[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Xen-users] strange behaviour of top



Hello,

I have a strange problem inside my Xen DomU.
Inside my domU, I have a tool running (lookbusy) that generates a certain amount of cpu load.
Let's say, I generate 50% of CPU load on 1 core.

When I check the cpu load with top, I get something strange.
Here's the output from top:

top - 11:36:06 up 52 min,  1 user,  load average: 0.54, 0.52, 0.46
Tasks:  68 total,   2 running,  66 sleeping,   0 stopped,   0 zombie
Cpu0 : 0.0%us, 0.0%sy, 0.0%ni, 99.8%id, 0.0%wa, 0.0%hi, 0.0%si, 0.2%st Cpu1 : 34.0%us, 0.2%sy, 0.0%ni, 65.8%id, 0.0%wa, 0.0%hi, 0.0%si, 0.0%st
Mem:    262468k total,    32896k used,   229572k free,        0k buffers
Swap:        0k total,        0k used,        0k free,    10092k cached

 PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME+  COMMAND
1384 root      20   0  1800  492  384 R   52  0.2  26:11.09 lookbusy


So, inside the detailed list, I see a CPU load of 52% (which is indeed more or less what I expect), but at the summary on the top of the output, I see the CPU1 is 65.8% idle or thus only 34.2% loaded.

If I look with xen top from Domain-0, I see more or less 50%.

I also tried some other tools:
1) with ps: I see the 52% load
2) with vmstat: I get the 34.2% load

Is there anyone who has an idea what could cause this difference?
Which tool should I believe?


Here is my configuration:
hypervisor: Xen 3.4.0
Domain 0:  2.6.26-2-xen-686 image from Lenny distribution
Domain U: minimal version of a 2.6.26-kernel for xen domU (lenny).
Hardware: 4 cores in total, in the example described above, I had 1 cpu core pinned for Domain-0 and 2 cpu cores pinned for the DomU.


Kind regards,
Lien



_______________________________________________
Xen-users mailing list
Xen-users@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-users


 


Rackspace

Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.