pfS+ 23.01-RELEASE crash on multiple PPPoE reconects
-
Now I hope this is the right place to submit a "crash report/programming bug".
It only happened once and it is was during my ISP outage on PPPoE connection. Additionally I must admit i was somewhat impatiently and/or frantically click manual reconnect and after some 30ish failed attempts it restarted.
So do i post my crash dump files here or mail directly to netgate or something else?
Thanks in advance.
-
You can post it here. Usually we would only need to see the backtrace and panic string at least initially. Any output leading up to it in the message buffer output can also help.
Steve
-
Thanks Steve, here's files...
-
Hmm, I'm unable to open that file for some reason. Did you modify it at all?
-
just sanitized some private urls, but i can download them just fine and open
want me to copy/paste to clear text?
-
Weird I can't open it in anything. Yes if you post the contained files directly I can review them.
-
it won't let me post entire thing because it's too long so here's another try...
info.0.txt
textdump.tar.0.txt -
Ah it looks like it was just a text file I was failing to decompress.
Ok so the backtrace there looks like:
db:1:pfs> bt Tracing pid 0 tid 100007 td 0xfffffe00119bd720 kdb_enter() at kdb_enter+0x32/frame 0xfffffe00101a86c0 vpanic() at vpanic+0x182/frame 0xfffffe00101a8710 panic() at panic+0x43/frame 0xfffffe00101a8770 trap_fatal() at trap_fatal+0x409/frame 0xfffffe00101a87d0 trap_pfault() at trap_pfault+0x4f/frame 0xfffffe00101a8830 calltrap() at calltrap+0x8/frame 0xfffffe00101a8830 --- trap 0xc, rip = 0xffffffff80f66369, rsp = 0xfffffe00101a8900, rbp = 0xfffffe00101a8930 --- pppoe_findsession() at pppoe_findsession+0x79/frame 0xfffffe00101a8930 ng_pppoe_rcvdata_ether() at ng_pppoe_rcvdata_ether+0x461/frame 0xfffffe00101a89b0 ng_apply_item() at ng_apply_item+0x2bf/frame 0xfffffe00101a8a40 ng_snd_item() at ng_snd_item+0x28e/frame 0xfffffe00101a8a80 ether_demux() at ether_demux+0x212/frame 0xfffffe00101a8ab0 ether_nh_input() at ether_nh_input+0x353/frame 0xfffffe00101a8b10 netisr_dispatch_src() at netisr_dispatch_src+0xb9/frame 0xfffffe00101a8b60 ether_input() at ether_input+0x69/frame 0xfffffe00101a8bc0 ether_demux() at ether_demux+0x9e/frame 0xfffffe00101a8bf0 ether_nh_input() at ether_nh_input+0x353/frame 0xfffffe00101a8c50 netisr_dispatch_src() at netisr_dispatch_src+0xb9/frame 0xfffffe00101a8ca0 ether_input() at ether_input+0x69/frame 0xfffffe00101a8d00 iflib_rxeof() at iflib_rxeof+0xbdb/frame 0xfffffe00101a8e00 _task_fn_rx() at _task_fn_rx+0x72/frame 0xfffffe00101a8e40 gtaskqueue_run_locked() at gtaskqueue_run_locked+0x15d/frame 0xfffffe00101a8ec0 gtaskqueue_thread_loop() at gtaskqueue_thread_loop+0xc3/frame 0xfffffe00101a8ef0 fork_exit() at fork_exit+0x7e/frame 0xfffffe00101a8f30 fork_trampoline() at fork_trampoline+0xe/frame 0xfffffe00101a8f30 --- trap 0x107a772c, rip = 0x11d295625b1b31a1, rsp = 0xf02460003c98dfb3, rbp = 0x41abfa0065646f ---We also see in the message buffers:
<5>ng_pppoe[1d]: no matching session <5>ng_pppoe[1d]: no matching sessionNeither of those are errors I'm familiar with.
Do you have any traffic shaping in play? The output form Snort has obscured the boot output in the message buffer.
Steve
-
yep, a guest wifi network limiter
-
Just a limiter though? No AltQ shaping? And not on WAN?
-
right, no shaping on nothing, only limiter 100/50 on wifi subnet
however, i can offload that to unifi if that's the problem
-
Do you have another PPPoE session already running, pppoe0?
The only other reference to that crash I can see is this: https://forum.netgate.com/post/902217
That appeared to be a one off and this may also be if you were spamming the reconnect button. It's possible the manual connect killed the session while the script was running. I would not expect it to crash if/when that happens though.
-
@stephenw10 said in pfS+ 23.01-RELEASE crash on multiple PPPoE reconects:
Do you have another PPPoE session already running, pppoe0?
no, there's only pppoe1 which is tied on ISP internet vlan
The only other reference to that crash I can see is this: https://forum.netgate.com/post/902217
That appeared to be a one off and this may also be if you were spamming the reconnect button. It's possible the manual connect killed the session while the script was running. I would not expect it to crash if/when that happens though.
well, as I said, and a bit embarrassed, i was kinda frantic to get back online, and it happened as i was doing it, and never before or after, so occam's razor...
now since you said you wouldn't expect it to crash/reboot, would it be beneficial to netgate/pfsense devs if i try to reproduce it?
although i'm sure it wont be exactly the same scenario as true ISP outagep.s. do you by any chance want a ppp log?
-
I wouldn't expect the ppp log to show anything further.
If you can replicate it easily it would be interesting to have two crash reports.
Steve
