[ovirt-users] Sharing iSCSI data stroage domain across multiple clusters in the same datacenter

Itamar Heim iheim at redhat.com
Mon Aug 4 20:42:45 UTC 2014


On 08/04/2014 05:46 PM, santosh wrote:
>
> On 08/03/2014 03:01 PM, Itamar Heim wrote:
>> On 07/30/2014 10:35 PM, santosh wrote:
>>> Hi,
>>> *
>>> **Can we share the iSCSI data storage domain across multiple clusters in
>>> the same datacenter?*
>>>
>>> Following are the setup details which I tried.
>>>
>>>      -    One datacenter, Say DC1
>>>      -    in DC1, two clusters, say CL1 and CL2
>>>      -    In CL1, one host, say H1. And in CL2 one host, say H2
>>>      -    iSCSI Data Storage domain is configured where external storage
>>>      LUNs are exported to host H1(A host in CL1 of Datacenter).
>>>
>>>
>>> While adding H1 to CL1 is succeeded; addition of H2 in CL2 is failing
>>> with following error in vdsm.log.
>>>
>>>      Traceback (most recent call last):
>>>         File "/usr/share/vdsm/storage/task.py", line 873, in _run
>>>           return fn(*args, **kargs)
>>>         File "/usr/share/vdsm/logUtils.py", line 45, in wrapper
>>>           res = f(*args, **kwargs)
>>>         File "/usr/share/vdsm/storage/hsm.py", line 1020, in
>>>      connectStoragePool
>>>           spUUID, hostID, msdUUID, masterVersion, domainsMap)
>>>         File "/usr/share/vdsm/storage/hsm.py", line 1091, in
>>>      _connectStoragePool
>>>           res = pool.connect(hostID, msdUUID, masterVersion)
>>>         File "/usr/share/vdsm/storage/sp.py", line 630, in connect
>>>           self.__rebuild(msdUUID=msdUUID, masterVersion=masterVersion)
>>>         File "/usr/share/vdsm/storage/sp.py", line 1153, in __rebuild
>>>           self.setMasterDomain(msdUUID, masterVersion)
>>>         File "/usr/share/vdsm/storage/sp.py", line 1360, in setMasterDomain
>>>           raise se.StoragePoolMasterNotFound(self.spUUID, msdUUID)
>>>      StoragePoolMasterNotFound: Cannot find master domain:
>>>      'spUUID=a4dfaf64-adfa-4cfa-88d5-986fbdb2b2b2,
>>>      msdUUID=741f7913-09ad-4d96-a225-3bda6d06e042'
>>>      Thread-13::DEBUG::2014-07-30
>>>      15:24:49,780::task::885::TaskManager.Task::(_run)
>>>      Task=`07997682-8d6b-42fd-acb3-1360f14860d6`::Task._run:
>>>      07997682-8d6b-42fd-acb3-1360f14860d6
>>>      ('a4dfaf64-adfa-4cfa-88d5-986fbdb2b2b2', 2,
>>>      '741f7913-09ad-4d96-a225-3bda6d06e042', 1, None) {} failed -
>>>      stopping task
>>>      Thread-13::DEBUG::2014-07-30
>>>      15:24:49,780::task::1211::TaskManager.Task::(stop)
>>>      Task=`07997682-8d6b-42fd-acb3-1360f14860d6`::stopping in state
>>>      preparing (force False)
>>>      Thread-13::DEBUG::2014-07-30
>>>      15:24:49,780::task::990::TaskManager.Task::(_decref)
>>>      Task=`07997682-8d6b-42fd-acb3-1360f14860d6`::ref 1 aborting True
>>>      *Thread-13::INFO::2014-07-30
>>>      15:24:49,780::task::1168::TaskManager.Task::(prepare)
>>>      Task=`07997682-8d6b-42fd-acb3-1360f14860d6`::aborting: Task is
>>>      aborted: 'Cannot find master domain' - code 304*
>>>      Thread-13::DEBUG::2014-07-30
>>>      15:24:49,781::task::1173::TaskManager.Task::(prepare)
>>>      Task=`07997682-8d6b-42fd-acb3-1360f14860d6`::Prepare: aborted:
>>>      Cannot find master domain
>>>      Thread-13::DEBUG::2014-07-30
>>>      15:24:49,781::task::990::TaskManager.Task::(_decref)
>>>      Task=`07997682-8d6b-42fd-acb3-1360f14860d6`::ref 0 aborting True
>>>      Thread-13::DEBUG::2014-07-30
>>>      15:24:49,781::task::925::TaskManager.Task::(_doAbort)
>>>      Task=`07997682-8d6b-42fd-acb3-1360f14860d6`::Task._doAbort: force False
>>>      Thread-13::DEBUG::2014-07-30
>>>      15:24:49,781::resourceManager::977::ResourceManager.Owner::(cancelAll)
>>>      Owner.cancelAll requests {}
>>>      Thread-13::DEBUG::2014-07-30
>>>      15:24:49,781::task::595::TaskManager.Task::(_updateState)
>>>      Task=`07997682-8d6b-42fd-acb3-1360f14860d6`::moving from state
>>>      preparing -> state aborting
>>>      Thread-13::DEBUG::2014-07-30
>>>      15:24:49,781::task::550::TaskManager.Task::(__state_aborting)
>>>      Task=`07997682-8d6b-42fd-acb3-1360f14860d6`::_aborting: recover
>>>      policy none
>>>      Thread-13::DEBUG::2014-07-30
>>>      15:24:49,782::task::595::TaskManager.Task::(_updateState)
>>>      Task=`07997682-8d6b-42fd-acb3-1360f14860d6`::moving from state
>>>      aborting -> state failed
>>>      Thread-13::DEBUG::2014-07-30
>>>      15:24:49,782::resourceManager::940::ResourceManager.Owner::(releaseAll)
>>>      Owner.releaseAll requests {} resources {}
>>>      Thread-13::DEBUG::2014-07-30
>>>      15:24:49,782::resourceManager::977::ResourceManager.Owner::(cancelAll)
>>>      Owner.cancelAll requests {}
>>>      Thread-13::ERROR::2014-07-30
>>>      15:24:49,782::dispatcher::65::Storage.Dispatcher.Protect::(run)
>>>      {'status': {'message': "Cannot find master domain:
>>>      'spUUID=a4dfaf64-adfa-4cfa-88d5-986fbdb2b2b2,
>>>      msdUUID=741f7913-09ad-4d96-a225-3bda6d06e042'", 'code': 304}}
>>>
>>> _*Please advise if I need to have one Storage Domain per cluster in
>>> given datacenter.*_
>>>
>>> Thanks, Santosh.
>>>
>>>
>>> ***************************Legal Disclaimer***************************
>>> "This communication may contain confidential and privileged material for the
>>> sole use of the intended recipient. Any unauthorized review, use or distribution
>>> by others is strictly prohibited. If you have received the message by mistake,
>>> please advise the sender by reply email and delete the message. Thank you."
>>> **********************************************************************
>>>
>>>
>>>
>>> _______________________________________________
>>> Users mailing list
>>> Users at ovirt.org
>>> http://lists.ovirt.org/mailman/listinfo/users
>>>
>> no. you can access a storage domain from multiple clusters in same DC.
> [Santosh] That is my understanding also.
>> which type of storage?
> [Santosh] 'Storage' is of 'iSCSI' type. Domain type is 'Data (Master)'
>> could it be its "protecting" shared access to the DC?
> [Santosh] How should I find it out?

which type of storage array?

>> is moving the host to same cluster as the SPM resolving the issue for
>> you (it shouldn't)
> [Santosh] Moving the host to the same cluster as that of SPM did not
> resolve the issue.

and if you move the SPM host to maint, does the 2nd host able to become 
an SPM/access the storage (which would indicate an issue with storage 
side 'protection'.
else, are you sure all hosts can access the storage?

>
>
> Thanks for the reply.
> Regards, Santosh.
>
>
> ***************************Legal Disclaimer***************************
> "This communication may contain confidential and privileged material for the
> sole use of the intended recipient. Any unauthorized review, use or distribution
> by others is strictly prohibited. If you have received the message by mistake,
> please advise the sender by reply email and delete the message. Thank you."
> **********************************************************************
>




More information about the Users mailing list