[Pacemaker] drbd doesn't work on simple 2-node cluster.
Kees
chkoehoorn at live.nl
Wed Feb 10 07:42:42 EST 2010
Hello Andrew/All,
Thanks. It works again. This is what I had running after I wanted to
move on to a 3-node cluster.
Two drbd-nodes for storage and one as a webserver via NFS.
So i have to set the cluster back to non-symmetric and fiddle around
with location constraints to get the
tho drbd-nodes back to where they are now. Right?
PS I'm a newbie to this so sorry this stupid mistake!
Greetings,
Kees Koehoorn
> target-role="Started"<--- this isnt going to be helping.
>
> nor is
> symmetric-cluster="false"
> without defining any location constraints.
>
> On Wed, Feb 10, 2010 at 12:43 PM, Kees<chkoehoorn at live.nl> wrote:
>
>> Hello,
>>
>> I changed my config, but it sill won't run?
>> The warings i get are the same:
>> Anyone?
>>
>> Feb 10 12:15:49 test1 pengine: [5152]: notice: clone_print: Master/Slave
>> Set: ms_drbd_websites
>> Feb 10 12:15:49 test1 pengine: [5152]: notice: short_print: Stopped: [
>> drbd_websites:0 drbd_websites:1 ]
>> Feb 10 12:15:49 test1 pengine: [5152]: debug: native_assign_node: Could not
>> allocate a node for drbd_websites:0
>> Feb 10 12:15:49 test1 pengine: [5152]: WARN: native_color: Resource
>> drbd_websites:0 cannot run anywhere
>> Feb 10 12:15:49 test1 pengine: [5152]: debug: native_assign_node: Could not
>> allocate a node for drbd_websites:1
>> Feb 10 12:15:49 test1 pengine: [5152]: WARN: native_color: Resource
>> drbd_websites:1 cannot run anywhere
>> Feb 10 12:15:49 test1 pengine: [5152]: debug: clone_color: Allocated 0
>> ms_drbd_websites instances of a possible 2
>> Feb 10 12:15:49 test1 pengine: [5152]: debug: master_color: drbd_websites:0
>> master score: 0
>> Feb 10 12:15:49 test1 pengine: [5152]: debug: master_color: drbd_websites:1
>> master score: 0
>>
>> node test1 \
>> attributes standby="off"
>> node test2 \
>> attributes standby="off"
>> primitive drbd_websites ocf:linbit:drbd \
>> params drbdconf="/etc/drbd.conf" drbd_resource="r0" \
>> op monitor interval="15s" role="Master" timeout="30s" \
>> op monitor interval="30s" role="Slave" timeout="30s"
>> primitive fs_websites ocf:heartbeat:Filesystem \
>> params device="/dev/drbd/by-res/r0" directory="/websites/"
>> fstype="ext3"
>> primitive ip_storage ocf:heartbeat:IPaddr2 \
>> params ip="10.100.5.4" nic="eth0:0"
>> group websites fs_websites ip_storage
>> ms ms_drbd_websites drbd_websites \
>> meta master-max="1" master-node-max="1" clone-max="2"
>> clone-node-max="1" notify="true" target-role="Started"
>> colocation websites_on_drbd inf: websites ms_drbd_websites:Master
>> order websites_after_drbd inf: ms_drbd_websites:promote websites:start
>> property $id="cib-bootstrap-options" \
>> expected-quorum-votes="1024" \
>> dc-version="1.0.7-54d7869bfe3691eb723b1d47810e5585d8246b58" \
>> cluster-infrastructure="openais" \
>> no-quorum-policy="ignore" \
>> stonith-enabled="false" \
>> symmetric-cluster="false" \
>> last-lrm-refresh="1265725342"
>>
>> Greetings,
>>
>> Kees Koehoorn
>>
>> it seems on the new versions of corosync/pacemaker you will have to specify
>> monitor interval for both master and slave drbd resource. I had this problem
>> before but not sure if you are having the same.
>>
>>
>> ----- Original Message -----
>> From: Kees
>> To: pacemaker at oss.clusterlabs.org
>> Sent: Tuesday, February 09, 2010 10:18 AM
>> Subject: [Pacemaker] drbd doesn't work on simple 2-node cluster.
>> Hi,
>>
>> I'm trying to install a simple 2-node cluster, drbd/failover-IP/filesystem.
>> Followed the usual examples, wich are mostly the same. I got this working
>> before, but now it doesn't work anymore.
>> The problem (i think) is this (from de log):
>>
>> Feb 9 14:03:04 test1 pengine: [5152]: notice: clone_print: Master/Slave
>> Set: ms_drbd_websites
>> Feb 9 14:03:04 test1 pengine: [5152]: notice: short_print: Stopped: [
>> drbd_websites:0 drbd_websites:1 ]
>> Feb 9 14:03:04 test1 pengine: [5152]: debug: native_assign_node: Could not
>> allocate a node for drbd_websites:0
>> Feb 9 14:03:04 test1 pengine: [5152]: WARN: native_color: Resource
>> drbd_websites:0 cannot run anywhere
>> Feb 9 14:03:04 test1 pengine: [5152]: debug: native_assign_node: Could not
>> allocate a node for drbd_websites:1
>> Feb 9 14:03:04 test1 pengine: [5152]: WARN: native_color: Resource
>> drbd_websites:1 cannot run anywhere
>> Feb 9 14:03:04 test1 pengine: [5152]: debug: clone_color: Allocated 0
>> ms_drbd_websites instances of a possible 2
>> Feb 9 14:03:04 test1 pengine: [5152]: debug: master_color: drbd_websites:0
>> master score: 0
>> Feb 9 14:03:04 test1 pengine: [5152]: debug: master_color: drbd_websites:1
>> master score: 0
>> Feb 9 14:03:04 test1 pengine: [5152]: info: master_color: ms_drbd_websites:
>> Promoted 0 instances of a possible 1 to master
>> Feb 9 14:03:04 test1 pengine: [5152]: debug: master_color: drbd_websites:0
>> master score: 0
>> Feb 9 14:03:04 test1 pengine: [5152]: debug: master_color: drbd_websites:1
>> master score: 0
>> Feb 9 14:03:04 test1 pengine: [5152]: info: master_color: ms_drbd_websites:
>> Promoted 0 instances of a possible 1 to master
>> Feb 9 14:03:04 test1 pengine: [5152]: debug: native_create_probe: Probing
>> drbd_websites:0 on test1 (Stopped)
>> Feb 9 14:03:04 test1 pengine: [5152]: debug: native_create_probe: Probing
>> drbd_websites:0 on test2 (Stopped)
>> Feb 9 14:03:04 test1 pengine: [5152]: debug: master_create_actions:
>> Creating actions for ms_drbd_websites
>> Feb 9 14:03:04 test1 pengine: [5152]: debug: child_starting_constraints:
>> ms_drbd_websites has no active children
>> Feb 9 14:03:04 test1 pengine: [5152]: debug: child_stopping_constraints:
>> ms_drbd_websites has no active children
>> Feb 9 14:03:04 test1 pengine: [5152]: notice: LogActions: Leave resource
>> drbd_websites:0#011(Stopped)
>>
>> Why can't my drbd not run anywhere? Must I tell my the drbd-resource which
>> nodes it can use or something?
>> (PS the drbd alone without the cluster-stuff is working)
>>
>> My config:
>>
>> Debian Lenny
>> pacemaker 1.0.7+hg20100203-1 (madkiss)
>> corosync 1.2.0-1 (madkiss)
>> drbd-8.3.4 (compiled from soure)
>>
>>
>> node test1 \
>> attributes standby="off"
>> node test2 \
>> attributes standby="off"
>> primitive drbd_websites ocf:linbit:drbd \
>> params drbdconf="/etc/drbd.conf" drbd_resource="r0" \
>> op monitor interval="29s"
>> primitive fs_websites ocf:heartbeat:Filesystem \
>> params device="/dev/drbd/by-res/r0" directory="/websites/"
>> fstype="ext3"
>> primitive ip_storage ocf:heartbeat:IPaddr2 \
>> params ip="10.100.5.4" nic="eth0:0"
>> group websites fs_websites ip_storage
>> ms ms_drbd_websites drbd_websites \
>> meta master-max="1" master-node-max="1" clone-max="2"
>> clone-node-max="1" notify="true" target-role="Started"
>> colocation websites_on_drbd inf: websites ms_drbd_websites:Master
>> order websites_after_drbd inf: ms_drbd_websites:promote websites:start
>> property $id="cib-bootstrap-options" \
>> expected-quorum-votes="1024" \
>> dc-version="1.0.7-54d7869bfe3691eb723b1d47810e5585d8246b58" \
>> cluster-infrastructure="openais" \
>> no-quorum-policy="ignore" \
>> stonith-enabled="false" \
>> symmetric-cluster="false" \
>> last-lrm-refresh="1265725342"
>>
>> drbd.conf:
>>
>> global {
>> usage-count yes;
>> }
>> common {
>> protocol C;
>> }
>> resource r0 {
>> device /dev/drbd0;
>> meta-disk internal;
>> on test1 {
>> disk /dev/sda11;
>> address 10.100.5.1:7789;
>> }
>> on test2 {
>> disk /dev/sda10;
>> address 10.100.5.2:7789;
>> }
>> }
>>
>>
>> Thanks in advance for your help.
>>
>> Kees Koehoorn
>>
>> ________________________________
>> _______________________________________________
>> Pacemaker mailing list
>> Pacemaker at oss.clusterlabs.org
>> http://oss.clusterlabs.org/mailman/listinfo/pacemaker
>>
>> _______________________________________________
>> Pacemaker mailing list
>> Pacemaker at oss.clusterlabs.org
>> http://oss.clusterlabs.org/mailman/listinfo/pacemaker
>>
>>
>> _______________________________________________
>> Pacemaker mailing list
>> Pacemaker at oss.clusterlabs.org
>> http://oss.clusterlabs.org/mailman/listinfo/pacemaker
>>
>>
>>
> _______________________________________________
> Pacemaker mailing list
> Pacemaker at oss.clusterlabs.org
> http://oss.clusterlabs.org/mailman/listinfo/pacemaker
>
>
>
More information about the Pacemaker
mailing list