1

On a shared web server I have a strange readings in /proc/loadavg:

# cat /proc/loadavg
2145237816.60 2145245358.02 2145270263.79 10/1409 410568

The load is in fact normal and server is very responsive. It is a 8 core 16GB server.

# uname -a
Linux XXXX 2.6.18-338.19.1.el5.lve0.8.36 #1 SMP Wed Jul 27 13:20:06 EEST 2011 x86_64 x86_64 x86_64 GNU/Linux
# vmstat
procs -----------memory---------- ---swap-- -----io---- --system-- -----cpu------
 r  b   swpd   free   buff  cache   si   so    bi    bo   in   cs us sy id wa st
 5  0 383520 533624 837152 10716692    0    1   280   240    0    0 24 14 55  7  0

Part of top output is below:

top - 12:10:22 up 83 days, 12:32,  2 users,  load average: 2145210247.30, 2145219511.25, 2145245934.60
Tasks: 1257 total,   2 running, 1249 sleeping,   3 stopped,   3 zombie
Cpu(s): 16.1%us,  7.1%sy,  0.0%ni, 76.5%id,  0.2%wa,  0.0%hi,  0.2%si,  0.0%st
Mem:  16403852k total, 15910568k used,   493284k free,   853720k buffers
Swap: 33543712k total,   383520k used, 33160192k free, 10780112k cached
masegaloeh
  • 17,978
  • 9
  • 56
  • 104
FelixEnescu
  • 197
  • 1
  • 10
  • That seems like a lot. Have you checked for processes in uninterruptible sleep mode? (ps aux and look under "STAT" for processes with a D) – Frands Hansen Aug 18 '12 at 10:16
  • Nope. No uninterruptible sleep processes: `#ps -aux | tail -n +1 | awk '{ print $8}' |sort | uniq -c | sort -g Warning: bad syntax, perhaps a bogus '-'? See /usr/share/doc/procps-3.2.7/FAQ 1 R+ 1 S – FelixEnescu Aug 18 '12 at 11:22
  • This is Cloud Linux kernel for Centos 5, you could report it there, as well what hardware are you using. – Andrew Smith Aug 18 '12 at 16:17

1 Answers1

3

It was a known CloudLinux kernel bug. After updating the kernel and rebooting, everything went back to normal.

FelixEnescu
  • 197
  • 1
  • 10