There is bug in PCS 0.9.26: configure pacemaker resource agent for PG stream replication
Hi all,
I just complete my HA configuration on Redhat 6.4 enterprise:
Clusterware: Pacemaker 1.1.8 ( CMAN .0.12.1, corosync 1.4.1 )
Resource manager: PCS 0.9.26
PostgreSQL 9.3.3.1
Detailed configuration follow:
http://clusterlabs.org/wiki/PgSQL_Replicated_Cluster#PostgreSQL_.28node1_only.29
Complete the configuration, we find all PostgreSQL is slave status. We
finally find the root cause after google searching. PCS don't apply all
resource configuration definition, it cut configuration parameter at end of
the first "op"
For example:
pcs -f pgsql_cfg resource create pgsql pgsql \
pgctl="/usr/bin/pg_ctl" \
psql="/usr/bin/psql" \
pgdata="/var/lib/pgsql/data/" \
rep_mode="sync" \
node_list="node1 node2" \
restore_command="cp /var/lib/pgsql/pg_archive/%f %p" \
primary_conninfo_opt="keepalives_idle=60 keepalives_interval=5
keepalives_count=5" \
master_ip="192.168.2.3" \
restart_on_promote='true' \
op start timeout="60s" interval="0s" on-fail="restart" \
==============Other op definition is cut
After we upgrade PCS version to the latest 0.9.90. All problem disappear
:)
BWT, we find the app connection does not be interrupted when RA
automatically switch sync replication to async replication ( for example we
kill the slave node).
--
View this message in context: http://postgresql.1045698.n5.nabble.com/There-is-bug-in-PCS-0-9-26-configure-pacemaker-resource-agent-for-PG-stream-replication-tp5795081.html
Sent from the PostgreSQL - general mailing list archive at Nabble.com.
--
Sent via pgsql-general mailing list (pgsql-general@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-general
On Fri, Mar 7, 2014 at 3:32 PM, leo <dazhoufei@gmail.com> wrote:
I just complete my HA configuration on Redhat 6.4 enterprise:
Clusterware: Pacemaker 1.1.8 ( CMAN .0.12.1, corosync 1.4.1 )
Resource manager: PCS 0.9.26
PostgreSQL 9.3.3.1
Detailed configuration follow:
http://clusterlabs.org/wiki/PgSQL_Replicated_Cluster#PostgreSQL_.28node1_only.29
As this issue is related to Pacemaker (or PCS, the Pacemaker/Corosync
configuration system AFAIK) and not directly Postgres, you should ask
this question directly to the Pacemaker mailing list:
http://oss.clusterlabs.org/mailman/listinfo/pacemaker
Regards,
--
Michael
--
Sent via pgsql-general mailing list (pgsql-general@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-general
Michael Paquier wrote
On Fri, Mar 7, 2014 at 3:32 PM, leo <
dazhoufei@
> wrote:
I just complete my HA configuration on Redhat 6.4 enterprise:
Clusterware: Pacemaker 1.1.8 ( CMAN .0.12.1, corosync 1.4.1 )
Resource manager: PCS 0.9.26
PostgreSQL 9.3.3.1
Detailed configuration follow:
http://clusterlabs.org/wiki/PgSQL_Replicated_Cluster#PostgreSQL_.28node1_only.29As this issue is related to Pacemaker (or PCS, the Pacemaker/Corosync
configuration system AFAIK) and not directly Postgres, you should ask
this question directly to the Pacemaker mailing list:
http://oss.clusterlabs.org/mailman/listinfo/pacemaker
Regards,
I just about wrote the same...
This is being posted to -general, not -bugs. It reads like an FYI to people
who may be using PCS and PostgreSQL and as such is on-topic for -general.
A little more preamble saying such would have helped...
David J.
--
View this message in context: http://postgresql.1045698.n5.nabble.com/There-is-bug-in-PCS-0-9-26-configure-pacemaker-resource-agent-for-PG-stream-replication-tp5795081p5795159.html
Sent from the PostgreSQL - general mailing list archive at Nabble.com.
--
Sent via pgsql-general mailing list (pgsql-general@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-general