1

On my RHEL7 VM, the kernel process migration/[some_number] cpu usage is 100% and continuously getting cpu soft lockup messages on the terminal. 

PID USER      PR  NI    VIRT    RES    SHR S  %CPU %MEM     TIME+ COMMAND
37 root      rt   0       0      0      0 R 100.0  0.0  10:26.94 migration/1
63 root      rt   0       0      0      0 R 100.0  0.0  10:16.72 migration/6

kernel:[2130030.151956] BUG: soft lockup - CPU#1 stuck for 23s! [migration/1:37]
kernel:[2130030.187921] BUG: soft lockup - CPU#4 stuck for 23s! [migration/4:53]

The system freezes after this and I've to hard reboot it. It happens at least twice in a month. I did some research on this problem and found this is a bug in kernels before 3.6.11 and the workaround is to execute the command # echo 0 > /proc/sys/kernel/sched_cpulimit_nr_balance. But my VM kernel version is much higher than this,

[root@jboss-iotapp-qa ~]# uname -a
Linux jboss-iotapp-qa 3.10.0-229.1.2.el7.x86_64 #1 SMP Fri Mar 6 17:12:08 EST 2015 x86_64 x86_64 x86_64 GNU/Linux

[root@jboss-iotapp-qa ~]# cat /etc/redhat-release
Red Hat Enterprise Linux Server release 7.2 (Maipo)

I could not find the sched_cpulimit_nr_balance file,

[root@jboss-iotapp-qa ~]# ls -l /proc/sys/kernel/sched_cpulimit_nr_balance
ls: cannot access /proc/sys/kernel/sched_cpulimit_nr_balance: No such file or directory

Any suggestions would be appreciated.

Gerald Schneider
  • 23,274
  • 8
  • 57
  • 89
George John
  • 11
  • 1
  • Does this answer your question? [Kernel Processes Periodically Eating CPU During High Load](https://serverfault.com/questions/674685/kernel-processes-periodically-eating-cpu-during-high-load) – digijay May 07 '20 at 12:40
  • I'm planning to upgrade it to latest rhel version. Let's see that fix the problem. – George John May 07 '20 at 16:39
  • Did you try the recommendation from the link above (setting swappiness to 1)? – digijay May 08 '20 at 14:01

0 Answers0