IPMI CPU Usage on 1950

Matt Saladna msaladna at apisnetworks.com
Thu Dec 14 14:48:14 CST 2006


Hi everyone,
	We are migrating from the Poweredge 1600 platform to 1950 with the
fifth generation RAC.  After setting up CentOS 4.4, installing OMSA 5.1,
building the 2.6.19 kernel with IPMI support, and configuring DRAC's
watchdog to check in 480 second intervals, CPU usage seems extremely
high on the kernel's IPMI daemon:

top - 15:37:31 up 1 day, 19:55,  4 users,  load average: 0.01, 0.04, 0.00
Tasks:  81 total,   2 running,  79 sleeping,   0 stopped,   0 zombie
Cpu(s):  0.0% us,  0.0% sy,  0.0% ni, 100.0% id,  0.0% wa,  0.0% hi,
0.0% si
Mem:   4049344k total,   525480k used,  3523864k free,    43748k buffers
Swap:  2104496k total,        0k used,  2104496k free,   361872k cached

  PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME+  COMMAND

 3557 root      34  19     0    0    0 R    0  0.0  65:14.90 kipmi0

  880 root      10  -5     0    0    0 S    0  0.0   0:05.06 md5_raid1

 3727 root      16   0 62908 3236 1668 S    0  0.1   0:04.24
dsm_sa_datamgr3

The same high CPU behavior appears in CentOS 4.4's 2.6.9 kernel with
OMSA's DKMS-built IPMI modules.  This server has been sitting idle since
last reboot.  Is this normal for IPMI?  It seems as if something is
wrong, but sadly I have never ventured down the IPMI path before.

uname output:
Linux echelon 2.6.19 #9 SMP Tue Dec 12 15:13:32 EST 2006 x86_64 x86_64
x86_64 GNU/Linux

lsmod output:
Module                  Size  Used by
ipmi_devintf            9872  2
ipmi_si                34360  1
ipmi_msghandler        33012  2 ipmi_devintf,ipmi_si
dell_rbu                7968  0
dm_mod                 50384  0

Thanks!
- Matt Saladna



More information about the Linux-PowerEdge mailing list