If one virtual machine on the server is overloaded, the other virtual machines stop responding

[Musa123]

I have 3 servers, any virtual machine, even the most powerful one, stops responding under weak load. If I slightly overload a virtual machine with the command ab -T 'application/x-www-form-urlencoded' -l -r -n 10000 -c 100 -k, the virtual machine will overload and if we overload a virtual machine on pve02, the other virtual machines on pve02 and pve03 also stop working, this is also true for the reverse operation, overloading a virtual machine on pve03 also causes all virtual machines on pve02 and pve03 to freeze. In addition, pve02 shows a status of "No route to host (595)". The virtuals don't shut down, they just hang for a while, then after about 3-4 minutes they start working again. Help, maybe someone has encountered this problem.

 

[shrdlicka]

Hi,

do you have more details about you setup. For example:

• What storage are you using? Does it use spinning drives or SSDs?
• Config of one of the VMs (run qm config <vmid>[/icode} [*]Can you provide the output of [icode]pveversion -v

 

[Musa123]

Hi,

do you have more details about you setup. For example:

• What storage are you using? Does it use spinning drives or SSDs?
• Config of one of the VMs (run qm config <vmid>[/icode} [*]Can you provide the output of [icode]pveversion -v

Hi, thanks for trying to help. The SSDs used are for ceph.
I realized that the problem was with the ceph, one of its disks that is on pve03 was overloaded 91% and it was causing crashes, I tried to free up memory but when I removed the extra virtual machine, the memory did not change. Eventually the proxmox completely froze. I decided to add another SSD drive on pve03 to the one I already had. I partitioned the new drive and added it to ceph, but the problem is that the two drives that are on pve03 keep going into down status, even though the memory was supposed to be distributed evenly among the 4 SSDs.
UPD: The keys were prescribed

 

[shrdlicka]

More information about your setup would be helpful. Can you post the output of the following commands?

ceph osd df tree

ceph -s

cat /etc/ceph/ceph.conf
cat /etc/network/interfaces
pveceph pool ls --noborder > pools.txt

ls -l /dev/disk/by-id