after migration to proxmox , error shows in guest OS

[Ayush]

Hi Team ,
We have 3 node cluster , we recently migrate a vm from our KVM based environment to proxmox. But after migration to proxmox we face issue with guest vm. Our guest vm gives following errors .

x_tables autofs4 hid_generic usbhid hid psmouse virtio_net virtio_scsi i2c_piix4 pata_acpi floppy
[Fri Mar 22 14:54:57 2024] CPU: 2 PID: 55160 Comm: BcastRecv Tainted: G L 4.15.0-213-generic #224-Ubuntu
[Fri Mar 22 14:54:57 2024] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.16.2-0-gea1b7a073390-prebuilt.qemu.org 04/01/2014
[Fri Mar 22 14:54:57 2024] RIP: 0010:rw_verify_area+0x2a/0xb0
[Fri Mar 22 14:54:57 2024] RSP: 0018:ffffa5ddcc803e98 EFLAGS: 00010206 ORIG_RAX: ffffffffffffff11
[Fri Mar 22 14:54:57 2024] RAX: 00000000ffffffea RBX: 0000000000000d9a RCX: 0000000000000033
[Fri Mar 22 14:54:57 2024] RDX: 0000000000000d67 RSI: ffff8b1cf9057100 RDI: 0000000000000001
[Fri Mar 22 14:54:57 2024] RBP: ffffa5ddcc803ea8 R08: 00007ffffffff000 R09: ffff8b1adfa41a98
[Fri Mar 22 14:54:57 2024] R10: 0000000000000000 R11: 0000000000000000 R12: fffffffffffffff2
[Fri Mar 22 14:54:57 2024] R13: ffffa5ddcc803ef8 R14: 0000563ddb4ffb30 R15: ffff8b1cf9057100
[Fri Mar 22 14:54:57 2024] FS: 00007efd01cc4700(0000) GS:ffff8b18cdc80000(0000) knlGS:0000000000000000
[Fri Mar 22 14:54:57 2024] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[Fri Mar 22 14:54:57 2024] CR2: 000056208f700092 CR3: 00000007e5f5c000 CR4: 00000000000006e0
[Fri Mar 22 14:54:57 2024] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[Fri Mar 22 14:54:57 2024] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
[Fri Mar 22 14:54:57 2024] Call Trace:
[Fri Mar 22 14:54:57 2024] vfs_write+0x73/0x1a0
[Fri Mar 22 14:54:57 2024] SyS_write+0x5c/0xe0
[Fri Mar 22 14:54:57 2024] do_syscall_64+0x73/0x130

additionally we observe that vm high cpu utilisation intermittently . What is the suggestion that make it work it in a smooth way.

 

[Ayush]

We observe same error for our each application

at Mar 23 16:21:43 2024] CPU: 13 PID: 3901 Comm: python3 Not tainted 4.15.0-213-generic #224-Ubuntu
[Sat Mar 23 16:21:43 2024] CPU: 16 PID: 3333 Comm: -basket-backe Tainted: G L 4.15.0-213-generic #224-Ubuntu
[Sat Mar 23 16:21:43 2024] CPU: 18 PID: 3810 Comm: node Tainted: G L 4.15.0-213-generic #224-Ubuntu
[Sat Mar 23 16:21:43 2024] CPU: 19 PID: 3579 Comm:-basket-back Tainted: G L 4.15.0-213-generic #224-Ubuntu
[Sat Mar 23 16:21:43 2024] CPU: 21 PID: 3639 Comm: v2-baske Tainted: G L 4.15.0-213-generic #224-Ubuntu
[Sat Mar 23 16:21:43 2024] CPU: 30 PID: 2161 Comm: nats-server Tainted: G L 4.15.0-213-generic #224-Ubuntu
[Sat Mar 23 16:21:43 2024] CPU: 31 PID: 3056 Comm: python3.6 Tainted: G L 4.15.0-213-generic #224-Ubuntu

This type of issue didn't observe in kvm based virtualization

 

[pille99]

you added 60 cores to the VM from the first Picture

 

[pille99]

i am not very familar with KVM itself. Proxmox is based on KVM. so i think - the VM is "compatible" on the new environment.
it looks to me it has some issues with the VM Hardware you added.

 

[Ayush]

you added 60 cores to the VM from the first Picture

Yes , initially it has 40 core and we face interface reset and high cpu utilization. So we upgrade its core from 40 to 60. Base box have 40 Core and with hyperthreading it can go to 80 cores.
Additionally no other vm is running on that node.

 

[pille99]

what is running there that you need 40 cores ?
may you should check the high cpu utilization instead of add more CPU.
i have seen database server with 40 CPU, which has an extensive load
i have my doubt you have such environment. App server run, and never seen more, mostly on 2 or 4 CPU, very rare 8 or more.

 

[Ayush]

what is running there that you need 40 cores ?
may you should check the high cpu utilization instead of add more CPU.
i have seen database server with 40 CPU, which has an extensive load
i have my doubt you have such environment. App server run, and never seen more, mostly on 2 or 4 CPU, very rare 8 or more.

Yes , there database is running on it . But size of database is around 10Gb.

 

[Ayush]

Hello Team,

I make some change the type of CPU in guest machine. After that I did't see those messages.

My node CPUs is


Model name: Intel(R) Xeon(R) CPU E5-2696 v4 @ 2.20GHz
BIOS Model name: Intel(R) Xeon(R) CPU E5-2696 v4 @ 2.20GHz CPU @ 2.2GHz
BIOS CPU family: 179


and I did changes to
2 Socket -30 Cores x86-64-v3 , Numa=1

But I still face vm stuck issue. But the frequency is less. Any other way I can optimize it ?