May 1 17:49:01 cn036 rsyslogd: [origin software="rsyslogd" swVersion="7.4.7" x-pid="1025" x-info="http://www.rsyslog.com"] start
May 1 17:49:01 cn036 rsyslogd-2307: warning: ~ action is deprecated, consider using the 'stop' statement instead [try http://www.rsyslog.com/e/2307 ]
May 1 17:49:01 cn036 rsyslogd-2027: imjournal: fscanf on state file `/var/lib/rsyslog/imjournal.state' failed
[try http://www.rsyslog.com/e/2027 ]
May 2 16:39:29 cn036 rsyslogd: [origin software="rsyslogd" swVersion="7.4.7" x-pid="1148" x-info="http://www.rsyslog.com"] start
May 2 16:39:29 cn036 rsyslogd-2307: warning: ~ action is deprecated, consider using the 'stop' statement instead [try http://www.rsyslog.com/e/2307 ]
May 2 16:39:29 cn036 rsyslogd-2027: imjournal: fscanf on state file `/var/lib/rsyslog/imjournal.state' failed
[try http://www.rsyslog.com/e/2027 ]

Hi...!!! I can't find any solution for this compute node restarts in /var/log/messages file. Please tell me how to find the answer for automatic node restart. There is no problem in Centos 7.2.1511 . But it gets restart every 7 hrs a day.

Thanks in advance..

Since you tell us absolutely nothing about your environment, aside from the fact you're using CentOS 7, what do you think we'll be able to tell you? A few lines from a log file (that have nothing to do with the system restarting, and appear to be just warnings), give us nothing to go on. You mention a 'compute node', but tell us nothing about the cluster.

Checked the CRON jobs? Are you *SURE* it's exactly every 7 hours? Are there any messages in the syslogs on the specific node around that time that have ANYTHING to do with a restart/dump/some-fatal-error??

Hello Boopathi Ram,

Welcome!

It doesn't look like a problem. Did you check if rsyslog is running or not. A lot of things hup rsyslog. Unless it never starts up again, it's OK.

Thanks #AwesomeMachine... But rsyslog is running fine, when i gave "ps -A|grep rsyslog" it return some pid on that server node. If the server is idle, its not restarting automatically, but whenever some calculations are running in that server node, it does restarts by an average of 10- 12Hrs a day. I'm using "3.10.0-327.el7.x86_64" this kernel version and centos 7.2.1511. But remaining nodes are working properly with the same setup.

[hpc3@cn036 ~]$ last reboot
reboot system boot 3.10.0-327.el7.x Fri May 4 06:33 - 13:04 (06:30)
reboot system boot 3.10.0-327.el7.x Fri May 4 00:26 - 13:04 (12:38)
reboot system boot 3.10.0-327.el7.x Wed May 2 16:39 - 13:04 (1+20:24)
reboot system boot 3.10.0-327.el7.x Tue May 1 17:48 - 13:04 (2+19:15)
reboot system boot 3.10.0-327.el7.x Tue May 1 13:11 - 13:04 (2+23:52)
reboot system boot 3.10.0-327.el7.x Tue May 1 08:06 - 13:04 (3+04:57)
reboot system boot 3.10.0-327.el7.x Mon Apr 30 20:22 - 13:04 (3+16:41)

wtmp begins Mon Apr 30 03:48:32 2018

This is what happening in this node. I can't find any answer for why it's restarting frequently!!! Please help me to find the solution. Thanks in advance

...and....
And you're welcome for my advice too. Again, as you were asked by me in my first reply:

• You mention a 'compute node', but tell us nothing about the cluster.
• Have you checked the CRON jobs?
• Are you *SURE* it's exactly every 7 hours?
• Are there any messages in the syslogs on the specific node around that time that have ANYTHING to do with a restart/dump/some-fatal-error??

You have answered none of these questions, and have still not posted ANY information that looks like an error. AGAIN, the questions asked are starting points. Showing a "last reboot" again, tells us absolutely nothing.

Find out the **EXACT TIME** the system reboots, and look at the log files from a few minutes before, to a few minutes after. Check the rest of the things mentioned. Do some actual troubleshooting.

:[ 512.659296] mce: [Hardware Error]: CPU 6: Machine Check Exception: 5 Bank 3: b200000000300165
:[ 512.659300] mce: [Hardware Error]: Machine check events logged
:[ 512.659360] mce: [Hardware Error]: RIP !INEXACT! 33:<000000000118d9aa>
:[ 512.659383] mce: [Hardware Error]: TSC 1741938e3cd
:[ 512.659402] mce: [Hardware Error]: PROCESSOR 0:406f1 TIME 1517247164 SOCKET 0 APIC 1a microcode b000012
:[ 512.659432] mce: [Hardware Error]: Run the above through 'mcelog --ascii'
:[ 512.664493] [Firmware Warn]: ERST: Firmware does not respond in time
:[ 512.664515] mce: [Hardware Error]: Machine check: Processor context corrupt
:[ 512.664537] Kernel panic - not syncing: Fatal machine check


I got this from "/var/spool/mail/root".

There is no such log in CRON jobs.

Regarding compute node.
"Intel(R) Xeon(R) CPU E5-2680 v4 @ 2.40GHz"
"Core = 28"
"PBS Pro 14.1.0"

Ok...did you not read or understand what you posted??? Did you try to look up what the MCE error is??
http://www.advancedclustering.com/ac...ptions-or-mce/

Your machine is having a hardware problem. Fix it.

1 members found this post helpful.