You can probably use "Skylake-Server-v4". Generally you want every feature turned on that is supported by every CPU in your cluster. I don't know specifically for cascade lake vs skylake, but usually just picking the one that still works with...
You could try to catch the boot at the grub menu, and add "nomodeset" to the kernel command line. That should tell it not to change the video mode (resolution), so hopefully your screen won't go blank during the boot anymore. Hopefully that...
For your local LVM-thin results, a couple things to check: is iothread=1 actually set on the SCSI controller? Should be default with virtio-scsi-single on recent PVE but I've seen cases where it wasn't. Also try aio=io_uring instead of the...
The easiest option is to backup any VMs that matter from one of the nodes, then delete all of that node's VMs. Then you should be able to click in the UI on Datacenter and go to the storage section. Remove the storage for that node and create...
Was @Falk R. sagt: PBS schreibt deduplizierte Chunks, die alten Backups bleiben intakt. Das ist schon mal die halbe Miete.
Und @bl1mp hat recht, die Dedup-Rate ist genau die Metrik die du willst. In den PBS-Task-Logs siehst du nach jedem Backup...
Nicht vergessen, der PBS ist durch seine Deduplizierung nur mit SSDs optimal, was gerade beim Backup zu delays bei dem VMs führen kann.
Das sollte man auf jeden Fall im Hinterkopf behalten und ist kein Mangel an der Software.
So...the lookup happens in RAM?
Just out of curiosity - how much RAM does the CRUSH map usually consume?
Got it.
Then why did the ceph devs (in the Erasure Coding Enhancement document) quote:
"We want clients to submit small I/Os directly to...
I managed to fix my talos cluster by running the following (or adding the args line manually to the config):
qm set 195 --args '-global virtio-net-pci.host_tunnel_csum=off -global virtio-net-pci.csum=off'
@fba hat das richtig beschrieben. Einfach ein LV aus dem Thin-Pool anlegen, ext4 drauf, mounten. Dann im PVE unter Datacenter → Storage ein neues Directory-Storage hinzufügen mit Pfad /mnt/iso-store und als Content "ISO image" und "Container...
Falk hat recht, ohne Redundanz wird das auf Dauer eng. Aber praktischer ist es, ZFS Replication (pvesr) zwischen den Hosts einzurichten, das geht auch mit lokalem Storage. Dann kannst du VMs live migrieren, den Node in Ruhe rebooten und...
Genau, Bugzilla ist die richtige Anlaufstelle. Wenn du dort ein Ticket erstellst, wähl als Severity "enhancement" aus, dann landets auch richtig. Pro Feature am besten ein eigenes Ticket, nicht alles in eins packen.
Know little about Gluster, so can't compare features/performance, but wanted to clarify some points regarding Ceph (again):
Among other tasks, Ceph MON will provide a copy of the CRUSH map to every component in the Ceph cluster, including...
That sounds good then. I think I'll go the VPN route since that'll just be a case of creating a new client and feels like it would be a bit more secure to create the connection.
Reading through the documentation it seems I would configure my...
I'm aware of the ZFS special VDEV approach but that's not what I'm looking for here. I'm looking for two separate repos: one that handles all the snapshot files and the other that handles all the chunks. This way I'd be able to sync them...
Never heard of Fusion-io SSDs and I doubt my ssd is connected to one, though it is on a extend pcb. Again when I'm back at the machine I can take a look