Hi,

On Mon, Jun 15, 2009 at 08:25:14PM +0530, Lasantha Marian wrote:
> Hi,
> 
> I have a two node cluster setup using Heartbeat 2.99.3/Pacemaker 1.0.2
> on an Ubuntu 8.10 server built on two identical Dell 2900 servers.
> Basically Heartbeat works as expected. My configuration uses CRM to
> manage and DRBD for storages.
> 
> I have noticed the following.
> 
> If node A is running and when node node B has failed, and then if node A
> is rebooted (whilst node B still reaming failed), all the resources on
> the node A would not start. Is this the expected behavior ? But I was
> expecting node A to bring all resources up even when node B is being
> temporarily unavailable. Can this be achieved ?
> 
> If both nodes (A and B) are started again, all resources come up correctly.
> 
> In the log it mostly shows "WARN: native_color: Resource dr0:1 cannot
> run anywhere" and similar.

You don't have fencing (stonith) configured. With shared storage
it's a must. It's a must anyway with only two nodes. Though you
can disable stonith in properties, while you're testing and
don't care about the data.

Thanks,

Dejan

> Any help/explanation is appreciated.
> 
> Best regards,
> 
> Lasantha.

> primitive dr0 ocf:heartbeat:drbd \
>       params drbd_resource="www" \
>       op monitor interval="59s" role="Master" timeout="30s" \
>       op monitor interval="60s" role="Slave" timeout="30s"
> primitive dr1 ocf:heartbeat:drbd \
>       params drbd_resource="spool" \
>       op monitor interval="59s" role="Master" timeout="30s" \
>       op monitor interval="60s" role="Slave" timeout="30s"
> primitive dr3 ocf:heartbeat:drbd \
>       params drbd_resource="data" \
>       op monitor interval="59s" role="Master" timeout="30s" \
>       op monitor interval="60s" role="Slave" timeout="30s"
> primitive fs0 ocf:heartbeat:Filesystem \
>       params fstype="ext3" directory="/ha/www" device="/dev/drbd0" 
> options="rw,relatime"
> primitive fs1 ocf:heartbeat:Filesystem \
>       params fstype="ext3" directory="/ha/spool" device="/dev/drbd1" 
> options="rw,relatime"
> primitive fs3 ocf:heartbeat:Filesystem \
>       params fstype="ext3" directory="/ha/data" device="/dev/drbd3" 
> options="rw,relatime"
> primitive ip0 ocf:heartbeat:IPaddr2 \
>       params ip="192.168.0.180" broadcast="192.168.0.255" nic="bond0" 
> iflabel="0" cidr_netmask="24" \
>       op monitor interval="21s" timeout="5s"
> primitive pg0 ocf:heartbeat:pgsql \
>       params pgctl="/usr/local/PostgreSQL/bin/pg_ctl" 
> psql="/usr/local/PostgreSQL/bin/psql" pgdata="/ha/data/PostgreSQL" 
> pgdba="postgresql" pgport="5432" pgdb="template1" \
>       op monitor interval="10s" timeout="10s"
> group rg0 fs0 fs1 fs3 ip0 pg0
> ms ms0 dr0 \
>       meta clone-max="2" notify="true" globally-unique="false" 
> target-role="Started"
> ms ms1 dr1 \
>       meta clone-max="2" notify="true" globally-unique="false" 
> target-role="Started"
> ms ms3 dr3 \
>       meta clone-max="2" notify="true" globally-unique="false" 
> target-role="Started"
> location ms0-master-on-connected-node ms0 \
>       rule $id="ms0-master-on-connected-node-rule" $role="master" -inf: 
> not_defined pingd or pingd lte 0
> location ms1-master-on-connected-node ms1 \
>       rule $id="ms1-master-on-connected-node-rule" $role="master" -inf: 
> not_defined pingd or pingd lte 0
> location ms3-master-on-connected-node ms3 \
>       rule $id="ms3-master-on-connected-node-rule" $role="master" -inf: 
> not_defined pingd or pingd lte 0
> colocation rg0-on-ms0 inf: rg0 ms0:Master
> colocation rg0-on-ms1 inf: rg0 ms1:Master
> colocation rg0-on-ms3 inf: rg0 ms3:Master
> order ms0-before-rg0 inf: ms0:promote rg0:start
> order ms1-before-rg0 inf: ms1:promote rg0:start
> order ms3-before-rg0 inf: ms3:promote rg0:start
> property $id="cib-bootstrap-options" \
>       dc-version="1.0.2-c02b459053bfa44d509a2a0e0247b291d93662b7" \
>       no-quorum-policy="ignore"

> _______________________________________________
> Linux-HA mailing list
> [email protected]
> http://lists.linux-ha.org/mailman/listinfo/linux-ha
> See also: http://linux-ha.org/ReportingProblems
_______________________________________________
Linux-HA mailing list
[email protected]
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems

Reply via email to