Recent content by AppState95

After setting the zfs_arc_min and zfs_arc_max as mentioned above, the RAM utilized in my cluster nodes appears to now be stable, hovering around 19-20% or ~150gb out of 756gb. Thanks again for the help. I think for my environment, at least, this issue is resolved.

Thanks, VictorSTS and Fiona. I set the minimum and maximum arc values for my server nodes as instructed in the article. I now have the minimum set to 24gb (I have roughly 21.8TB ZFS per node, so ~22gb + 2gb=24gb) and max at 128gb (or 1/6th total RAM). That immediately trimmed the RAM usage...

Hi Fiona, This morning the RAM usage is up to ~44-45% or ~336-345gb out of 756gb. There were new updates again today, so I installed them, but none appeared to require rebooting, so I have not rebooted the nodes yet. Here is the output from the cat /proc/spl/kstat/zfs/arcstats | grep -e...

This morning all three nodes are using close to 34% of available RAM (~256 out of 756gb). root@phantom-pve1:~# cat /proc/cmdline BOOT_IMAGE=/boot/vmlinuz-6.8.8-4-pve root=/dev/mapper/pve-root ro quiet root@phantom-pve1:~# lscpu Architecture: x86_64 CPU op-mode(s)...

RAM usage is now up to 23-24% (175-176gb) on each node after a little over 2 days of uptime, which is an increase of >10% in a day. No change in usage and no running VMs or containers on 2 of the nodes, but still it increased.

This morning, RAM usage is on my cluster nodes is up to between 12.33-13% (93-94gb of 756gb used). Nothing else has changed, so still seems to be increasing daily. Will continue to observe and report.

I found a new kernel was released over the weekend and I have updated the nodes in my cluster. Initially, now after rebooting from the update, the nodes are running with about 1.16% of CPU or 8.8-9gb of RAM in use. We'll see how it goes over the next few days and I will check to see if and how...

I'm seeing the same issue. 3-node pve 8.2.4 cluster running latest kernel (6.8.8-3-pve) with Ceph 17.2.7 (SSD osds) and glusterFS (SATA drives in RAIDZ) shared storage on Dell r730xd servers. Only running one Windows VM with HA and 3 Ubuntu Linux containers for NAT/DHCP server. RAM usage for the...

I ran into the same problem as the OP. After rebooting the node from the next round of updates after following the Cloud-init doc, the node booted up with the hostname of the Cloud-init VM that I had created and was no longer showing as active in the cluster. It did, however, still use the root...

I was running into the "failed: exit code 133" when attempting to restore a VM after reconfiguring our ProxMox server. I think it was because the directory structure was slightly different. I overcame the error by adding the '-storage local' flag to my qmrestore command and the restore completed...