<html>
  <head>
    <meta content="text/html; charset=UTF-8" http-equiv="Content-Type">
  </head>
  <body bgcolor="#FFFFFF" text="#000000">
    <div class="moz-cite-prefix">    I patched python source managers.py
      to retry recv() after EINTR, supervdsm works well and the issue
      gone.<br>
          Even declared in python doc that:"only the main thread
      can set a new signal handler, and the main thread will be the only
      one to
      receive signals (this is enforced by the Python <a
        class="reference internal"
        href="http://docs.python.org/2/library/signal.html#module-signal"
        title="signal: Set handlers for asynchronous events."><tt
          class="xref py py-mod docutils literal"><span class="pre">signal</span></tt></a>
      module, even
      if the underlying thread implementation supports sending signals
      to
      individual threads).
      "(<a class="moz-txt-link-freetext" href="http://docs.python.org/2/library/signal.html">http://docs.python.org/2/library/signal.html</a>)<br>
          But according to my test script bellow, the child process
      forker event if it is not the main thread, also got SIGCHILD, Will
      it be a python BUG or feature?<br>
          I agree with Mark maybe we should use synchronised way to deal
      with child process instead of signal handler. <br>
      <pre>import threading</pre>
      <pre>import signal</pre>
      <pre>import time</pre>
      <pre>import os</pre>
      <pre>from multiprocessing import Process,Pipe</pre>
      <pre>
</pre>
      <pre>def _zombieReaper(signum, frame):</pre>
      <pre>    print 'sigchild!!!'

def child():</pre>
      <pre>    time.sleep(5)</pre>
      <pre>
def sleepThread():</pre>
      <pre>    proc = Process(target=child)</pre>
      <pre>    proc.start()</pre>
      <pre>    pip,pop = Pipe()</pre>
      <pre>    pip.recv()&lt;----------<font color="#6666cc">This line will get IOError.EINTR by SIGCHLD</font></pre>
      <pre>
def main():</pre>
      <pre>    signal.signal(signal.SIGCHLD, _zombieReaper)
</pre>
      <pre>    servThread = threading.Thread(target = sleepThread)</pre>
      <pre>    servThread.setDaemon(True)</pre>
      <pre>    servThread.start()</pre>
      <pre>    time.sleep(30)</pre>
      <pre>
if __name__ == '__main__':</pre>
      <pre>    main()</pre>
      On 01/25/2013 03:20 PM, Mark Wu wrote:<br>
    </div>
    <blockquote cite="mid:51023254.3070204@linux.vnet.ibm.com"
      type="cite">Great work!
      <br>
      The default action for SIGCHLD is ignore, so there's no problems
      reported before a signal handler is installed by zombie reaper.
      <br>
      But I still have one problem: the python multiprocessing.manager
      code is running a new thread and according to the implementation
      of python's signal, only the main thread can receive the signal.
      <br>
      So how is the signal delivered to the server thread?
      <br>
      <br>
      <br>
      On Fri 25 Jan 2013 12:30:39 PM CST, Royce Lv wrote:
      <br>
      <blockquote type="cite">
        <br>
        Hi,
        <br>
           I reproduced this issue, and I believe it's a python bug.
        <br>
           1. How to reproduce:
        <br>
           with the test case attached, put it under
        /usr/share/vdsm/tests/,
        <br>
        run #./run_tests.sh superVdsmTests.py
        <br>
           and this issue will be reproduced.
        <br>
           2.Log analyse:
        <br>
           We notice a strange pattern in this log: connectStorageServer
        be
        <br>
        called twice, first supervdsm call succeed, second fails becasue
        of
        <br>
        validateAccess().
        <br>
           That is because for the first call validateAccess returns
        normally
        <br>
        and leave a child there, when the second validateAccess call
        arrives
        <br>
        and multirprocessing manager is receiving the method message, it
        is
        <br>
        just the time first child exit and SIGCHLD comming, this signal
        <br>
        interrupted multiprocessing receive system call, python
        managers.py
        <br>
        should handle INTR and retry recv() like we do in vdsm but it's
        not,
        <br>
        so the second one raise error.
        <br>
        &gt;Thread-18::DEBUG::2013-01-22
        10:41:03,570::misc::85::Storage.Misc.excCmd::(&lt;lambda&gt;)
        '/usr/bin/sudo -n /bin/mount -t nfs -o
        soft,nosharecache,timeo=600,retrans=6,nfsvers=3
        192.168.0.1:/ovirt/silvermoon
        /rhev/data-center/mnt/192.168.0.1:_ovirt_silvermoon' (cwd None)
        <br>
        &gt;Thread-18::DEBUG::2013-01-22
        10:41:03,607::misc::85::Storage.Misc.excCmd::(&lt;lambda&gt;)
        '/usr/bin/sudo -n /bin/mount -t nfs -o
        soft,nosharecache,timeo=600,retrans=6,nfsvers=3
        192.168.0.1:/ovirt/undercity
        /rhev/data-center/mnt/192.168.0.1:_ovirt_undercity' (cwd None)
        <br>
        &gt;Thread-18::ERROR::2013-01-22
        10:41:03,627::hsm::2215::Storage.HSM::(connectStorageServer)
        Could not connect to storageServer
        <br>
        &gt;Traceback (most recent call last):
        <br>
        &gt;  File "/usr/share/vdsm/storage/hsm.py", line 2211, in
        connectStorageServer
        <br>
        &gt;    conObj.connect()
        <br>
        &gt;  File "/usr/share/vdsm/storage/storageServer.py", line 303,
        in connect
        <br>
        &gt;    return self._mountCon.connect()
        <br>
        &gt;  File "/usr/share/vdsm/storage/storageServer.py", line 209,
        in connect
        <br>
        &gt;   
        fileSD.validateDirAccess(self.getMountObj().getRecord().fs_file)
        <br>
        &gt;  File "/usr/share/vdsm/storage/fileSD.py", line 55, in
        validateDirAccess
        <br>
        &gt;    (os.R_OK | os.X_OK))
        <br>
        &gt;  File "/usr/share/vdsm/supervdsm.py", line 81, in __call__
        <br>
        &gt;    return callMethod()
        <br>
        &gt;  File "/usr/share/vdsm/supervdsm.py", line 72, in
        &lt;lambda&gt;
        <br>
        &gt;    **kwargs)
        <br>
        &gt;  File "&lt;string&gt;", line 2, in validateAccess
        <br>
        &gt;  File "/usr/lib64/python2.6/multiprocessing/managers.py",
        line 740, in _callmethod
        <br>
        &gt;    raise convert_to_error(kind, result)
        <br>
        the vdsm side receive RemoteError because of supervdsm server
        <br>
        multiprocessing manager raise error KIND='TRACEBACK'
        <br>
          &gt;RemoteError:
        <br>
        The upper part is the trace back from the client side, the
        following
        <br>
        part is from server side:
        <br>
&gt;---------------------------------------------------------------------------
        <br>
        &gt;Traceback (most recent call last):
        <br>
        &gt;  File "/usr/lib64/python2.6/multiprocessing/managers.py",
        line 214, in serve_client
        <br>
        &gt;    request = recv()
        <br>
        &gt;IOError: [Errno 4] Interrupted system call
        <br>
&gt;---------------------------------------------------------------------------
        <br>
        <br>
        Corresponding Python source code:managers.py(Server side)
        <br>
             def serve_client(self, conn):
        <br>
                 '''
        <br>
                 Handle requests from the proxies in a particular
        process/thread
        <br>
                 '''
        <br>
                 util.debug('starting server thread to service %r',
        <br>
                            threading.current_thread().name)
        <br>
                 recv = conn.recv
        <br>
                 send = conn.send
        <br>
                 id_to_obj = self.id_to_obj
        <br>
                 while not self.stop:
        <br>
                     try:
        <br>
                         methodname = obj = None
        <br>
                         request = recv()&lt;------------------this line
        been interrupted by SIGCHLD
        <br>
                         ident, methodname, args, kwds = request
        <br>
                         obj, exposed, gettypeid = id_to_obj[ident]
        <br>
                         if methodname not in exposed:
        <br>
                             raise AttributeError(
        <br>
                                 'method %r of %r object is not in
        exposed=%r' %
        <br>
                                 (methodname, type(obj), exposed)
        <br>
                                 )
        <br>
                         function = getattr(obj, methodname)
        <br>
                         try:
        <br>
                             res = function(*args, **kwds)
        <br>
                         except Exception, e:
        <br>
                             msg = ('#ERROR', e)
        <br>
                         else:
        <br>
                             typeid = gettypeid and
        gettypeid.get(methodname, None)
        <br>
                             if typeid:
        <br>
                                 rident, rexposed = self.create(conn,
        typeid, res)
        <br>
                                 token = Token(typeid, self.address,
        rident)
        <br>
                                 msg = ('#PROXY', (rexposed, token))
        <br>
                             else:
        <br>
                                 msg = ('#RETURN', res)
        <br>
                     except AttributeError:
        <br>
                         if methodname is None:
        <br>
                             msg = ('#TRACEBACK', format_exc())
        <br>
                         else:
        <br>
                             try:
        <br>
                                 fallback_func =
        self.fallback_mapping[methodname]
        <br>
                                 result = fallback_func(
        <br>
                                     self, conn, ident, obj, *args,
        **kwds
        <br>
                                     )
        <br>
                                 msg = ('#RETURN', result)
        <br>
                             except Exception:
        <br>
                                 msg = ('#TRACEBACK', format_exc())
        <br>
        <br>
                     except EOFError:
        <br>
                         util.debug('got EOF -- exiting thread serving
        %r',
        <br>
                                    threading.current_thread().name)
        <br>
                         sys.exit(0)
        <br>
        <br>
                     except Exception:&lt;------does not handle
        IOError,INTR here should retry recv()
        <br>
                         msg = ('#TRACEBACK', format_exc())
        <br>
        <br>
        <br>
        3. Actions we will take:
        <br>
        (1)As a work round we can first remove the zombie reaper from
        <br>
        supervdsm server
        <br>
        (2)I'll see whether python has a fixed version for this
        <br>
        (3)Yaniv is working on changing vdsm/svdsm communication channel
        to
        <br>
        pipe and handle it ourselves, I believe we'll get rid of this
        with
        <br>
        that properly handled.
        <br>
        <br>
        <br>
        On 01/25/2013 06:00 AM, Dead Horse wrote:
        <br>
        <blockquote type="cite">Tried some manual edits to SD states in
          the dbase. The net result was
          <br>
          I was able to get a node active. However as reconstructing the
          master
          <br>
          storage domain kicked in it was unable to do so. It was also
          not able
          <br>
          to recognize the other SD with similar failure modes to the
          <br>
          unrecognized master above. Guessing the newer VDSM version
          borked
          <br>
          things pretty good. So being as this is a test harness and the
          SD
          <br>
          data is not worth saving I just smoked the all SD. I ran
          <br>
          engine-cleanup and started from fresh and all is well now.
          <br>
          <br>
          - DHC
          <br>
          <br>
          <br>
          On Thu, Jan 24, 2013 at 11:53 AM, Dead Horse
          <br>
          &lt;<a class="moz-txt-link-abbreviated" href="mailto:deadhorseconsulting@gmail.com">deadhorseconsulting@gmail.com</a>
          <br>
          <a class="moz-txt-link-rfc2396E" href="mailto:deadhorseconsulting@gmail.com">&lt;mailto:deadhorseconsulting@gmail.com&gt;</a>&gt; wrote:
          <br>
          <br>
              This test harness setup here consists of two servers tied
          to NFS
          <br>
              storage via IB (NFS mounts are via IPoIB, NFS over RDMA is
          <br>
              disabled) . All storage domains are NFS. The issue does
          occur
          <br>
              with both servers on when attempting to bring them out of
          <br>
              maintenance mode with the end result being non-operational
          due to
          <br>
              storage attach fail.
          <br>
          <br>
              The current issue is now that with a working older commit
          the
          <br>
              master storage domain is "stuck" in state "locked" and I
          see the
          <br>
              secondary issue wherein VDSM cannot seem to find or
          contact the
          <br>
              master storage domain even though it is there. I am can
          mount the
          <br>
              master storage domain manually and and all content appears
          to be
          <br>
              accounted for accordingly on either host.
          <br>
          <br>
              Here is the current contents of the master storage domain
          metadata:
          <br>
              CLASS=Data
          <br>
              DESCRIPTION=orgrimmar
          <br>
              IOOPTIMEOUTSEC=1
          <br>
              LEASERETRIES=3
          <br>
              LEASETIMESEC=5
          <br>
              LOCKPOLICY=
          <br>
              LOCKRENEWALINTERVALSEC=5
          <br>
              MASTER_VERSION=417
          <br>
              POOL_DESCRIPTION=Azeroth
          <br>
             
POOL_DOMAINS=0549ee91-4498-4130-8c23-4c173b5c0959:Active,d8b55105-c90a-465d-9803-8130da9a671e:Active,67534cca-1327-462a-b455-a04464084b31:Active,c331a800-839d-4d23-9059-870a7471240a:Active,f8984825-ff8d-43d9-91db-0d0959f8bae9:Active,c434056e-96be-4702-8beb-82a408a5c8cb:Active,f7da73c7-b5fe-48b6-93a0-0c773018c94f:Active,82e3b34a-6f89-4299-8cd8-2cc8f973a3b4:Active,e615c975-6b00-469f-8fb6-ff58ae3fdb2c:Active,5bc86532-55f7-4a91-a52c-fad261f322d5:Active,1130b87a-3b34-45d6-8016-d435825c68ef:Active<br>
              POOL_SPM_ID=1
          <br>
              POOL_SPM_LVER=6
          <br>
              POOL_UUID=f90a0d1c-06ca-11e2-a05b-00151712f280
          <br>
              REMOTE_PATH=192.168.0.1:/ovirt/orgrimmar
          <br>
              ROLE=Master
          <br>
              SDUUID=67534cca-1327-462a-b455-a04464084b31
          <br>
              TYPE=NFS
          <br>
              VERSION=3
          <br>
              _SHA_CKSUM=1442bb078fd8c9468d241ff141e9bf53839f0721
          <br>
          <br>
              So now with the older working commit I now get this the
          <br>
              "StoragePoolMasterNotFound: Cannot find master domain"
          error
          <br>
              (prior details above when I worked backwards to that
          commit)
          <br>
          <br>
              This is odd as the nodes can definitely reach the master
          storage
          <br>
              domain:
          <br>
          <br>
              showmount from one of the el6.3 nodes:
          <br>
              [root@kezan ~]# showmount -e 192.168.0.1
          <br>
              Export list for 192.168.0.1 <a class="moz-txt-link-rfc2396E" href="http://192.168.0.1">&lt;http://192.168.0.1&gt;</a>:
          <br>
              /ovirt/orgrimmar 192.168.0.0/16
          <a class="moz-txt-link-rfc2396E" href="http://192.168.0.0/16">&lt;http://192.168.0.0/16&gt;</a>
          <br>
          <br>
              mount/ls from one of the nodes:
          <br>
              [root@kezan ~]# mount 192.168.0.1:/ovirt/orgrimmar /mnt
          <br>
              [root@kezan ~]# ls -al
          <br>
              /mnt/67534cca-1327-462a-b455-a04464084b31/dom_md/
          <br>
              total 1100
          <br>
              drwxr-xr-x 2 vdsm kvm    4096 Jan 24 11:44 .
          <br>
              drwxr-xr-x 5 vdsm kvm    4096 Oct 19 16:16 ..
          <br>
              -rw-rw---- 1 vdsm kvm 1048576 Jan 19 22:09 ids
          <br>
              -rw-rw---- 1 vdsm kvm       0 Sep 25 00:46 inbox
          <br>
              -rw-rw---- 1 vdsm kvm 2097152 Jan 10 13:33 leases
          <br>
              -rw-r--r-- 1 vdsm kvm     903 Jan 10 13:39 metadata
          <br>
              -rw-rw---- 1 vdsm kvm       0 Sep 25 00:46 outbox
          <br>
          <br>
          <br>
              - DHC
          <br>
          <br>
          <br>
          <br>
              On Thu, Jan 24, 2013 at 7:51 AM, ybronhei
          &lt;<a class="moz-txt-link-abbreviated" href="mailto:ybronhei@redhat.com">ybronhei@redhat.com</a>
          <br>
              <a class="moz-txt-link-rfc2396E" href="mailto:ybronhei@redhat.com">&lt;mailto:ybronhei@redhat.com&gt;</a>&gt; wrote:
          <br>
          <br>
                  On 01/24/2013 12:44 AM, Dead Horse wrote:
          <br>
          <br>
                      I narrowed down on the commit where the originally
          <br>
                      reported issue crept in:
          <br>
                     
          commitfc3a44f71d2ef202cff18d7203b9e4165b546621building
          <br>
                      and testing with
          <br>
          <br>
                      this commit or subsequent commits yields the
          original issue.
          <br>
          <br>
                  Interesting.. it might be related to this commit and
          we're
          <br>
                  trying to reproduce it.
          <br>
          <br>
                  Did you try to remove that code and run again? does it
          work
          <br>
                  without the additional of zombieReaper?
          <br>
                  does the connectivity to the storage work well? when
          you run
          <br>
                  'ls' on the mounted folder you get see the files
          without a
          <br>
                  long delay ? it might related to too long timeout when
          <br>
                  validating access to this mount..
          <br>
                  we work on that.. any additional info can help
          <br>
          <br>
                  Thanks.
          <br>
          <br>
          <br>
                      - DHC
          <br>
          <br>
          <br>
                      On Wed, Jan 23, 2013 at 3:56 PM, Dead Horse
          <br>
                      &lt;<a class="moz-txt-link-abbreviated" href="mailto:deadhorseconsulting@gmail.com">deadhorseconsulting@gmail.com</a>
          <br>
                     
          <a class="moz-txt-link-rfc2396E" href="mailto:deadhorseconsulting@gmail.com">&lt;mailto:deadhorseconsulting@gmail.com&gt;</a>&gt;wrote:
          <br>
          <br>
                          Indeed reverting back to an older vdsm clears
          up the
          <br>
                          above issue. However
          <br>
                          now I the issue is see is:
          <br>
                          Thread-18::ERROR::2013-01-23
          <br>
                         
          15:50:42,885::task::833::TaskManager.Task::(_setError)
          <br>
                         
          Task=`08709e68-bcbc-40d8-843a-d69d4df40ac6`::Unexpected
          <br>
                          error
          <br>
          <br>
                          Traceback (most recent call last):
          <br>
                             File "/usr/share/vdsm/storage/task.py",
          line 840,
          <br>
                          in _run
          <br>
                               return fn(*args, **kargs)
          <br>
                             File "/usr/share/vdsm/logUtils.py", line
          42, in
          <br>
                          wrapper
          <br>
                               res = f(*args, **kwargs)
          <br>
                             File "/usr/share/vdsm/storage/hsm.py", line
          923,
          <br>
                          in connectStoragePool
          <br>
                               masterVersion, options)
          <br>
                             File "/usr/share/vdsm/storage/hsm.py", line
          970,
          <br>
                          in _connectStoragePool
          <br>
                               res = pool.connect(hostID, scsiKey,
          msdUUID,
          <br>
                          masterVersion)
          <br>
                             File "/usr/share/vdsm/storage/sp.py", line
          643, in
          <br>
                          connect
          <br>
                               self.__rebuild(msdUUID=msdUUID,
          <br>
                          masterVersion=masterVersion)
          <br>
                             File "/usr/share/vdsm/storage/sp.py", line
          1167,
          <br>
                          in __rebuild
          <br>
                               self.masterDomain =
          <br>
                          self.getMasterDomain(msdUUID=msdUUID,
          <br>
                          masterVersion=masterVersion)
          <br>
                             File "/usr/share/vdsm/storage/sp.py", line
          1506,
          <br>
                          in getMasterDomain
          <br>
                               raise
          se.StoragePoolMasterNotFound(self.spUUID,
          <br>
                          msdUUID)
          <br>
                          StoragePoolMasterNotFound: Cannot find master
          domain:
          <br>
                          'spUUID=f90a0d1c-06ca-11e2-a05b-00151712f280,
          <br>
                          msdUUID=67534cca-1327-462a-b455-a04464084b31'
          <br>
                          Thread-18::DEBUG::2013-01-23
          <br>
                         
          15:50:42,887::task::852::TaskManager.Task::(_run)
          <br>
                         
          Task=`08709e68-bcbc-40d8-843a-d69d4df40ac6`::Task._run:
          <br>
                          08709e68-bcbc-40d8-843a-d69d4df40ac6
          <br>
                          ('f90a0d1c-06ca-11e2-a05b-00151712f280', 2,
          <br>
                          'f90a0d1c-06ca-11e2-a05b-00151712f280',
          <br>
                          '67534cca-1327-462a-b455-a04464084b31', 433)
          {}
          <br>
                          failed - stopping task
          <br>
          <br>
                          This is with vdsm built from
          <br>
                          commit25a2d8572ad32352227c98a86631300fbd6523c1
          <br>
                          - DHC
          <br>
          <br>
          <br>
                          On Wed, Jan 23, 2013 at 10:44 AM, Dead Horse
          &lt;
          <br>
                          <a class="moz-txt-link-abbreviated" href="mailto:deadhorseconsulting@gmail.com">deadhorseconsulting@gmail.com</a>
          <br>
                         
          <a class="moz-txt-link-rfc2396E" href="mailto:deadhorseconsulting@gmail.com">&lt;mailto:deadhorseconsulting@gmail.com&gt;</a>&gt; wrote:
          <br>
          <br>
                              VDSM was built from:
          <br>
                              commit
          166138e37e75767b32227746bb671b1dab9cdd5e
          <br>
          <br>
                              Attached is the full vdsm log
          <br>
          <br>
                              I should also note that from engine
          perspective
          <br>
                              it sees the master
          <br>
                              storage domain as locked and the others as
          unknown.
          <br>
          <br>
          <br>
                              On Wed, Jan 23, 2013 at 2:49 AM, Dan
          Kenigsberg
          <br>
                              &lt;<a class="moz-txt-link-abbreviated" href="mailto:danken@redhat.com">danken@redhat.com</a>
          <a class="moz-txt-link-rfc2396E" href="mailto:danken@redhat.com">&lt;mailto:danken@redhat.com&gt;</a>&gt;wrote:
          <br>
          <br>
                                  On Tue, Jan 22, 2013 at 04:02:24PM
          -0600,
          <br>
                                  Dead Horse wrote:
          <br>
          <br>
                                      Any ideas on this one? (from VDSM
          log):
          <br>
                                      Thread-25::DEBUG::2013-01-22
          <br>
                                     
          15:35:29,065::BindingXMLRPC::914::vds::(wrapper)
          <br>
                                      client
          <br>
          <br>
                                  [3.57.111.30]::call
          <br>
          <br>
                                      getCapabilities with () {}
          <br>
                                      Thread-25::ERROR::2013-01-22
          <br>
                                     
          15:35:29,113::netinfo::159::root::(speed)
          <br>
                                      cannot read ib0 speed
          <br>
                                      Traceback (most recent call last):
          <br>
                                         File
          <br>
                                     
          "/usr/lib64/python2.6/site-packages/vdsm/netinfo.py",
          <br>
                                      line 155,
          <br>
          <br>
                                  in
          <br>
          <br>
                                      speed
          <br>
                                           s =
          <br>
                                      int(file('/sys/class/net/%s/speed'
          %
          <br>
                                      dev).read())
          <br>
                                      IOError: [Errno 22] Invalid
          argument
          <br>
          <br>
                                      Causes VDSM to fail to attach
          storage
          <br>
          <br>
          <br>
                                  I doubt that this is the cause of the
          <br>
                                  failure, as vdsm has always
          <br>
                                  reported "0" for ib devices, and still
          is.
          <br>
          <br>
                  it happens only when you call to getCapabilities.. so
          it
          <br>
                  doesn't related to the flow, and it can't effect the
          storage.
          <br>
                  Dan: I guess this is not the issue but why is the
          IOError?
          <br>
          <br>
          <br>
                                  Does a former version works with your
          Engine?
          <br>
                                  Could you share more of your vdsm.log?
          I
          <br>
                                  suppose the culprit lies in one
          <br>
                                  one of the storage-related commands,
          not in
          <br>
                                  statistics retrieval.
          <br>
          <br>
          <br>
                                      Engine side sees:
          <br>
                                      ERROR
          <br>
                                     
          [org.ovirt.engine.core.bll.storage.NFSStorageHelper]
          <br>
                                      (QuartzScheduler_Worker-96)
          [553ef26e]
          <br>
                                      The connection with details
          <br>
                                      192.168.0.1:/ovirt/ds failed
          because of
          <br>
                                      error code 100 and error
          <br>
          <br>
                                  message
          <br>
          <br>
                                      is: general exception
          <br>
                                      2013-01-22 15:35:30,160 INFO
          <br>
                                     
          [org.ovirt.engine.core.bll.SetNonOperationalVdsCommand]
          <br>
                                      (QuartzScheduler_Worker-96)
          [1ab78378]
          <br>
                                      Running command:
          <br>
                                      SetNonOperationalVdsCommand
          internal:
          <br>
                                      true. Entities affected :  ID:
          <br>
                                     
          8970b3fe-1faf-11e2-bc1f-00151712f280
          <br>
                                      Type: VDS
          <br>
                                      2013-01-22 15:35:30,200 INFO
          <br>
                                     
          [org.ovirt.engine.core.vdsbroker.SetVdsStatusVDSCommand]
          <br>
                                      (QuartzScheduler_Worker-96)
          [1ab78378] START,
          <br>
                                      SetVdsStatusVDSCommand(HostName =
          kezan,
          <br>
                                      HostId =
          <br>
                                     
          8970b3fe-1faf-11e2-bc1f-00151712f280,
          <br>
                                      status=NonOperational,
          <br>
                                     
          nonOperationalReason=STORAGE_DOMAIN_UNREACHABLE),
          <br>
                                      log id: 4af5c4cd
          <br>
                                      2013-01-22 15:35:30,211 INFO
          <br>
                                     
          [org.ovirt.engine.core.vdsbroker.SetVdsStatusVDSCommand]
          <br>
                                      (QuartzScheduler_Worker-96)
          [1ab78378]
          <br>
                                      FINISH, SetVdsStatusVDSCommand,
          <br>
          <br>
                                  log
          <br>
          <br>
                                      id: 4af5c4cd
          <br>
                                      2013-01-22 15:35:30,242 ERROR
          <br>
                                     
          [org.ovirt.engine.core.dal.dbbroker.auditloghandling.AuditLogDirector]
          <br>
                                      (QuartzScheduler_Worker-96)
          [1ab78378]
          <br>
                                      Try to add duplicate audit log
          <br>
                                      values with the same name. Type:
          <br>
                                      VDS_SET_NONOPERATIONAL_DOMAIN.
          Value:
          <br>
                                      storagepoolname
          <br>
          <br>
                                      Engine = latest master
          <br>
                                      VDSM = latest master
          <br>
          <br>
          <br>
                                  Since "latest master" is an unstable
          <br>
                                  reference by definition, I'm sure
          <br>
                                  that History would thank you if you
          post the
          <br>
                                  exact version (git hash?)
          <br>
                                  of the code.
          <br>
          <br>
                                      node = el6
          <br>
          <br>
          <br>
          <br>
          <br>
          <br>
          <br>
          <br>
          <br>
                      _______________________________________________
          <br>
                      Users mailing list
          <br>
                      <a class="moz-txt-link-abbreviated" href="mailto:Users@ovirt.org">Users@ovirt.org</a> <a class="moz-txt-link-rfc2396E" href="mailto:Users@ovirt.org">&lt;mailto:Users@ovirt.org&gt;</a>
          <br>
                      <a class="moz-txt-link-freetext" href="http://lists.ovirt.org/mailman/listinfo/users">http://lists.ovirt.org/mailman/listinfo/users</a>
          <br>
          <br>
          <br>
          <br>
                  --
          <br>
                  Yaniv Bronhaim.
          <br>
                  RedHat, Israel
          <br>
                  09-7692289
          <br>
                  054-7744187
          <br>
          <br>
          <br>
          <br>
        </blockquote>
        <br>
        <br>
        <br>
        <br>
        _______________________________________________
        <br>
        Users mailing list
        <br>
        <a class="moz-txt-link-abbreviated" href="mailto:Users@ovirt.org">Users@ovirt.org</a>
        <br>
        <a class="moz-txt-link-freetext" href="http://lists.ovirt.org/mailman/listinfo/users">http://lists.ovirt.org/mailman/listinfo/users</a>
        <br>
      </blockquote>
      <br>
      <br>
    </blockquote>
    <br>
    <br>
  </body>
</html>