On Wed, 2022-04-27 at 22:10 +0200, Valentin Vidić via Users wrote: > On Thu, Apr 28, 2022 at 12:25:37AM +0500, Umar Draz wrote: > > * sharedfs1_start_0 on g2fs-1 'error' (1): call=158, > > status='complete', > > exitreason='Couldn't mount device [/dev/shared_vg1/shared_lv1] as > > /mnt/webgfs', last-rc-change='Tue Apr 26 01:07:45 2022', > > queued=0ms, > > exec=806ms > > Maybe the gfs2 filesystem was created with 2 journals so it only > supports 2 hosts at the same time? >
This does seem unrelated to the cluster software per se, and more likely to be a GFS2 issue. But for the record, the upstream software has no official limit on node count; the only limits are in practice (network capabilities vs. Corosync's token passing algorithm). Commercial support like RHEL generally imposes their own limits based on practicality -- often 16 or 32 full cluster nodes (more are possible with Pacemaker Remote). -- Ken Gaillot <kgail...@redhat.com> _______________________________________________ Manage your subscription: https://lists.clusterlabs.org/mailman/listinfo/users ClusterLabs home: https://www.clusterlabs.org/