Hi admins,
I have a problem mixed with Oracle. This sometimes occur on my system. Sysptoms are;
Oracle locks. This occurs when the checkpoint is to be done (which occurs at log switch (which is the only cause for checkpoint in our Oracle system)
What Oracle and my DBA say is that; there is an I/O problem. When the I/O load is heavy it causes Oracle to be locked. No more.
Our Oracle RedoLog file is on re0 and, at the time of lock its tps/bps was zero (0) for a while and some times (in every 5-10 sec) a big data was passing (120 K, 264K, and this changes). Other times re0 shows 50 to 200 K data traffic(every second). But my root is also on re0 and I can easily got an ls on it at the time of crash. But my mouse sometimes stopped on the console which shows system sloweness. I could not observe the disks where oracle data files are( disks are numbered as re1..re4 but I could not observe them with the iostat, instead rz8..rz12 are shown. I have been missing something may be)
Also Oracle user processes starts to increase from 317 to 400 (which is our oracle max user process number).
Other users and shells are running without problem. And Oracle starts to continue after a while.
Now I ask, how can i analyze the problem, especially how can i determine whether there is a problem with disk I/O.
Thanks in advance for your time...
My system:
alpha 4100 /466 MHz* 2 CPU
2 GB RAM
100 GB HD
vmstat 1 3 command gave the following:
2 hours before :
Virtual Memory Statistics: (pagesize = 8192)
procs memory pages intr cpu
r w u act free wire fault cow zero react pin pout in sy cs us sy id
9910 33 128K 112 27K 6M 1M 2M 83K 1M 6619 50 9K 946 29 3 68
9912 32 128K 123 27K 1566 129 1173 0 149 0 501 46K 3K 77 23 0
10911 33 128K 95 27K 995 41 917 0 49 0 668 53K 3K 81 19 0
When the Oracle Locked, vmstat 2 3 output.
Virtual Memory Statistics: (pagesize = 8192)
procs memory pages intr cpu
r w u act free wire fault cow zero react pin pout in sy cs us sy id
7979 32 128K 104 28K 7M 1M 2M 105K 2M 8271 58 9K 975 30 4 66
8980 32 127K 127 28K 1729 190 1274 0 191 0 215 2K 1K 92 8 0
7982 32 127K 125 28K 937 0 937 0 0 0 127 1K 1K 94 6 0
In the Oracle logs, the folowing is written,
Current log# 11 seq# 8228 mem# 0: /ctl/redo/log11SEVDE.rdo
Tue Mar 7 13:09:46 2000
Thread 1 advanced to log sequence 8229
Current log# 12 seq# 8229 mem# 0: /ctl/redo/log12SEVDE.rdo
Tue Mar 7 13:25:46 2000
Thread 1 cannot allocate new log, sequence 8230
Checkpoint not complete
Current log# 12 seq# 8229 mem# 0: /ctl/redo/log12SEVDE.rdo
Tue Mar 7 13:40:10 2000
Thread 1 advanced to log sequence 8230
Current log# 13 seq# 8230 mem# 0: /ctl/redo/log13SEVDE.rdo
Tue Mar 7 14:10:08 2000
Received on Tue Mar 07 2000 - 16:19:57 NZDT