From: Gowda, Sanjay (Sanjay.Gowda@wiltel.com)
Date: Sun Nov 30 2003 - 17:48:50 EST
Hi,
This is on a V880 part of a 2-node Veritas cluster. I have a couple of
Oracle databases also running on the box. I am seeing an unusally
high Kernel CPU on the box. It is a 6 CPU box with 12 Gigs of RAM.
Here are some of the outputs on the box.
# top output:
load averages: 15.02, 14.27, 12.77 hostname 16:48:22
160 processes: 139 sleeping, 15 running, 1 stopped, 5 on cpu
CPU states: 2.0% idle, 9.5% user, 88.3% kernel, 0.3% iowait, 0.0% swap
Memory: 12.0G real, 6.8G free, 4.0G swap in use, 10.9G swap free
PID USERNAME THR PR NCE SIZE RES STATE TIME FLTS CPU COMMAND
23897 oraclecl 1 0 0 2.1G 34.3M run 10:17 0 5.40% oracle
27839 oraclecl 1 0 0 2.1G 28.6M run 52:07 0 4.97% oracle
24208 oraclecl 1 10 0 2.1G 34.3M sleep 9:45 0 4.86% oracle
27797 oraclecl 1 0 0 2.1G 29.6M run 53:47 0 4.85% oracle
24110 oraclecl 1 60 0 2.1G 34.1M cpu01 7:43 0 4.85% oracle
23877 oraclecl 1 30 0 2.1G 34.1M run 9:55 0 4.82% oracle
27783 oraclecl 1 0 0 2.1G 29.7M run 54:09 0 4.60% oracle
26215 oraclecl 1 10 0 2.1G 29.9M cpu04 53:42 0 4.57% oracle
23937 oraclecl 1 10 0 2.1G 34.4M run 8:45 0 4.49% oracle
27795 oraclecl 1 40 0 2.1G 29.8M run 52:32 0 4.45% oracle
27799 oraclecl 1 0 0 2.1G 29.7M cpu06 54:39 0 4.43% oracle
27792 oraclecl 1 0 0 2.1G 29.9M run 53:34 0 4.17% oracle
25911 oraclecl 1 60 0 2.1G 28.7M run 53:25 0 4.07% oracle
27837 oraclecl 1 30 0 2.1G 31.0M cpu04 53:25 0 3.84% oracle
26219 oraclecl 1 60 0 2.1G 30.9M run 52:39 0 3.81% oracle
# lockstat sleep 5 | more
Adaptive mutex spin: 1296604 events in 5.002 seconds (259224 events/sec)
Count indv cuml rcnt spin Lock Caller
-----------------------------------------------------------------------------
-- 1279438 99% 99% 1.00 20 0x30014be59b0 segspt_softunlock+0xcc 3442 0% 99% 1.00 10741 0x30014be59a8 spt_anon_getpages+0x4 70 0% 99% 1.00 6 ph_mutex+0x38 page_find+0x40 62 0% 99% 1.00 7 ph_mutex+0x40 page_find+0x40 60 0% 99% 1.00 86 0x310024470e0 hat_unlock+0xd4 59 0% 99% 1.00 77 0x31002447060 hat_unlock+0xd4 59 0% 99% 1.00 74 0x31002447040 hat_unlock+0xd4 59 0% 99% 1.00 82 0x310024471c0 hat_unlock+0xd4 59 0% 99% 1.00 77 0x310024471a0 hat_unlock+0xd4 58 0% 99% 1.00 83 0x310024470c0 hat_unlock+0xd4 58 0% 99% 1.00 82 0x31002447080 hat_unlock+0xd4 58 0% 99% 1.00 80 0x31002447020 hat_unlock+0xd4 58 0% 99% 1.00 86 0x31002447000 hat_unlock+0xd4 58 0% 99% 1.00 80 0x310024471e0 hat_unlock+0xd4 58 0% 99% 1.00 75 0x31002447180 hat_unlock+0xd4 58 0% 99% 1.00 80 0x31002447160 hat_unlock+0xd4 58 0% 99% 1.00 33 0x30017091670 lwp_mutex_wakeup+0x11c Can somebody explain to me what segspt_softunlock is? Or why is this box spending so much time making kernel system calls? Any insight would be deeply appreciated. Thanks. - Sanjay _______________________________________________ sunmanagers mailing list sunmanagers@sunmanagers.org http://www.sunmanagers.org/mailman/listinfo/sunmanagers
This archive was generated by hypermail 2.1.7 : Wed Apr 09 2008 - 23:27:35 EDT