdown a SFM. There should not be related stacktrace shown in syslog Signed-off-by: mlok <marty.lok@nokia.com>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Description
In the 202205 branch. on Supervisor card of a VOQ chassis, when remove or shutdown a Fabric card, swss.sh shows Stacktrace for all related empty SFM slots in the syslog file. Based on the investigation, this issue should be in all branches.
Steps to reproduce the issue:
Describe the results you received:
Describe the results you expected:
swss.sh should not show any stacktrace for this operation.
Output of
show version:Although the issue is shown on 202205 image, we believe it is in all branches.
admin@ixre-cpm-chassis17:~$ show version
SONiC Software Version: SONiC.20220532.55
SONiC OS Version: 11
Distribution: Debian 11.8
Kernel: 5.10.0-23-2-amd64
Build commit: 99f2e7b93b
Build date: Sat Feb 10 09:05:02 UTC 2024
Built by: cloudtest@65620ad2c000000
Platform: x86_64-nokia_ixr7250e_sup-r0
HwSKU: Nokia-IXR7250E-SUP-10
ASIC: broadcom
ASIC Count: 10
Serial Number: NS214510028
Model Number: N/A
Hardware Revision: 56
Uptime: 13:59:38 up 5 days, 15:01, 1 user, load average: 1.94, 1.98, 2.02
Date: Mon 18 Mar 2024 13:59:38
Output of
show techsupport:Additional information you deem important (e.g. issue happens only occasionally):
The text was updated successfully, but these errors were encountered: