[Pacemaker] configuring postgresql streaming replication cluster

[Andrew Beekhof]

On 11/06/2013, at 10:05 PM, Gregg Jaskiewicz <gryzman at gmail.com> wrote:

> Hi guys, 
> 
> I'm trying to wrap my head around the pacemaker, and setting up postgresql cluster using pcs on centos 6.4. 
> 
> I used so far following commands to set it up. And this seems to work, but all nodes are running as slaves (postgresql wise). 
> The 'master' node (IP address wise) even connects to itself to replicate (that's a bit of a snafu in postgresql for allowing it). 
> 
> So question is, how can this be improved, what did I miss, and why doesn't postgresql master get promoted to be 'master' on the 'master' node? 

There is no 'master' node.  Just nodes on which some things are running.
Does the postgresql agent call crm_master?  Thats the mechanism by which pacemaker knows which instance to promote

> 
> 
> pcs property set stonith-enabled=false
> pcs property set no-quorum-policy=ignore
> 
> pcs resource rsc defaults resource-stickiness="INFINITY" migration-threshold="1"
> 
> # add virtual IP resource for postgres services master
> pcs resource create vip-master ocf:heartbeat:IPaddr2 params ip=10.0.0.100 nic=eth1:0 cidr_netmask=24 meta migration-threshold=0 \
>    op start   timeout=60s interval=0s  on-fail=restart \
>    op monitor timeout=60s interval=10s on-fail=restart \
>    op stop    timeout=60s interval=0s  on-fail=block
> 
> # add virtual IP resource for postgres replication master_ip
> pcs resource create vip-rep ocf:heartbeat:IPaddr2 params ip=10.0.1.100 nic=eth2:0 cidr_netmask=24 meta migration-threshold=0 \
>    op start   timeout=60s interval=0s  on-fail=restart \
>    op monitor timeout=60s interval=10s on-fail=restart \
>    op stop    timeout=60s interval=0s  on-fail=block
> 
> # Group them up
> pcs resource group add master-group vip-master vip-rep
> 
> # Start it
> pcs resource start master-group
> 
> # PostgreSQL stuff
> pcs resource create pgsql ocf:heartbeat:pgsql \
>      params pgctl="/usr/pgsql-9.2/bin/pg_ctl" psql="/usr/pgsql-9.2/bin/psql" pgdata="/var/lib/pgsql/9.2/data/" \
>      restore_command="cp /var/lib/pgsql/9.2/data/pg_archive/\%f \%p" \
>      start_opt="-p 5432" \
>      rep_mode="async" \
>      node_list="hanode01 hanode02" \
>      primary_conninfo_opt="keepalives_idle=60 keepalives_interval=5 keepalives_count=5" \
>      master_ip="10.0.1.100" \
>      stop_escalate="0" \
>  op start   timeout="60s" interval="0s"  on-fail="restart" \
>  op monitor timeout="60s" interval="7s"  on-fail="restart" \
>  op monitor timeout="60s" interval="2s"  on-fail="restart" role="Master" \
>  op promote timeout="60s" interval="0s"  on-fail="restart" \
>  op demote  timeout="60s" interval="0s"  on-fail="stop" \
>  op stop    timeout="60s" interval="0s"  on-fail="block" \
>  op notify  timeout="60s" interval="0s"
> 
> pcs resource master msPostgresql pgsql master-max=1 master-node-max=1 clone-max=10 clone-node-max=10 notify=true
> 
> pcs constraint colocation add master-group msPostgresql role=Master 
> 
> # Score needs to be different, otherwise pcs thinks both lines below are trying to do the same
> pcs constraint order start master-group  then promote msPostgresql   symmetrical=false score=2
> pcs constraint order stop  master-group  then demote  msPostgresql   symmetrical=false score=3
> 
> pcs  resource update pgsql repuser=rep monitor_password=lemon monitor_user=monitor
> 
> 
> Thanks in advance !
> 
> -- 
> GJ
> _______________________________________________
> Pacemaker mailing list: Pacemaker at oss.clusterlabs.org
> http://oss.clusterlabs.org/mailman/listinfo/pacemaker
> 
> Project Home: http://www.clusterlabs.org
> Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf
> Bugs: http://bugs.clusterlabs.org