The PBS S3 client already has a retry logic (up to 3 times) with exponential backoff time, see [0]. So the 503 errors you get in your log are after having already retried. Currently 429 status codes are not retired, that can be easily added...
You could perform the full sync to a different namespace first, then use the move feature introduced in PBS 4.2 to merge contents. See https://pbs.proxmox.com/docs/storage.html#storage-move-namespaces-groups
Edit: fixed link
Those were hacks and only accelerated specific workloads or patterns. Cache misses still gave the worst case scenario performance and VM and databases both are workloads that should be calculated accordingly (unlike filer NAS).
Ceph/Linux has a...
Spiel doch erstmal mit deiner vorhanden Alt-HW rum. Installiere den vorgeschlagenen Standard. Verzichte auf RAID und weitere Optionen.
Mach dir dann Gedanken, welchen Speicherbedarf du hast.
Sehe ich auch so. Fehler steht ja schon in dmesg drin:
Jun 30 13:39:33 pve QEMU[64897]: KVM: entry failed, hardware error 0xffffffffffffffff
Jun 30 13:41:15 pve kernel: kvm: kvm [68075]: vcpu0, guest rIP: 0xffffffff8388840f Unhandled...
I am using a Supermicro X12SAE, the HBA I'm using takes up two slots so it's in the first x16 slot, the GPU is in the second x16 slot, the coral is in the x1 slot and the x4 slot is unused. I'm not sure the x4 slot will work with my x8 HBA though.
Could you please tell us on which API endpoints did you experienced issues? What issues exactly? Was it simply slow response times? You can use your browse debugger to see the outgoing API requests.
What does it actually want to fetch? there's a single api call that gives you all guests and basic (status) info already via /cluster/resources and that is cached and fast. Full config is different, there isn't also really a single API call for...
In my case it is using Proxmoxer and simply fetching all the VMs at the same time will time out the API. We have to paginate. There are also some race conditions, where creating a VM returns from the API, then it takes some time to return its...
Naja für virtuelle Maschinen eignet sich der Host natūrlich nicht wirklich, aber Container gängen schon. Würde dann aber auch eher omv plus docker-Container nehmen
Hi @Taledo,
Perhaps adding a few more full page screenshots would help here. Its unclear where this data came from exactly, or how it correlates to your actual hosts. Providing exact CLI metrics data from the hosts would be helpful as well...
Hi, there is a tool/command that can explore a vm on datastore where is present the backup and estimate the size?
Vm has sometimes more than one disk and also the prune keep few days in my configuration.
If I choose a backup, is possibile to...
Checked a few of those solutions and even if the concept is nice and sure they work perfectly, for me they fall into the niche solutions when you can't use Ceph for performance reasons, mostly disk latency. None beats Ceph in data resiliency and...
I'm migrating some VMs to different storage. Some of those VMs are linked clones, the parent of which would still be on the old storage.
Should this go well? I tried this in a lab environment, and it looks like the linked clone is converted to a...
if you read the topic, and if you have same problème, then fleecing greatly help.
But as we have not info about your case, nobody can guess.
Open new topic is better with your specs and your flow
Hi,
Seit kurzem funktioniert die Installation von Ubunt 26.04 Server nicht mehr.
Steps:
VM mit ubuntu-26.04-live-server-am64
8GB Ram
2 CPU
Wenn die Installation anlaeuft, gebe ich eine statische IP Adresse ein.
Danach kann der Mirror nicht...
I tried to ping with different MTU.
And after many pings 1410 bytes does work, so might be 1438 bytes without ICMP header.
So my IPsec header should be 62 bytes
Got it, thanks for the help. Is there any way to get the two PCIe slots (x16 and x8) to be in different IOMMU groups or is that a hardware limitation?
100 and 101 are different versions of the same VM I'm testing and they aren't running at the...