---
** [tickets:#1059] 2PBE: cluster reset observed during switchovers**
**Status:** unassigned
**Milestone:** 4.3.3
**Created:** Wed Sep 10, 2014 09:57 AM UTC by Sirisha Alla
**Last Updated:** Wed Sep 10, 2014 09:57 AM UTC
**Owner:** nobody
The issue is seen on SLES X86. OpenSAF is running
SC-2 logs
Attachment: SLOT2.tar.bz2 (8.9 MB; application/x-bzip)
---
** [tickets:#1059] 2PBE: cluster reset observed during switchovers**
**Status:** unassigned
**Milestone:** 4.3.3
**Created:** Wed Sep 10, 2014 09:57 AM UTC by Sirisha Alla
**Last Updated:** Wed Sep 10, 2014 09:57 AM UTC
- **Component**: unknown -- clm
- **Comment**:
The direct cause of the cluster reset is that CLM exits on receiving
ERR_EXIST on implementerSet.This could a case of #946
(fixed in 5722:d353ca39b3d9).
If not then someone needs to analyze what CLM is doing (it is detaching as main
OI and fails
The CLM problem actually only explains why this SLOT2 went down.
This was a switchover, not a failover, so the other SC should have
reverted the switchover.
---
** [tickets:#1059] 2PBE: cluster reset observed during switchovers**
**Status:** unassigned
**Milestone:** 4.3.3
**Created:** Wed
And one more comment:
I dont think that this incident is logically related to 2PBE.
A PRTO fails to get created.
That is possibly more likely to happen with 2PBE than 1PBE or 0PBE,
but logically it can at least also happen also with 1PBE.
---
** [tickets:#1059] 2PBE: cluster reset observed
A. SLOT1 node went down:
1. CLM got BAD_HANDLE and finalizes the handle
Sep 10 14:56:51.543332 osafclmd [7511:imma_oi_api.c:0622] saImmOiFinalize
Sep 10 14:56:51.543370 osafclmd [7511:imma_oi_api.c:0626] T2 ERR_BAD_HANDLE: No
initialized handle exists!
2. Discard implementer is called
Sep
] [opensaf:tickets] #1059 2PBE: cluster reset observed during
switchovers
A. SLOT1 node went down:
1. CLM got BAD_HANDLE and finalizes the handle
Sep 10 14:56:51.543332 osafclmd [7511:imma_oi_api.c:0622] saImmOiFinalize
Sep 10 14:56:51.543370 osafclmd [7511:imma_oi_api.c:0626] T2 ERR_BAD_HANDLE