Scheduling problem alarm

ExternalUser · October 15, 2024, 8:06pm

We have this constant error: Scheduling problem alarm = true.

SBC’s packet processing cores have detected scheduling problems in the last 60 seconds.

Please advice.

admin · October 15, 2024, 8:06pm

Please see below the troubleshooting guide that having a scheduling problem topic that can apply to your case.

FreeSBC Troubleshooting Guide - TBwiki (telcobridges.com)

Also, scheduling problems once in a while should not be a concern yet.

dcosta · February 5, 2026, 5:38pm

Hello,

I frequently see the same alarm “Scheduling problem sbc list” and “Warning sbc list” on my PROSBC. This happens with no load, zero calls on ProSBC.
I believe I have set all tuning required, moreover I have the ProSBC on a Dell R640 with 2x“Intel(R) Xeon(R) Gold 6130 CPU @ 2.10GHz” running a proxmox 9.1.4 and at the moment it doesn’t have any other VMs running.

What am I missing?

In annex you can find my alarms and VM configuration on Proxmox.

Thanks,

David Costa

DavePuckett · February 6, 2026, 7:32am

Hi David,

Can you download the TBreport and for us?

Upload it here: https://tbsupport.wetransfer.com *(File Upload)
*
-Support

dcosta · February 6, 2026, 11:48am

Hi Dave,

I already upload the TBReport.

Thanks,

David Costa

DavePuckett · February 9, 2026, 6:04am

Hi David,

I don’t see any scheduling errors in the tbreport uploaded on feb 7th. this is a bit strange, the tbreport seems a bit light to me. Like a new build, or it’s only capturing a small timeframe.

Scheduling problem alarms normaly point to not having cpu resources for calling the next instruction. Can you elaborate more about your proxmox setup and tell us if it might be you are sharing resources on the same system with other applciations besides the ProSBC?

If you have the exact error, please send it here. Thanks!

-Support

dcosta · February 9, 2026, 7:24pm

Hi Dave,

The TBreport is lite because I’m still testing ProSBC and it yet doesn’t have activity, only my test calls.
The alarms also appear with no activity on ProSBC, zero calls on ProSBC.

The Proxmox cluster is made of 3 Proxmox servers, is also new and doesn’t have load. To better debug I moved all the VMs to other servers on the cluster and the Proxmox server where I have the ProSBC there’s no other VMs running.

Where can I see/found the exact error?

On webportal I see the following errors:

Thanks,

David Costa

DavePuckett · February 10, 2026, 5:04am

Hi David,

It’s difficult for me to see this “edit processors” image. Can you confirm if the NUMA box is checked on your proxmox settings? If enabled please disable. Then let me know if you still are encountering same errors.

Thanks,
Support

Also you can send me output so we can confirm. cat /etc/pve/qemu-server/VM-ID

Your NUMA should be set to 0.

dcosta · February 10, 2026, 11:41am

Hi Dave,

NUMA was already set to 0.

We have on CPU per ProSBC interface, plus one CPU.

You can confirm in the annex files. I post the /etc/pve/qemu-server/400.conf and the same data with export qm config 400 > vm400.conf

vm.zip (1.5 KB)

You can find below some hardware specifications of our Proxmox servers:

PowerEdge R640
CPUs:
2x Intel(R) Xeon(R) Gold 6130 CPU @ 2.10GHz
PROXMOX Interfaces:

Component FW Version
Intel(R) Ethernet 10G 2P X520 Adapter - A0:36:9F:26:8F:8C 15.0.28
Intel(R) Ethernet 10G 2P X520 Adapter - A0:36:9F:26:8F:8E 15.0.28

Component FW Version
Intel(R) Ethernet 10G 4P X550/I350 rNDC - E4:43:4B:29:A2:50 24.0.5
Intel(R) Ethernet 10G 4P X550/I350 rNDC - E4:43:4B:29:A2:51 24.0.5
Intel(R) Gigabit 4P X550/I350 rNDC - E4:43:4B:29:A2:52 24.0.5
Intel(R) Gigabit 4P X550/I350 rNDC - E4:43:4B:29:A2:53 24.0.5

Thanks,

David Costa

DavePuckett · February 11, 2026, 11:40am

Hey David,

Let’s try something. Can you ssh into your ProSBC?

lsmod | grep balloon

If you get a return that indicates virtio_balloon this indicates to me you are sharing ram with the host. Go into the proxmox settings and set ballooning off. Then restart the sbc and tell me if you have any schedule errors.

btw: How much does RAM does the host machine have?

Cheers,
Support

dcosta · February 11, 2026, 1:06pm

Hey Dave,

The host machine has 384GB.

You were right, we had ballooning active.

[root@prosbc01 ~]# lsmod | grep balloon
virtio_balloon 18015 0
virtio_ring 22991 5 virtio_net,virtio_pci,virtio_balloon,virtio_console,virtio_scsi
virtio 14959 5 virtio_net,virtio_pci,virtio_balloon,virtio_console,virtio_scsi