ZFS partition unstable on ProxMox
Hello, for a very long time I face the following problem I have many node with ZFS partitions on raid 0/5/10 and all well have the same problem, the slightest data copy is done a lot I / O and all node becomes unstable, how can I solve this? I want to say that I have enough RAM, for example I have a node with 192 GB ram and only 64 GB ram are used by VMs and I / O is about 30% but the node becomes unstable and even drops uneoiri I tested the same thing on file partitions and this did not happen, I want to say that for both tests I used ProxMox 7 Who else was facing this type of problem? I thought to pass all the knots on file partitions because is more efficient than some ZFS partitions that when you do test YABS have good results but when try copy just a simple database node sits to fall Or what solution to this or so does ZFS work?
Do you running any cache on this setup ?
Share your full disk layout
Now you just realized this when years ago it had the same problem. It even cost one of the hosts here their entire company.
Good to know proxmox hasn't solved this issue yet.
Hello , the most recently node it's
4x2 TB HDD SAS HGS
Hello , no , only ZFS Partitions , partitions it s defaul , how did he create proxmox at installation, I did not touch them