Im trying to cluster the PostgreSQL database in active-passive mode with two nodes.
I have tested it and If I fence one node, it fails over the other node, but when I run the command to manually relocate the service:
clusvcadm -r BBDD -m NODO1
when I have open connections open on the database I get this error messages on /var/log/messages:
Jul 15 09:27:29 NODO2 clurgmgrd[2493]: <notice> Stopping service service:BBDD
Jul 15 09:27:33 NODO2 clurgmgrd: [2493]: <err> Stopping Service postgres-8:BBDD > Failed - Application Is Still Running
Jul 15 09:27:33 NODO2 clurgmgrd: [2493]: <err> Stopping Service postgres-8:BBDD > Failed
Jul 15 09:27:33 NODO2 clurgmgrd[2493]: <notice> stop on postgres-8 "BBDD" returned 1 (generic error)
Jul 15 09:27:33 NODO2 clurgmgrd: [2493]: <err> Stopping Service postgres-8:BBDD > Failed
Jul 15 09:27:33 NODO2 clurgmgrd[2493]: <notice> stop on postgres-8 "BBDD" returned 1 (generic error)
Jul 15 09:27:33 NODO2 avahi-daemon[2304]: Withdrawing address record for 192.168.1.183 on eth0.
Jul 15 09:27:43 NODO2 clurgmgrd[2493]: <crit> #12: RG service:BBDD failed to stop; intervention required
Jul 15 09:27:43 NODO2 clurgmgrd[2493]: <notice> Service service:BBDD is failed
Jul 15 09:27:43 NODO2 clurgmgrd[2493]: <warning> #70: Failed to relocate service:BBDD; restarting locally
Jul 15 09:27:43 NODO2 clurgmgrd[2493]: <err> #43: Service service:BBDD has failed; can not start.
Jul 15 09:27:44 NODO2 clurgmgrd[2493]: <alert> #2: Service service:BBDD returned failure code. Last Owner: NODO2.
Jul 15 09:27:44 NODO2 clurgmgrd[2493]: <alert> #4: Administrator intervention required.
If I check the status of the service:
Service Name Owner (Last) State
------- ---- ----- ------ -----
service:BBDD (NODO2) failed
and If I check with ps:
# ps aux | grep postgres
root 21552 0.0 0.2 2844 1120 ? S< 09:27 0:00 su - postgres -c /usr/bin/postmaster -c config_file="/etc/cluster/postgres-8/postgres-8:BBDD/postgresql.conf" ??-D /nfsvol/pgsql/data
postgres 21553 0.1 0.5 21504 3076 ? S<s 09:27 0:00 /usr/bin/postmaster -c config_file=/etc/cluster/postgres-8/postgres-8:BBDD/postgresql.conf -D /nfsvol/pgsql/data
postgres 21591 0.0 0.1 11284 608 ? S< 09:27 0:00 postgres: logger process
postgres 21593 0.0 0.1 21504 896 ? S< 09:27 0:00 postgres: writer process
postgres 21594 0.0 0.1 12284 608 ? S< 09:27 0:00 postgres: stats buffer process
postgres 21595 0.0 0.1 11428 804 ? S< 09:27 0:00 postgres: stats collector process
postgres 21720 0.0 0.8 22280 4328 ? S< 09:27 0:00 postgres: postgres postgres 192.168.1.170(2849) idle
So, I think the problem is with the stop script that comes with the cluster suite. I think it must close all the open connections or wait until they finish...
Anyone has this problem? and how can it be solved
Thanks in advance
ESG
-- Linux-cluster mailing list Linux-cluster@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/linux-cluster
