12-26-2009, 01:17 PM
DBSPI20-8 DBSPI collector <.....> is still running or ww_dbmon.exe is still running
Solution:
It shows dbspi was not able to execute all metrics since previous schedule.
If issue is continuous (see dbspierror), check (or ask DBA to check) if one of the metric is not fast enough compared to the schedule.
It may be necessary to modify frequency of long running metric within db_mon.cfg. This change should be done by DBA.
Root cause also be some blocked processed. Look at running process and search for ‘old’ dbspi* or *dbmon* processes.
If some dbspi* or *dbmon* processes are stucked (i.e. been there for more than an hour or so), kill them.
It can also come from strange OVO agent behaviour ( sometime OVO buffers actions, and then execute many actions at the same time).
When it happens after HealthCheck sends “No HC message received for the last xx min”, issue is not from dbspi but from temporary issue with ovo agent.
You also need to check opcagt processes. Sometime, 2 OVO agent are running:
eclipse:/var/opt/OV/dbspi (root) ps -elf | grep opc
1 R root 15488 1 0 152 20 792f1ec0 296 - Oct 4 ? 3:18 opcacta
1 R root 15490 1 0 152 20 61252d00 545 - Oct 4 ? 10:35 opcmona
1 R root 15491 1 0 152 20 4b766040 377 - Oct 4 ? 0:06 opcmsgi
1 R root 15489 1 0 152 20 4e013700 727 - Oct 4 ? 9:03 opcle
1 R root 15326 1 0 152 20 5c1eb580 1010 - Oct 11 ? 1:47 opcctla -start
1 R root 15448 15326 0 152 20 4abe8480 417 - Oct 11 ? 0:13 opcmsgi
1 R root 15447 15326 0 152 20 48ab8040 553 - Oct 11 ? 7:27 opcmona
1 R root 15445 15326 0 152 20 4abe8980 308 - Oct 11 ? 2:09 opcacta
1 R root 15443 15326 0 152 20 4abe8d40 581 - Oct 11 ? 2:04 opcmsga
1 S root 1163 22231 0 154 20 48891900 23 927df4ee 10:49:40 pts/1 0:00 grep opc
1 R root 15446 15326 0 152 20 6012b340 723 - Oct 11 ? 5:54 opcle
For such issue, you need to kill both and restart it. Also try to schedule an agent upgrade if running old OVO agent version.
Solution:
It shows dbspi was not able to execute all metrics since previous schedule.
If issue is continuous (see dbspierror), check (or ask DBA to check) if one of the metric is not fast enough compared to the schedule.
It may be necessary to modify frequency of long running metric within db_mon.cfg. This change should be done by DBA.
Root cause also be some blocked processed. Look at running process and search for ‘old’ dbspi* or *dbmon* processes.
If some dbspi* or *dbmon* processes are stucked (i.e. been there for more than an hour or so), kill them.
It can also come from strange OVO agent behaviour ( sometime OVO buffers actions, and then execute many actions at the same time).
When it happens after HealthCheck sends “No HC message received for the last xx min”, issue is not from dbspi but from temporary issue with ovo agent.
You also need to check opcagt processes. Sometime, 2 OVO agent are running:
eclipse:/var/opt/OV/dbspi (root) ps -elf | grep opc
1 R root 15488 1 0 152 20 792f1ec0 296 - Oct 4 ? 3:18 opcacta
1 R root 15490 1 0 152 20 61252d00 545 - Oct 4 ? 10:35 opcmona
1 R root 15491 1 0 152 20 4b766040 377 - Oct 4 ? 0:06 opcmsgi
1 R root 15489 1 0 152 20 4e013700 727 - Oct 4 ? 9:03 opcle
1 R root 15326 1 0 152 20 5c1eb580 1010 - Oct 11 ? 1:47 opcctla -start
1 R root 15448 15326 0 152 20 4abe8480 417 - Oct 11 ? 0:13 opcmsgi
1 R root 15447 15326 0 152 20 48ab8040 553 - Oct 11 ? 7:27 opcmona
1 R root 15445 15326 0 152 20 4abe8980 308 - Oct 11 ? 2:09 opcacta
1 R root 15443 15326 0 152 20 4abe8d40 581 - Oct 11 ? 2:04 opcmsga
1 S root 1163 22231 0 154 20 48891900 23 927df4ee 10:49:40 pts/1 0:00 grep opc
1 R root 15446 15326 0 152 20 6012b340 723 - Oct 11 ? 5:54 opcle
For such issue, you need to kill both and restart it. Also try to schedule an agent upgrade if running old OVO agent version.