Hi,
we are using clustered database with 2 nodes on MTS
we have 2 databases, each with 2 nodes on 2 HP-UX servers
I observed that dispatcher of one particular process is taking 98-99% CPU constantly
which is ora_d000_devdbnod1
Also the following command output is showing c=255 for ora_d000 constantly
I assume it is CPU scheduling though I not much aware of the 'C' field in ps -ef
ps -ef|grep ora_d00
oracle 27271 1 0 Jan 6 ? 0:56 ora_d001_testdbnod1
oracle 10380 1 0 Jan 22 ? 0:06 ora_d002_devdbnod1
oracle 10378 1 0 Jan 22 ? 0:22 ora_d001_devdbnod1
oracle 27279 1 0 Jan 6 ? 1:06 ora_d002_testdbnod1
oracle 10376 1 255 Jan 22 ? 60:05 ora_d000_devdbnod1
oracle 27218 1 0 Jan 6 ? 0:18 ora_d000_testdbnod1
oracle 22955 26042 0 10:46:14 pts/0 0:00 grep ora_d00
What could be the reason for constantly c=255 in ps command and 98-99% cpu usage in top command for this particular process?
Is it because connection requests are picked by a particular dispatcher? If yes, How I can avoid it?
following are some details for disptachers on the server
select * from v$dispatcher;
NAME NETWORK PADDR STATUS ACCEPT MESSAGES BYTES BREAKS OWNED CREATED IDLE BUSY LISTENER CONF_INDX
D000 (ADDRESS=(PROTOCOL=tcp)(HOST=server1.codomain.com)(PORT=63038)) C00000015E89BF98 WAIT YES 1538492 71581793 4 7 394 33847244 14824 0 0
D001 (ADDRESS=(PROTOCOL=tcp)(HOST=server1.codomain.com)(PORT=63039)) C00000015E89C4B8 WAIT YES 373918 200098257 5 4 523 33851281 10750 0 0
D002 (ADDRESS=(PROTOCOL=tcp)(HOST=server1.codomain.com)(PORT=63040)) C00000015E89C9D8 WAIT YES 73427 13669449 0 6 1144 33861236 759 0 0
SQL> show parameter disp
NAME TYPE VALUE
------------------------------------ ----------- ------------------------------
dispatchers string (ADDRESS=(PROTOCOL=TCP))(DISPATCHERS=3)
max_dispatchers integer 10
mts_dispatchers string (ADDRESS=(PROTOCOL=TCP))(DISPATCHERS=3)
mts_max_dispatchers integer 10
Thanks and Regards,
Chetana