Procedure for cleaning up extra lock managers running

[Troy Morton]

We have a site that has multiple lock managers running on our production server.

Code:

hsprim1b>ps -ef |grep “seb ” |grep lm_ip
    hci  590044       1   0   Aug 28      – 111:16 lm_ip -u 500 -a lm_qdx5.4_seb -z /qdx/qdx5.4/integrator/seb/exec/databases/
    hci  872512       1   0   May 18      – 1888:26 lm_ip -u 500 -a lm_qdx5.4_seb -z /qdx/qdx5.4/integrator/seb/exec/databases/
    hci 1593354       1   0   Sep 14      –  0:25 lm_ip -u 500 -a lm_qdx5.4_seb -z /qdx/qdx5.4/integrator/seb/exec/databases/
    hci 1605724       1   0   Aug 28      –  0:49 lm_ip -u 500 -a lm_qdx5.4_seb -z /qdx/qdx5.4/integrator/seb/exec/databases/
    hci 2105428       1   0   Apr 16      – 529:23 lm_ip -u 500 -a lm_qdx5.4_seb -z /qdx/qdx5.4/integrator/seb/exec/databases/
    hci 2154580       1   0   Jun 30      – 392:18 lm_ip -u 500 -a lm_qdx5.4_seb -z /qdx/qdx5.4/integrator/seb/exec/databases/
    hci 2674728       1   0   Aug 28      –  0:51 lm_ip -u 500 -a lm_qdx5.4_seb -z /qdx/qdx5.4/integrator/seb/exec/databases/
    hci 3362860       1   0   Aug 28      –  0:49 lm_ip -u 500 -a lm_qdx5.4_seb -z /qdx/qdx5.4/integrator/seb/exec/databases/
    hci 3399872       1   0   Sep 14      – 105:37 lm_ip -u 500 -a lm_qdx5.4_seb -z /qdx/qdx5.4/integrator/seb/exec/databases/

The pid file in $HCISITEDIR/exec/hcilockmanager shows that process 3399872 is the active lock manager for this site.

Code:

hsprim1b>setsite seb
hsprim1b>cat $HCISITEDIR/exec/hcilockmgr/pid
3399872

Can I kill the extra lock managers without causing the processes in those sites to panic?  I’d like to be able to clean this up without taking the entire site down.

Thanks

Troy

Topic