Glad it helped, @ns33.
With U.2 and 100Gb, PCIe placement matters. On dual socket, keep your NVMe and the NIC carrying Ceph traffic on the same socket so I/O isn't crossing the inter-socket link. cat /sys/bus/pci/devices/<addr>/numa_node will...
Hi @Zernox,
Your post is confusing because you seem to jump between the VM and Template but the connection is not there.
For example, in your Proxmoxer config (btw this is a 3rd party tool and if you really have questions/issues with it - you...
Hi @ns33,
It really depends on your applications. Neither vendor is a universally "better" choice, and the right answer comes from the workloads you plan to run.
That said, my general preference is single-socket servers for anything network or...
@ghusson, this thread is specifically about FC connectivity with a particular QLogic card, so I don't think recommending iSCSI would be directly applicable to the posters' situation.
That said, the best way to confirm hardware compatibility...
A new article finally showing what a small ARM box running PVE is capable of:
https://forum.proxmox.com/threads/pve-9-2-arm64-on-asus-gx10-gb10-6-gb-s-of-nvme-tcp-out-of-the-box-29-gb-s-after-two-kernel-flags.186770/
Blockbridge : Ultra low...
One more GX10 post, the last one for now. After battling the black screen, the hard resets and the power cap, we were finally able to see what this box and PVE on arm64 can do.
Out of the box, the kernel initiator managed NVMe/TCP reads at 6.3...
we already have previous reports about HPE MSA (and Dell) appliances misbehaving in exactly this fashion - they advertise support for zeroing unmap, but only do it when the request matches some conditions, and simply ignore the zeroing part...
Hi @aday , welcome to the forum.
The journal line Cannot find device='z…' nor node-name='z…' on the target comes from the incoming dirty-bitmap migration: it couldn't find the block node the bitmap was attached to, so QEMU dropped the bitmap and...
I'd recommend trying to reproduce this directly on the PVE host, bypassing the VM. Allocate a raw volume with pvesm alloc and test against the block device directly (or mkfs and mount it).
My suspicion is that LVM doesn't clear extents on...
Hi @eatonjb, welcome to the forum.
The first thing I would check is the system log:journalctl. There may be some clues there.
If the customer has to manually power-cycle the server, that may indicate that there are messages on the console. If a...
Hi @mira,
I’ll check with the developers on the current status. They are currently working through a number of existing deliverables, so this is in the queue but is not something I expect to be completed immediately.
Since we have not seen this...
You could try virt-sparsify on your file to see if that helps.
Blockbridge : Ultra low latency all-NVME shared storage for Proxmox - https://www.blockbridge.com/proxmox
@bbgeek17 no disagreement. I was simply pointing out that the premise that a full node fence is required isnt necessarily true for the proposed solution to work. I'm not suggesting its easy, not fraught with gaping holes, probably too slow to be...
Sorry, I missed the forum name - that documentation/procedure is for the PVE Cluster.
If you have collection of single nodes under PDM, take a look at this thread:
https://forum.proxmox.com/threads/how-to-remove-node-from-remotes.163437/...
Hi @bishoptf ,
Have you checked the cluster guide? https://pve.proxmox.com/pve-docs/chapter-pvecm.html#_remove_the_cluster_node
Blockbridge : Ultra low latency all-NVME shared storage for Proxmox - https://www.blockbridge.com/proxmox
Hi Mike,
It's a good idea, but it's worth noting that discard has a rough history. SATA TRIM often didn't guarantee zeros, a bunch of drives corrupted data on queued TRIM and are still on the kernel blacklist, and arrays have been all over the...
A new article for ARM system users covering performance troubleshooting
Blockbridge : Ultra low latency all-NVME shared storage for Proxmox - https://www.blockbridge.com/proxmox
Now that the system is "stable" after dealing with the black screen and the watchdog resets, we've started looking at performance. Our first look at storage performance was absolutely terrible. And after some head scratching, learning a bit more...
Hi @alexskysilk , you know I love a good bash loop, but ACL toggling is vendor-specific. It often only gates new logins, so an established session keeps writing until the target decides to drop it, and whether it does that on revoke is up to the...