Can you also check the time differences between nodes?
We had a situation recently where the server time mismatch caused failures.
On Thu, Jun 28, 2018 at 2:50 AM, Kevin D Johnson
wrote:
> You can also try to convert to the old primary/secondary model to back it
> away from the default CCR
You can also try to convert to the old primary/secondary model to back it away from the default CCR configuration.
mmchcluster --ccr-disable -p servername
Then, temporarily go with only one quorum node and add more once the cluster comes back up. Once the cluster is back up and has at least
Hi Renata,
You may want to reduce the set of quorum nodes. If your version supports
the --force option, you can run
mmchnode --noquorum -N --force
It is a good idea to configure tiebreaker disks in a cluster that has only
2 quorum nodes.
Regards, The Spectrum Scale (GPFS) team
Hi Simon, yes I ran
mmsdrrestore -p
and that helped to create the /var/mmfs/ccr directory which was
missing. But it didn't create a ccr.nodes file, so I ended up scp'ng
that over by hand which I hope was the right thing to do. The one
host that is no longer in service is still in that
Hi, any gpfs commands fail with:
root@ocio-gpu01 ~]# mmlsmgr
get file failed: Not enough CCR quorum nodes available (err 809)
gpfsClusterInit: Unexpected error from ccr fget mmsdrfs. Return code: 158
mmlsmgr: Command failed. Examine previous error messages to determine cause.
The two "working"
Hi, Have you check if there is any manager node available? #mmlsmgrIf not could you try to asig a new cluster/gpfs_fs manager.Mmchmgr gpfs_fs. Manager_nodeMmchmgr. -c. Cluster_manager_nodeCheers. ___
gpfsug-discuss mailing list
gpfsug-discuss at
Have you tried running mmsdrestore in the reinstalled node to reads to the
cluster and then try and startup gpfs on it?
https://www.ibm.com/support/knowledgecenter/STXKQY_4.2.3/com.ibm.spectrum.scale.v4r23.doc/bl1pdg_mmsdrrest.htm
Simon
From:
Hi, we have a client cluster of 4 nodes with 3 quorum nodes. One of the
quorum nodes is no longer in service and the other was reinstalled with
a newer OS, both without informing the gpfs admins. Gpfs is still
"working" on the two remaining nodes, that is, they continue to have access
to the
Hi,
we currently support the SMB protocol method of quering snapshots, which is used by the Windows "Previous versions" dialog. Mac clients unfortunately do not implement these explicit queries. Browsing the snapshot directories with the @GMT names through SMB currently is not supported.
Could
Hi Andreas,
Output of the debug log – no clue, but maybe you can interpret it better
[root@icgpfsq1 ~]# /usr/lpp/mmfs/gui/cli/runtask pm_monitor --debug
debug: locale=en_US
debug: Raising event: gui_pmcollector_connection_ok, for node:
localhost.localdomain
err:
Hi Richard,
if you double-click the event there should be some additional help available. The steps under "User Action" will hopefully help to identify the root cause:
1.) Check if there is additional information available by executing '/usr/lpp/mmfs/gui/cli/lstasklog [taskname]'.2.) Run the
Hi Renar,
No, it all runs over the same network.
Thanks,
Richard
From: gpfsug-discuss-boun...@spectrumscale.org
[mailto:gpfsug-discuss-boun...@spectrumscale.org] On Behalf Of Grunenberg, Renar
Sent: 27 June 2018 12:29
To: 'gpfsug main discussion list'
Subject: Re: [gpfsug-discuss] PM_MONITOR
Hi all,
our (Windows) users are familiared with the 'previous versions' self-recover
feature.
We honor this by creating regular snapshots with the default @GMT prefix
(non-@-heading prefixes are not visible in 'previous versions').
Unfortunately, MacOS clients having the same share mounted via
Hallo Richard,
do have a private admin-interface-lan in your cluster if yes than the logic of
query the collector-node, and the representing ccr value are wrong. Can you
‘mmperfmon query cpu’? If not then you hit a problem that I had yesterday.
Renar Grunenberg
Abteilung Informatik – Betrieb
Hi all,
I'm getting the following error in the GUI, running 5.0.1:
"The following GUI refresh task(s) failed: PM_MONITOR".
As yet, this is the only node I've upgraded to 5.0.1 - the rest are running
(healthily, according to the GUI) 4.2.3.7. I'm not sure if this version
mismatch is relevant
15 matches
Mail list logo
