dmp_errd_loop

[Paul Thomson_2]

Hi

I was wondering if anyone had ever come accross this before ?

I have noted on three redhat servers the same behaviour


Red Hat Enterprise Linux AS release 3 (Taroon Update 3)
Linux version 2.4.21-20.ELsmp (bhcompile@tweety.build.redhat.com) (gcc version 3
.2.3 20030502 (Red Hat Linux 3.2.3-42)) #1 SMP Wed Aug 18 20:46:40 EDT 2004

I see two processes use high CPU usage for a number of hours then these processes return to a normal usage.

They are dmp_errd_loop and kswapd

root 435 13.0 0.0 0 0 ? SW May03 18591:43 [dmp_errd_loop]
root 7 6.2 0.0 0 0 ? SW May03 8907:44 [kswapd]


These processes use high CPU usage for a number of hours before returing to normal.

19:10:00 all 0.78 0.08 0.93 98.22
19:20:00 all 0.47 0.06 0.84 98.63
19:30:00 all 0.40 0.07 0.69 98.83
19:40:00 all 0.61 0.07 21.71 77.61
19:50:02 all 0.56 0.09 80.97 18.38
20:00:00 all 0.64 0.12 81.03 18.22
20:10:00 all 0.71 0.10 81.32 17.86

Cut parts to save space and next the following morning

03:30:00 all 0.69 0.11 81.04 18.16
03:40:00 all 0.56 0.10 81.22 18.11
03:50:01 all 0.55 0.10 80.85 18.50
04:00:00 all 0.58 0.11 81.14 18.17
04:10:00 all 1.41 0.26 80.43 17.91
04:20:00 all 0.54 0.11 81.13 18.22
04:30:00 all 0.52 0.13 80.98 18.37
04:40:00 all 0.58 0.10 81.00 18.32
04:50:00 all 0.58 0.10 81.28 18.04
05:00:00 all 1.47 0.11 80.61 17.81
05:10:00 all 0.57 0.05 4.57 94.81
05:20:00 all 0.46 0.06 0.78 98.70
05:30:00 all 0.33 0.08 0.72 98.87
05:40:00 all 0.95 0.07 3.32 95.65
05:50:01 all 0.45 0.06 0.83 98.67

I dont see any pgouts or swap usage increase during this time, so I dont believe kswapd is running high due to paging, and the dmp_errd_loop was to do with dynamic multi pathing, yet the only disks are internal RAIDED CCISS disks for a DL380 Proliant server.

Has anyone ever seen this behaviour ?

Thanks