Pfsense rebooting too often

JKnott

@rafapoch Your hardware might be failing. I've never seen pfSense reboot on it's own, except when the computer I was running it on was about to croak!

stephenw10

This is a kernel panic, it didn't just reboot.

And it's running in KVM so unlikely to be a hardware issue. Though not impossible!

The backtrace is inconclusive unfortunately:

db:0:kdb.enter.default>  bt
Tracing pid 11 tid 100003 td 0xfffffe00021ec3a0
kdb_enter() at kdb_enter+0x32/frame 0xfffffe0048fc8d60
vpanic() at vpanic+0x163/frame 0xfffffe0048fc8e90
panic() at panic+0x43/frame 0xfffffe0048fc8ef0
trap_fatal() at trap_fatal+0x40c/frame 0xfffffe0048fc8f50
calltrap() at calltrap+0x8/frame 0xfffffe0048fc8f50
--- trap 0x9, rip = 0xffffffff8126b570, rsp = 0xfffffe0048fc9020, rbp = 0xfffffe0048fc9d70 ---
Xprot_pti() at Xprot_pti/frame 0xfffffe0048fc9d70
acpi_cpu_idle() at acpi_cpu_idle+0x2fe/frame 0xfffffe0048fc9db0
cpu_idle_acpi() at cpu_idle_acpi+0x46/frame 0xfffffe0048fc9dd0
cpu_idle() at cpu_idle+0x9d/frame 0xfffffe0048fc9df0
sched_idletd() at sched_idletd+0x576/frame 0xfffffe0048fc9ef0
fork_exit() at fork_exit+0x7f/frame 0xfffffe0048fc9f30
fork_trampoline() at fork_trampoline+0xe/frame 0xfffffe0048fc9f30
--- trap 0, rip = 0, rsp = 0, rbp = 0 ---

Have all the crashes produced similar backtraces?

rafapoch

My pfsense is running on VirtualBox, although there are more two VM's running in the same box, just on the Pfsense i'm getting problems.
So, I don't think the hardware is failuring.

Here's the latest crash report

msgbuf.txt ddb.txt config.txt panic.txt version.txt

stephenw10

Hmm, are you sure it's Virtual Box? Everything shown in the log says it isn't!

Anyway that crash is very different:

db:0:kdb.enter.default>  bt
Tracing pid 69265 tid 100341 td 0xfffffe005e6d8900
kdb_enter() at kdb_enter+0x32/frame 0xfffffe005c7b6d90
vpanic() at vpanic+0x163/frame 0xfffffe005c7b6ec0
panic() at panic+0x43/frame 0xfffffe005c7b6f20
trap_fatal() at trap_fatal+0x40c/frame 0xfffffe005c7b6f80
calltrap() at calltrap+0x8/frame 0xfffffe005c7b6f80
--- trap 0x9, rip = 0xffffffff81292d20, rsp = 0xfffffe005c7b7058, rbp = 0x820280650 ---
handle_ibrs_entry() at handle_ibrs_entry/frame 0x820280650

What hardware is the hypervisor running on? This seems similar the known issue with Jasper Lake CPUs and power management.

rafapoch

Yes, I'm sure its a virtual enviroment, it's a virtual machine running from a Proxmox Server.
The hardware is a Beelink Mini Pc.

As I said, i have two more vm's running without any problem ( all linux )

stephenw10

Ok Proxmox is not VirtualBox. But that also lines up with that know issue.

What version of Proxmox is it? What hardware is in that Beelink device?

rafapoch

Sorry, I wrote wrong in the first post, it's a ProxmoxServer, not VirtuaBox.

The hardware is:
Intel(R) Celeron(R) N5095 @ 2.00GHz
Beelink MINI S12 Pro
8G RAM

PFsense version is 2.7.2-RELEASE (amd64)

stephenw10

Ok that is a Jasper Lake CPU so you're probably hitting that issue in Proxmox.

Are you running a current version of Proxmox? With all the patches?

Otherwise check: https://forums.servethehome.com/index.php?threads/jasper-lake-proxmox-kvm-qemu-vm-guest-stability.38824/

rafapoch

I'm running the Proxmox Edge Version 6.0.19 with the latest updates available.

I've tried to update the microde, even apt install intel-microcode running successfuly the version from server stills showing a differnt version. ( reboot server was done )

stephenw10

Ok 6.0.19 is pretty old (2023?) That's still in Proxmox 7 I assume. You should upgrade that IMO.

This is almost certainly that bug though which is not a pfSense issue.