Reboot no explanation

Ricardo_Gros · ‎2019-04-12

Hi,

I recently have the issue that a customer from us has an appliance that reboots without explanation.

We have noticed following behaviour:

[Expert@clusterFW2:0]# last -x |head |tac
reboot   system boot 2.6.18-92cpx86_6 Fri Apr 12 02:44          (00:03)
runlevel (to lvl 3)   2.6.18-92cpx86_6 Fri Apr 12 02:44 - 02:48 (00:03)
runlevel (to lvl 6)   2.6.18-92cpx86_6 Fri Apr 12 02:48 - 02:48 (00:00)
shutdown system down 2.6.18-92cpx86_6 Fri Apr 12 02:48 - 15:59 (13:10)
reboot   system boot 2.6.18-92cpx86_6 Fri Apr 12 02:51          (13:07)
runlevel (to lvl 3)   2.6.18-92cpx86_6 Fri Apr 12 02:51 - 15:59 (13:07)
sseidewi pts/2        dez7acomdv010.in Fri Apr 12 06:05 - 06:25 (00:20)
admin    pts/2        dez7acomdv002.in Fri Apr 12 09:29 - 09:41 (00:12)
admin    pts/2        dez7acomdv001.in Fri Apr 12 14:23 - 14:57 (00:33)
admin    pts/2        dez7acomdv001.in Fri Apr 12 15:46   still logged in

This looks like a normal reboot, however runlevel 6 is making me wonder, a normal reboot should not show runlevel6,

On messages file I can see the message Restart, but no errors previous to this, system reboots normally.

There are no crash dumps available or errors.

Can I somehow confirm that the system was not rebooted by simply pressing the power or imputing a command?

Nick_Doropoulos · ‎2019-04-12

Try the dmesg log file:

less /var/log/dmesg

Ricardo_Gros · ‎2019-04-16

Hi DSMEG shows nothing relevant. Any other ideas?

I also found

Apr 12 02:43:06 2019 clusterFW2 kernel:
Apr 12 02:43:06 2019 clusterFW2 kernel: FW-1: stopping debug messages for the next 54 seconds
Apr 12 02:57:39 2019 clusterFW2 syslogd 1.4.1: restart.
Apr 12 02:57:39 2019 clusterFW2 syslogd: local sendto: Network is unreachable

Nick_Doropoulos · ‎2019-04-16

Another log I would check is the /var/log/messages file. Could you provide the entries whose time stamp corresponds with the last time that the reboot took place?

Ricardo_Gros · ‎2019-04-16

Apr 12 02:43:06 2019 clusterFW2 kernel: [fw4_5];FW-1 - cifs_process_read_andx: /var/log/jail/sys/class/misc/mcelog
Apr 12 02:43:06 2019 clusterFW2 last message repeated 98 times
Apr 12 02:43:06 2019 clusterFW2 kernel:
Apr 12 02:43:06 2019 clusterFW2 kernel: FW-1: stopping debug messages for the next 54 seconds
Apr 12 02:57:39 2019 clusterFW2 syslogd 1.4.1: restart.
Apr 12 02:57:39 2019 clusterFW2 syslogd: local sendto: Network is unreachable
Apr 12 02:57:39 2019 clusterFW2 kernel: klogd 1.4.1, log source = /proc/kmsg started.
Apr 12 02:57:39 2019 clusterFW2 kernel: Linux version 2.6.18-92cpx86_64 (builder@Lnx50BccCmp9.checkpoint.com) (gcc version 4.1.1 20061011 (Red Hat 4.1.1-30)) #1 SMP Mon Oct 8 10:34:42 IDT 2018

Reboot happend on this time frame.

G_W_Albrecht · ‎2019-04-16

I would rather involve TAC here - and be sure to save all log files before they are overwritten !

Ricardo_Gros · ‎2019-04-16

Hi Gunter,

Tac is involved, we already had 1 RMA, now the new maschine has the exact same problem, this lead me to reasearch because this is very strange...

I'm searching for mcelog or something similar but could not find it..

Vladimir · ‎2019-05-13

@Ricardo_Gros , can you let us know if the problem was identified and resolved?

If so, what was causing the reboots and how was it remedied.

Thank you,

Vladimir

Ricardo_Gros · ‎2019-09-13

Hi,

Sorry for the late reply, this issue was a Kernel Bug of some sort and was solved by installing at the time the newest Jumbo Take.

Nick_Doropoulos · ‎2019-04-16

Could you provide the following info as well please:

fw ver
enabled_blades

JozkoMrkvicka · ‎2019-04-16

Are syslog messages forwarded to some syslog receiver ? From there you should see what happened exactly.

Do you have LOM ? There you can find some interesting logs.

What is status of PSU ?

Wasnt there any on-site activity for this node? Manual on-site reboot, for example.

Kind regards,
Jozko Mrkvicka

Alissa20 · ‎2019-04-17

Agreed.....involving TAC will make things much more easier...Also ensure that u dont overwrite log files...So yeah,involve TAC files and it make things much more easier.