[gpfsug-discuss] How to ignore ib_rdma_nic_unrecognized event on nodes where an IB link is not used.
Ran Pergamin
rpergamin at ddn.com
Wed May 29 13:26:40 BST 2019
Thanks All. Solved it.
The other port Link Layer was in autosense rather than IB.
Once changed the Link Layer to IB the false report cleared. I assume that’s the auth fix that was applied.
Regards,
Ran
From: <gpfsug-discuss-bounces at spectrumscale.org> on behalf of Mathias Dietz <MDIETZ at de.ibm.com>
Reply-To: gpfsug main discussion list <gpfsug-discuss at spectrumscale.org>
Date: Wednesday, 29 May 2019 at 15:20
To: gpfsug main discussion list <gpfsug-discuss at spectrumscale.org>
Subject: Re: [gpfsug-discuss] How to ignore ib_rdma_nic_unrecognized event on nodes where an IB link is not used.
Hi Ran,
please double check that port 2 config is not yet active for the running mmfsd daemon. When changing the verbsPorts, the daemon keeps using the old value until a restart is done.
mmdiag --config | grep verbsPorts
Mit freundlichen Grüßen / Kind regards
Mathias Dietz
Spectrum Scale Development - Release Lead Architect (4.2.x)
Spectrum Scale RAS Architect
---------------------------------------------------------------------------
IBM Deutschland
Am Weiher 24
65451 Kelsterbach
Phone: +49 70342744105
Mobile: +49-15152801035
E-Mail: mdietz at de.ibm.com
-----------------------------------------------------------------------------
IBM Deutschland Research & Development GmbH
Vorsitzender des Aufsichtsrats: Martina Koederitz, Geschäftsführung: Dirk WittkoppSitz der Gesellschaft: Böblingen / Registergericht: Amtsgericht Stuttgart, HRB 243294
From: Ran Pergamin <rpergamin at ddn.com>
To: gpfsug main discussion list <gpfsug-discuss at spectrumscale.org>
Date: 29/05/2019 13:54
Subject: [EXTERNAL] [gpfsug-discuss] How to ignore ib_rdma_nic_unrecognized event on nodes where an IB link is not used.
Sent by: gpfsug-discuss-bounces at spectrumscale.org
________________________________
Hi All,
My customer has some nodes in the cluster which current have their second IB port disabled.
Spectrum scale 4.2.3 update 13.
Port 1 is defined in verbs port, yet sysmoncon monitor and reports error on port 2 despite not being used.
I found an old listing claiming it will be solved in in 4.2.3-update5, yet nothing in 4.2.3-update7 release notes, about it.
https://www.spectrumscale.org/pipermail/gpfsug-discuss/2018-January/004395.html
Filters in sensor file say filters are not support + apply to ALL nodes, so no relevant where I need to ignore it.
Any idea how can I disable the check of sensor on mlx4_0/2 on some of the nodes ?
Node name: cff003-ib0.chemfarm
Node status: DEGRADED
Status Change: 2019-05-29 12:29:49
Component Status Status Change Reasons
-------------------------------------------------------------------------------------------------------------------------------------------------
GPFS TIPS 2019-05-29 12:29:48 gpfs_pagepool_small
NETWORK DEGRADED 2019-05-29 12:29:49 ib_rdma_link_down(mlx4_0/2), ib_rdma_nic_down(mlx4_0/2), ib_rdma_nic_unrecognized(mlx4_0/2)
ib0 HEALTHY 2019-05-29 12:29:49 -
mlx4_0/1 HEALTHY 2019-05-29 12:29:49 -
mlx4_0/2 FAILED 2019-05-29 12:29:49 ib_rdma_link_down, ib_rdma_nic_down, ib_rdma_nic_unrecognized
FILESYSTEM HEALTHY 2019-05-29 12:29:48 -
apps HEALTHY 2019-05-29 12:29:48 -
data HEALTHY 2019-05-29 12:29:48 -
PERFMON HEALTHY 2019-05-29 12:29:33 -
THRESHOLD HEALTHY 2019-05-29 12:29:18 -
Thanks !
Regards,
Ran
_______________________________________________
gpfsug-discuss mailing list
gpfsug-discuss at spectrumscale.org
http://gpfsug.org/mailman/listinfo/gpfsug-discuss
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://gpfsug.org/pipermail/gpfsug-discuss_gpfsug.org/attachments/20190529/b7006ec1/attachment.htm>
More information about the gpfsug-discuss
mailing list