I may have figured this out.  The systems that "failed" are running the Oracle "unbreakable" kernel:

3.8.13-98.el6uek.x86_64

The working systems are running the default CentOS 6 2.6 kernel.

and the error from the vdsm.log only show up on the UEK kernel. 

    -- Chris




On Wed, Aug 12, 2015 at 9:34 AM, Chris Liebman <chris.l@taboola.com> wrote:
Hi,
   I'm new to oVirt and recently built a 10 node ovirt 3.5 DC with shared storage using gluster configured as distributed-replicated (replication = 2).  Shortly after 7 of the 10 nodes dropped, one at a time over a few hours, into "Non Operational" state.  Attempting to activate one of these nodes gives the error:  "Failed to connect Host ovirt-node260 to Storage Pool LADC-TBX". Attempting to put the node into Maintenance eaves the node stuck in "Preparing For maintenance".  

When I rebooted one of the nodes I see this in the nodes event list:

"Host ovirt-node269 reports about one of the Active Storage Domains as Problematic."

I see many of these errors in the vdsm log from the failed nodes:

Thread-10000::ERROR::2015-08-12 10:01:17,748::__init__::506::jsonrpc.JsonRpcServer::(_serveRequest) Internal server error

Traceback (most recent call last):

  File "/usr/lib/python2.6/site-packages/yajsonrpc/__init__.py", line 501, in _serveRequest

    res = method(**params)

  File "/usr/share/vdsm/rpc/Bridge.py", line 267, in _dynamicMethod

    result = fn(*methodArgs)

  File "/usr/share/vdsm/API.py", line 1330, in getStats

    stats.update(self._cif.mom.getKsmStats())

  File "/usr/share/vdsm/momIF.py", line 60, in getKsmStats

    stats = self._mom.getStatistics()['host']

  File "/usr/lib/python2.6/site-packages/mom/MOMFuncs.py", line 75, in getStatistics

    host_stats = self.threads['host_monitor'].interrogate().statistics[-1]

AttributeError: 'NoneType' object has no attribute 'statistics'

Any help here is appreciated.

    -- Chris