Hi list,
I upgraded to pfSense 2.4.1 (from 2.3) recently, and ever since the
upgrade, the machine has been crashing multiple times daily. Here's a list
of Dumptimes extracted from the bug reporting module:
Line 20: Dumptime: Thu Nov 16 19:59:59 2017
Line 36: Dumptime: Thu Nov 16 20:02:04 2017
Line 52: Dumptime: Sat Nov 18 19:11:50 2017
Line 68: Dumptime: Sat Nov 18 20:06:03 2017
Line 84: Dumptime: Sat Nov 18 20:08:33 2017
Line 100: Dumptime: Sat Nov 18 20:10:41 2017
Line 116: Dumptime: Sat Nov 18 20:12:57 2017
Line 132: Dumptime: Sat Nov 18 20:18:13 2017
Line 148: Dumptime: Sat Nov 18 20:20:51 2017
Line 164: Dumptime: Sat Nov 18 22:13:42 2017
Line 180: Dumptime: Sun Nov 19 00:01:08 2017
Line 196: Dumptime: Sun Nov 19 09:14:27 2017
Line 212: Dumptime: Thu Nov 16 20:04:19 2017
Line 228: Dumptime: Sun Nov 19 12:24:57 2017
Line 244: Dumptime: Sun Nov 19 19:00:57 2017
Line 260: Dumptime: Sun Nov 19 19:16:59 2017
Line 276: Dumptime: Sun Nov 19 19:19:04 2017
Line 292: Dumptime: Sun Nov 19 19:21:34 2017
Line 308: Dumptime: Sun Nov 19 19:31:04 2017
Line 324: Dumptime: Sun Nov 19 20:19:43 2017
Line 340: Dumptime: Sun Nov 19 21:19:33 2017
Line 356: Dumptime: Sun Nov 19 21:55:31 2017
Line 372: Dumptime: Mon Nov 20 18:52:17 2017
Line 388: Dumptime: Thu Nov 16 20:08:31 2017
Line 404: Dumptime: Mon Nov 20 18:54:32 2017
Line 420: Dumptime: Mon Nov 20 18:57:50 2017
As you can see, it seems to occur around 7 to 8pm (GMT+8) daily and
occurs in rapid succession. There are exceptions though (e.g. the crashes
that happened at midnight, 9am and 1pm above). Eventually it does stabilise
and things appear well for another 24 hours or so.
Looking through the dumps, the backtrace is always this:
db:0:kdb.enter.default> bt
Tracing pid 12 tid 100085 td 0xfffff80006613560
strlen() at strlen+0x1f/frame 0xfffffe02390bf070
kvprintf() at kvprintf+0x93c/frame 0xfffffe02390bf160
vlog() at vlog+0x9b/frame 0xfffffe02390bf240
log() at log+0x3f/frame 0xfffffe02390bf2a0
ip6_forward() at ip6_forward+0xc5/frame 0xfffffe02390bf3f0
pf_refragment6() at pf_refragment6+0x17a/frame 0xfffffe02390bf4b0
pf_test6() at pf_test6+0x2176/frame 0xfffffe02390bf710
pf_check6_out() at pf_check6_out+0x1d/frame 0xfffffe02390bf730
pfil_run_hooks() at pfil_run_hooks+0x7b/frame 0xfffffe02390bf7c0
bridge_pfil() at bridge_pfil+0x61b/frame 0xfffffe02390bf850
bridge_broadcast() at bridge_broadcast+0x249/frame 0xfffffe02390bf8c0
bridge_forward() at bridge_forward+0x23b/frame 0xfffffe02390bf920
bridge_input() at bridge_input+0x2bd/frame 0xfffffe02390bf990
ether_nh_input() at ether_nh_input+0x292/frame 0xfffffe02390bf9f0
netisr_dispatch_src() at netisr_dispatch_src+0xa0/frame 0xfffffe02390bfa40
ether_input() at ether_input+0x26/frame 0xfffffe02390bfa60
vmxnet3_rxq_eof() at vmxnet3_rxq_eof+0x811/frame 0xfffffe02390bfae0
vmxnet3_legacy_intr() at vmxnet3_legacy_intr+0xf1/frame 0xfffffe02390bfb20
intr_event_execute_handlers() at intr_event_execute_handlers+0xec/frame
0xfffffe02390bfb60
ithread_loop() at ithread_loop+0xd6/frame 0xfffffe02390bfbb0
fork_exit() at fork_exit+0x85/frame 0xfffffe02390bfbf0
fork_trampoline() at fork_trampoline+0xe/frame 0xfffffe02390bfbf0
While it takes less than a minute for pfSense to reboot, the frequency
of it happening is starting to become a big nuisance.
Anyone has any idea what's going on? Restoring to pfSense 2.3 seems to
solve this problem, so it is more likely a software than hardware issue.
Warm regards,
Liwei
--
Clear Skies,LiweiCo-Founder, CTO
TinyMOS
<http://tinymos.com/> <https://www.facebook.com/thetinymos/>
<https://www.instagram.com/thetinymos/> <https://twitter.com/thetinymos>
21 Heng Mui Keng Terrace, Level 1 The Hangar, Singapore 119613
_______________________________________________
pfSense mailing list
https://lists.pfsense.org/mailman/listinfo/list
Support the project with Gold! https://pfsense.org/gold
