Search results

Right What I am suggesting is that it would be great if we could define the allowed cache types for a given storage. That would allow me to prevent the use of specific cache types that are known to be bad for that particular storage. Something like this in storage.cfg: lvm: lms-storage vgname...

I was reading a thread where @wbumiller mentioned that using O_DIRECT with mdraid can result in inconsistent arrays. https://forum.proxmox.com/threads/proxmox-4-4-virtio_scsi-regression.31471/page-2#post-159574 On DRBD we have the same issue where some cache types can result in out of sync...

I'm running jewel 10.2.5-1~bpo80+1

I use fstrim and it was not running at the time this occurred. I changed to virtio-scsi so I could use fstrim to reclaim some space in CEPH. Ran fstrim once a couple of weeks ago.

When this problem happens the KVM process dies. Never had this problem untilI changed from virtio to virtio-scsi-single, also happened with virtio-scsi vm.conf: args: -D /var/log/pve/105.log boot: cd bootdisk: ide0 cores: 4 ide0: ceph_rbd:vm-105-disk-1,cache=writeback,size=512M ide2...

That's great and all that linbit providers repo but will that last? Will they improve the proxmox plugin? Continue to keep it updated? Assuming they do all that, when I have an issue and I have some storage related issue on my DRBD backed volume I now have to get support from two companies who...

I have had lots of issues with drbd 9 in proxmox 4.x early on, seems to be ok now tho. In Proxmox 4 the drbd storage plugin used drbdmanage to create individual drbd volumes for each vm disk. I've not followed all the mailing list threads but it looks like linbit plans to maintain the proxmox...

I applied this change to the host. I use drbd on many hosts, I think that same drbd link you provided is where I got the idea to change min_free_kbytes. On any host with 32GB ram or more and in all my vms I disabled disk swap and added zram swap. Got tired of high ram usage vms that sat idle...

No idea if this will help but it seems to be related to video so have you tried different video options on the vm hardware tab?

Try running "apt-get update" first, then "apt-get install ncdu"

@gkovacs in the past I've increased min_free_kbytes on some servers that were having io stall issues during vzdump operations and it helped. Any thoughs as to why this is necessary?

CEPH is a shared nothing, distributed, highly available system and ZFS is none of that. Backup storage is mission critical when the only way to restore failed services is to restore data from backup.

A kernel panic in the guest leaves the guest running, I'd open the console for that guest in Proxmox and see the panic. In my case the kvm process itself exited, likely due to a segfault or some other serious error. Might be a different problem but it happened for the first time after changing...

I would not consider any online storage system a backup, maybe part of a backup system but not a complete backup. More related recommendations here: https://forum.proxmox.com/threads/howto-upgrade-ceph-hammer-to-jewel.31692/#post-158359 I think it would be great to be able to use cephfs for...

I changed a vm from virtio to virtio scsi single Friday. The vm disk is 4TB ceph rbd formatted in guest as ext4. Made the change so i could run fstrim in the guest. The kvm process died while vzdump was running on Sunday. Logs in guest just abruptly end and only log in host was vzdump: 105: Jan...

I recently converted a vm to virtio scsi and it crashed during vzdump backup. The kvm process vanished.

I've run a windows VM inside Proxmox that was nested inside Proxmox but that was like 4 or 5 years ago with, I think, Proxmox 1.x My Google fu says that only windows 2016 can run a vm inside of nested hyper-v and it looks like that only works when nested in hyper-v and maybe vmware too.

You could write a vzdump hook script that performs the actions desired. Read the comments, it has an example of scp the backup file to remote host, add in your copy to tape commands there. After copied to tape you could delete the file too. You can write the hook script in any language, the...

I don't use zfs so I have no clue if that might be part of the issue or not. The pveperf looks OK. How are your VM disks configured? Maybe you can just post the vm.conf files.

When changing windows boot disks I have had luck with this process: Shutdown widows Add a temporary disk of the new type Start windows Install driver for new controller Shutdown windows Change boot disk to new type Start windows I've not tried virtio scsi on windows so no idea if that even...