<html><head><style type='text/css'>p { margin: 0; }</style></head><body><div style='font-family: arial,helvetica,sans-serif; font-size: 12pt; color: #000000'>Sorry for the delay -- here's what the volume looks like. It's pretty basic:<div><br></div><div><div>volume elkguests0-server</div><div> type protocol/server</div><div> option rpc-auth-allow-insecure on</div><div> option auth.addr./mnt/bulk.allow *</div><div> option auth.login.39da7ae0-8cee-4152-a612-674c48da544e.password 62a8ff81-cd3d-4872-a9b9-bad5da242f10</div><div> option auth.login./mnt/bulk.allow 39da7ae0-8cee-4152-a612-674c48da544e</div><div> option transport-type tcp</div><div> subvolumes /mnt/bulk</div><div>end-volume</div></div><div><br></div><div><div>type=2</div><div>count=6</div><div>status=1</div><div>sub_count=2</div><div>stripe_count=1</div><div>replica_count=2</div><div>version=12</div><div>transport-type=0</div><div>volume-id=7c50ebe2-337b-4bed-a53b-5fa706150b26</div><div>username=39da7ae0-8cee-4152-a612-674c48da544e</div><div>password=62a8ff81-cd3d-4872-a9b9-bad5da242f10</div><div>op-version=2</div><div>client-op-version=2</div><div>server.allow-insecure=on</div><div>nfs.disable=0</div><div>brick-0=fs-elkpo01.suntradingllc.com:-mnt-bulk</div><div>brick-1=fs-elkpo02.suntradingllc.com:-mnt-bulk</div><div>brick-2=fs-elkpo03.suntradingllc.com:-mnt-bulk</div><div>brick-3=fs-elkpo04.suntradingllc.com:-mnt-bulk</div><div>brick-4=fs-elkpo05.suntradingllc.com:-mnt-bulk</div><div>brick-5=fs-elkpo06.suntradingllc.com:-mnt-bulk</div><div><br></div><div>I'll try and dig up some client logs, but the last event was over a month ago.</div><br><hr id="zwchr"><div style="color:#000;font-weight:normal;font-style:normal;text-decoration:none;font-family:Helvetica,Arial,sans-serif;font-size:12pt;"><b>From: </b>"Vijay Bellur" <vbellur@redhat.com><br><b>To: </b>"Nicholas Majeran" <nmajeran@suntradingllc.com>, gluster-users@gluster.org, "Shyamsundar Ranganathan" <srangana@redhat.com><br><b>Sent: </b>Monday, February 17, 2014 10:16:56 AM<br><b>Subject: </b>Re: [Gluster-users] add-brick and fix-layout takes some VMs offline<br><br>On 02/13/2014 08:52 PM, Nicholas Majeran wrote:<br>> Hi there,<br>><br>> We have a distributed-replicated volume hosting KVM guests running<br>> Gluster 3.4.1.<br>> We've grown from 1 x 2 -> 2 x 2 -> 3 x 2,but each time we've added nodes<br>> or run a fix layout,<br>> some of our guests go offline (or worse with error=continue they<br>> silently error).<br>> After the last addition we didn't even run fix-layout as the guests are<br>> becoming increasingly important.<br><br>Would it be possible to share the client log files and your volume <br>configuration?<br><br>><br>> Those guests are currently are using a combination of FUSE and libgfapi.<br>> Is there a setting or group of settings we should use to ameliorate this<br>> problem?<br>> Is FUSE or libgfapi more forgiving when add-brick or fix-layout is run?<br>><br><br>The behavior of FUSE or libgfapi should mostly be the same with either <br>add-brick or fix-layout.<br><br>Thanks,<br>Vijay<br><br></div><br></div></div></body></html>