Search results

Siehe auch https://pve.proxmox.com/wiki/Upgrade_from_8_to_9#VM_Memory_Consumption_Shown_is_Higher

Swap is more than just an escape for low memory: https://chrisdown.name/2018/01/02/in-defence-of-swap.html But given that the host has ~185G of memory, you could consider disabling swap, as that is a lot of memory, and if you run out of memory, those 8G of swap are most likely not enough...

Hmm, those versions look new enough. Can you please restart the pvestatd service on the hosts? Either in the Node→System panel, or with systemctl restart pvestatd Does that help to get rid the log messages?

Yeah. I assume you have one interface for everything on the hosts, that goes to the switch, right? The single point of failure there is the switch. If you can add a direct cable between the hosts, without a switch in between, you can configure a second IP subnet on it and add it as second...

Well, those long timeouts are most likely the explanation. If corosync takes too long to form a new quorum with just the QDevice, it might take longer than the 60s timeout of the LRM! Please set it back to defaults, from one of my test clusters: quorum { provider: corosync_votequorum } totem...

Can you please post your /etc/pve/corosync.conf file? And make sure that the /etc/pve/corosync.conf and /etc/corosync/corosync.conf files are the same.

Any tool that allows booting a live system on the physical host and the target VM to transfer disk contents. That can be just a regular Linux live system and dd + ssh on both sides. Or something more guided like Clonezilla. There are surely also other more commercial tools out there that can do...

Has that host been installed a while ago? Because, IIRC, since about 8.1, the installer limits the ARC by default. If you installed earlier, you can manually set a limit on the ARC: https://pve.proxmox.com/pve-docs/chapter-sysadmin.html#sysadmin_zfs_limit_memory_usage What has changed with the...

yeah, if you can do another test, I am interested in the pvecm status output of node pve/Node1 a few seconds after you disconnected pve1/Node2, but before it will eventually fence itself (if something is wrong). I think, that info is not yet in the previous posts, unless I missed it. Because...

There is a misunderstanding in how fencing works. It is handled by the LRM on each node. If it is in "active" mode and the host lost the connection to the quorum for more than 60 seconds, it will not renew the watchdog. Once the watchdog runs out, the host will reboot/fence. So you would see a...

Can you disable the HA resource, wait ~10 minutes until all LRMs are idle, and then do the following please? With no active LRM, the nodes won't fence. 1. get pvecm status while all nodes are up and working 2. disconnect one of the nodes 3. get pvecm status from the node that can still talk to...

Looks a lot better :)

not great, because the Ceph Cluster network is only used for the replication between the OSDs. Everything else, also IO from the guests to their virtual disks, is going via the Ceph Public network. Which is probably why you see the rather meh performance in the benchmark, as the benchmark client...

Does the Ceph network actually provide 10Gbit? Check with ethtool if they autonegiotated to the expected speed, and if, run some iperf tests between the nodes to see how much bandwidth it can provide. Are both, Ceph Public and Ceph Cluster network on a 10Gbit Link or is one on a slower link?

Hmm, can you post the output of pveversion -v on the host where you have/had guest 130? Ideally inside of tags (or use the formatting options in the buttons above (</> for a code block) Is guest 130 a VM or a CT?

In the ceph pg dump_stuck output you have the columns ACTING_PRIMARY, where the replica(s) currently are. and UP_PRIMARY, where they should be.

okay. that is curious. are all guests powered on or are some powered off? For example, guest VMID 130 in that error message from the first post. Was it on or off at that time?

Weils im englischen Forum auch gerade vorkam, hier meine Antwort dort mit ein paar Details bez. des jetzt einfacheren Pinnings: https://forum.proxmox.com/threads/network-drops-on-boot.65210/#post-793255

Since the Proxmox VE 9 release, and I think in the very latest 8.4, there is now the pve-network-interface-pinning tool. This makes it a lot easier to ping NICs to a specific name. And you can even choose a more fitting name. For example enphys0 or enguest0. This avoids automatic renames if the...

Hey, are all nodes running on Proxmox VE 9 by now? If so, do you see files for all guests (VMs and CTs) on all hosts in the /var/lib/rrdcached/db/pve-vm-9.0 directory?