Vivek, different file systems will behave differently and the same goes for different OSes.
In past, I've seen/had issue where the XFS (CentOS7 mostly) VMs would go read-only during issues with storage, while the EXT4 (various Linux variants) would NOT go read-only - during the intermittent issues (long enough outages would bring those down as well) Then also Linux in general will go read only FS, while Windows Server will simply reboot itself (and possibly successfully boot if it was an intermittent issue - thus resulting in auto-fix which is awesome IMO). Different timeouts/FS behave differently - this is what I can tell for sure. Fixing the underlying storage issue is a must. Regards, Andrija On Mon, 27 Apr 2020 at 22:49, Vivek Kumar <vivek.ku...@indiqus.com> wrote: > Hello Folks, > > I have a very interesting issue which occurs very frequently, I have > CloudStack version 4.7 with XenServer 7.0 clusters. We are using NFS for > primary storage, we have been observing that file system of virtual router > goes to Read-Only mode very frequently. However other VMs running on the > same SR doest have such issues but it happened specific for VR’s. > > I understand that Read-Only filesystem indicates that the filesystem is > either not able to access to backend disk somehow or it looses the > connectivity from backend SR but if my backend SR has issues then other VMs > running on the same SR should also get affected. Has anyone came across > with this issue ? > > So every-time i need to reboot the VR in order to fix the file system > which obviously leads some downtime. > > Regards > Vivek Kumar > > -- Andrija Panić
