Separate names with a comma.
I know that it doesn't make a lot of sense, but it allows to restart VM on the same node when not running or fails to start.
Anyway in the docs:...
i have VMs in HA state, but their disks are locally (lvm). When i set parameters:
webgui sets: restart 3, relocation 0
webgui shows: restart...
Can't help you with all questions. But my servers:
1] vlan's works with no-subs repo
2] using ovs (from pve no-subs repo)
3] all config written...
one of our pve hosts totally freezed. The only notice in logs was information about corosync delay warning (more than 2500ms), those warnings...
Can you do this:
1] add dedicated HDD/SSDs as datastorage with zfs-> test job
2] add dedicated HDD/SSDs as datastorage withous zfs -> test job
You need create separate user in PVE because there isn't any ability to manage user permissions via AD groups.
Your dd tests are useless. Use fio.
And you forgot, backup is compressed. Thats a big slowdown.
I never saw more than 200MBps for proxmox backup...
Never rely on the proxmox built-in backup for applications (only when file based). Anything with db needs its own application backup.
Try different kernel.
You can make test.
Remove HDD from HPE what it will do. Check what HPE/PVE/VM logs/dumps about disks. I think, losing disk create some unstability...
What OS is on those two VMs vs others? There must be something what make VM system crash/spinlock. What about remote syslog/netdata etc?
Your problem is, you don't understand what LV vs VG is.
In case 1]:
- you created LV as raid 10, mounted to pve, but without FS its not usable for...
Any debug from inside of the VMs?
When HDD dies on HPE storage, root fs in VM can switch to ro state? Impressive.
How is HPE connected with PVE nodes?
1] try different kernel (even standard debian kernel, not just pve)
2] check firmware
Information, what you have configured is more important, than you think. Tell me, how much vCPU you have configured on node X. You need see this...
But i DON'T WANT to overprovision. I want be warned/stopped before overprovision in configuration.
So explain me (use case the first post):
That's not the problem. Problem is, there isn't any warning (or ability to block) when overprovisioning disk space in config occurs. None user...
But i am using THICK method. There is no way to create RAW image with max filesize from webgui. And even discard/fstrim etc doesn't resize file to...
i did this:
1] vg pve
2] lv local-lvm-data (lvm-thick), size 100GB, ext4, mounted AS directory -> VM images are files. Used 50%.