[Pacemaker] Drbd disk don't run

Dejan Muhamedagic dejanmm at fastmail.fm
Fri May 15 15:01:14 UTC 2009


Hi,

On Fri, May 15, 2009 at 08:54:31AM -0300, Rafael Emerick wrote:
> Hi, Dejan
> 
> The fist problem are solved, but now i have another.
> When i try to start de ms-drbd11 resource i don't get any error, but in the
> crm_mon i get the log:
> 
> ============
> Last updated: Fri May 15 08:44:11 2009
> Current DC: node1 (57e0232d-5b78-4a1a-976e-e5335ba8266d) - partition with
> quorum
> Version: 1.0.3-b133b3f19797c00f9189f4b66b513963f9d25db9
> 2 Nodes configured, unknown expected votes
> 2 Resources configured.
> ============
> 
> Online: [ node1 node2 ]
> 
> Clone Set: drbdinit
>         Started: [ node1 node2 ]
> 
> Failed actions:
>     drbd11:0_start_0 (node=node1, call=9, rc=1, status=complete): unknown
> error
>     drbd11_start_0 (node=node1, call=17, rc=1, status=complete): unknown
> error
>     drbd11:1_start_0 (node=node2, call=9, rc=1, status=complete): unknown
> error
>     drbd11_start_0 (node=node2, call=16, rc=1, status=complete): unknown
> error
> 
> So, in the messes log file, i get
> 
> 
> May 15 08:25:03 node1 pengine: [4749]: WARN: unpack_resources: No STONITH
> resources have been defined
> May 15 08:25:03 node1 pengine: [4749]: info: determine_online_status: Node
> node1 is online
> May 15 08:25:03 node1 pengine: [4749]: info: unpack_rsc_op: drbd11:0_start_0
> on node1 returned 1 (unknown error) instead of the expected value: 0 (ok)
> May 15 08:25:03 node1 pengine: [4749]: WARN: unpack_rsc_op: Processing
> failed op drbd11:0_start_0 on node1: unknown error
> May 15 08:25:03 node1 pengine: [4749]: WARN: process_orphan_resource:
> Nothing known about resource drbd11 running on node1
> May 15 08:25:03 node1 pengine: [4749]: info: log_data_element:
> create_fake_resource: Orphan resource <primitive id="drbd11" type="drbd"
> class="ocf" provider="heartbeat" />
> May 15 08:25:03 node1 pengine: [4749]: info: process_orphan_resource: Making
> sure orphan drbd11 is stopped
> May 15 08:25:03 node1 pengine: [4749]: info: unpack_rsc_op: drbd11_start_0
> on node1 returned 1 (unknown error) instead of the expected value: 0 (ok)
> May 15 08:25:03 node1 pengine: [4749]: WARN: unpack_rsc_op: Processing
> failed op drbd11_start_0 on node1: unknown error
> May 15 08:25:03 node1 pengine: [4749]: info: determine_online_status: Node
> node2 is online
> May 15 08:25:03 node1 pengine: [4749]: info: find_clone: Internally renamed
> drbdi:0 on node2 to drbdi:1
> May 15 08:25:03 node1 pengine: [4749]: info: unpack_rsc_op: drbd11:1_start_0
> on node2 returned 1 (unknown error) instead of the expected value: 0 (ok)
> May 15 08:25:03 node1 pengine: [4749]: WARN: unpack_rsc_op: Processing
> failed op drbd11:1_start_0 on node2: unknown error
> May 15 08:25:03 node1 pengine: [4749]: info: unpack_rsc_op: drbd11_start_0
> on node2 returned 1 (unknown error) instead of the expected value: 0 (ok)
> May 15 08:25:03 node1 pengine: [4749]: WARN: unpack_rsc_op: Processing
> failed op drbd11_start_0 on node2: unknown error
> May 15 08:25:03 node1 pengine: [4749]: notice: clone_print: Clone Set:
> drbdinit
> May 15 08:25:03 node1 pengine: [4749]: notice: print_list:     Started: [
> node1 node2 ]
> May 15 08:25:03 node1 pengine: [4749]: notice: clone_print: Master/Slave
> Set: ms-drbd11
> May 15 08:25:03 node1 pengine: [4749]: notice: print_list:     Stopped: [
> drbd11:0 drbd11:1 ]
> May 15 08:25:03 node1 pengine: [4749]: info: get_failcount: ms-drbd11 has
> failed 1000000 times on node1
> May 15 08:25:03 node1 pengine: [4749]: WARN: common_apply_stickiness:
> Forcing ms-drbd11 away from node1 after 1000000 failures (max=1000000)
> May 15 08:25:03 node1 pengine: [4749]: info: get_failcount: drbd11 has
> failed 1000000 times on node1
> May 15 08:25:03 node1 pengine: [4749]: WARN: common_apply_stickiness:
> Forcing drbd11 away from node1 after 1000000 failures (max=1000000)
> May 15 08:25:03 node1 pengine: [4749]: info: get_failcount: ms-drbd11 has
> failed 1000000 times on node2
> May 15 08:25:03 node1 pengine: [4749]: WARN: common_apply_stickiness:
> Forcing ms-drbd11 away from node2 after 1000000 failures (max=1000000)
> May 15 08:25:03 node1 pengine: [4749]: info: get_failcount: drbd11 has
> failed 1000000 times on node2
> May 15 08:25:03 node1 pengine: [4749]: WARN: common_apply_stickiness:
> Forcing drbd11 away from node2 after 1000000 failures (max=1000000)
> May 15 08:25:03 node1 pengine: [4749]: WARN: native_color: Resource drbd11:0
> cannot run anywhere
> May 15 08:25:03 node1 pengine: [4749]: WARN: native_color: Resource drbd11:1
> cannot run anywhere
> May 15 08:25:03 node1 pengine: [4749]: info: master_color: ms-drbd11:
> Promoted 0 instances of a possible 1 to master
> May 15 08:25:03 node1 pengine: [4749]: notice: LogActions: Leave resource
> drbdi:0      (Started node1)
> May 15 08:25:03 node1 pengine: [4749]: notice: LogActions: Leave resource
> drbdi:1      (Started node2)
> May 15 08:25:03 node1 pengine: [4749]: notice: LogActions: Leave resource
> drbd11:0     (Stopped)
> May 15 08:25:03 node1 pengine: [4749]: notice: LogActions: Leave resource
> drbd11:1     (Stopped)
> 
> 
> I had this problem with heartbeatV2, then i'm using pacemaker with the same
> error.
> My idea is that the crm does the management of the drbd, ocfs2 and vmxen

Can ocfs2 run on top of drbd? In that case you need master/master
resource. What you have is master/slave.

> resources to maintain them working...

It does, but this is a resource level problem. Funny that the
logs don't show much. You'll have to try by hand using drbdadm.

> To drbd resource init, the Sonith must be configured?

You must have stonith, in particular since it's shared storage.

Also, set

crm configure property no-quorum-policy=ignore

Thanks,

Dejan

> Thank you!
> 
> On Fri, May 15, 2009 at 7:02 AM, Dejan Muhamedagic <dejanmm at fastmail.fm>wrote:
> 
> > Hi,
> >
> > On Fri, May 15, 2009 at 06:47:37AM -0300, Rafael Emerick wrote:
> > > Hi, Dejan
> > >
> > > thanks for attention
> > > following my cib xml conf
> > > I am newbie with pacemaker, any hint is very welcome! : D
> >
> > The CIB as seen by crm:
> >
> > primitive drbd11 ocf:heartbeat:drbd \
> >        params drbd_resource="drbd11" \
> >        op monitor interval="59s" role="Master" timeout="30s" \
> >        op monitor interval="60s" role="Slave" timeout="30s" \
> >        meta target-role="started" is-managed="true"
> > ms ms-drbd11 drbd11 \
> >        meta clone-max="2" notify="true" globally-unique="false"
> > target-role="stopped"
> >
> > The target-role attribute is defined for both the primitive and
> > the container (ms). You should remove the former:
> >
> > crm configure edit drbd11
> >
> > and remove all meta attributes (the whole "meta" part). And don't
> > forget to remove the backslash in the line above it.
> >
> > Thanks,
> >
> > Dejan
> >
> > > thank you very much
> > > for the help
> > >
> > >
> > > On Fri, May 15, 2009 at 4:46 AM, Dejan Muhamedagic <dejanmm at fastmail.fm
> > >wrote:
> > >
> > > > Hi,
> > > >
> > > > On Thu, May 14, 2009 at 05:13:50PM -0300, Rafael Emerick wrote:
> > > > > Hi, Dejan
> > > > >
> > > > > There is no two set of meta-attributes.
> > > > >
> > > > > I remove the ms-drbd11, add again and the error is the same:
> > > > > Error performing operation: Required data for this CIB API call not
> > found
> > > >
> > > > Can you please post your CIB. As xml.
> > > >
> > > > Thanks,
> > > >
> > > > Dejan
> > > >
> > > > >
> > > > > Thanks,
> > > > >
> > > > >
> > > > > On Thu, May 14, 2009 at 3:43 PM, Dejan Muhamedagic <
> > dejanmm at fastmail.fm
> > > > >wrote:
> > > > >
> > > > > > Hi,
> > > > > >
> > > > > > On Thu, May 14, 2009 at 03:18:15PM -0300, Rafael Emerick wrote:
> > > > > > > Hi,
> > > > > > >
> > > > > > > I'm tryng to make a cluster with xen-ha using drbd and ocfs2...
> > > > > > >
> > > > > > > I want that crm management all resources (xen machines, drbd
> > disks
> > > > and
> > > > > > ocfs2
> > > > > > > filesystem ).
> > > > > > >
> > > > > > > First, a create a clone lsb resource to init drbd with gui
> > interface.
> > > > > > > Now, I'm following this manual
> > > > > > http://clusterlabs.org/wiki/DRBD_HowTo_1.0 to
> > > > > > > create the drbd disk managemnt and after make the ocfs2
> > filesystem.
> > > > > > >
> > > > > > > So, when i run:
> > > > > > > # crm resource start ms-drbd11
> > > > > > > # Multiple attributes match name=target-role
> > > > > > > # Value: stopped
> >  (id=ms-drbd11-meta_attributes-target-role)
> > > > > > > # Value: started        (id=drbd11-meta_attributes-target-role)
> > > > > > > # Error performing operation: Required data for this CIB API call
> > not
> > > > > > found
> > > > > >
> > > > > > As it says, there are multiple matches for the attribute. Don't
> > > > > > know how it came to be. Perhaps you can
> > > > > >
> > > > > > crm configure edit ms-drbd11
> > > > > >
> > > > > > and drop one of them. It could also be that there are two sets of
> > > > > > meta-attributes.
> > > > > >
> > > > > > If crm can't edit the resource (in that case please report it)
> > > > > > then you can try:
> > > > > >
> > > > > > crm configure edit xml ms-drbd11
> > > > > >
> > > > > > Thanks,
> > > > > >
> > > > > > Dejan
> > > > > >
> > > > > > > My messages:
> > > > > > > May 14 15:07:11 node1 pengine: [4749]: info: get_fail count:
> > > > ms-drbd11
> > > > > > has
> > > > > > > failed 1000000 times on node2
> > > > > > > May 14 15:07:11 node1 pengine: [4749]: WARN:
> > common_apply_stickiness:
> > > > > > > Forcing ms-drbd11 away from node2 after 1000000 failures
> > > > (max=1000000)
> > > > > > > May 14 15:07:11 node1 pengine: [4749]: WARN: native_color:
> > Resource
> > > > > > drbd11:0
> > > > > > > cannot run anywhere
> > > > > > > May 14 15:07:11 node1 pengine: [4749]: WARN: native_color:
> > Resource
> > > > > > drbd11:1
> > > > > > > cannot run anywhere
> > > > > > > May 14 15:07:11 node1 pengine: [4749]: info: master_color:
> > ms-drbd11:
> > > > > > > Promoted 0 instances of a possible 1 to master
> > > > > > > May 14 15:07:11 node1 pengine: [4749]: notice: LogActions: Leave
> > > > resource
> > > > > > > drbdi:0      (Started node1)
> > > > > > > May 14 15:07:11 node1 pengine: [4749]: notice: LogActions: Leave
> > > > resource
> > > > > > > drbdi:1      (Started node2)
> > > > > > > May 14 15:07:11 node1 pengine: [4749]: notice: LogActions: Leave
> > > > resource
> > > > > > > drbd11:0     (Stopped)
> > > > > > > May 14 15:07:11 node1 pengine: [4749]: notice: LogActions: Leave
> > > > resource
> > > > > > > drbd11:1     (Stopped)
> > > > > > >
> > > > > > >
> > > > > > > Thank you for any help!
> > > > > >
> > > > > > > _______________________________________________
> > > > > > > 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
> > > >
> >
> > > _______________________________________________
> > > 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