Hello and thanks for replying. I think this log was console dump actually coming from the Hbase shell rather than Stargate, but in any case I couldn't reconnect via Stargate until I restarted. I cycled Hbase and Stargate so I cannot say which of the two resolved it. I'll keep an eye out though, if I see this happen again I'll forward some decent log.
For the base64 stuff - don't worry about it, but it is not so obvious hidden in the XSD, that's all (I did see it there after I figured out what was going on by reading the usage notes on the old rest connector). If it wasn't obvious to me, it is going to confuse and disconcert others too. Kind regards Robert Gibbon Solutions Architect Integration Design & Solution Engineering [email protected] +32 474 53 52 51 Vodafone Group Service GmbH Mannesmannufer 2, D-40213 Düsseldorf, Germany Amtsgericht Düsseldorf, HRB 53554 Geschäftsführung: Helmut Hoffmann, Dr. Joachim Peters -----Original Message----- From: Lars George [mailto:[email protected]] Sent: Mittwoch, 23. Dezember 2009 16:06 To: [email protected] Subject: Re: Zookeeper session prb Hi Robert, The exceptions you got are indeed timeouts after the client going stale. They are all DEBUG and WARN level and therefore "normal". The client also subsequently reconnects and continues ad expected. You are saying it did not? What were the symptoms and do you have more logs *after* those that you already sent? Btw, the Stargate XML schemas do explain that all fields are indeed base64 encoded. The Stargate docs does include them. Maybe a high level getting started and feature set doc is needed? Lars On Dec 23, 2009, at 15:00, "Gibbon, Robert, VF-Group" <[email protected] > wrote: > Hello, > > I am playing around with Hbase and the Stargate interface; I got a > problem when I came back from lunch which rendered my little test rig > unavailable and I had to restart it. > > I'm running a pseudo cluster on an ubuntu vm, 0.20.2 of Hbase and all > the defaults. I haven't changed anything in limits.conf on the vm but > I think this is a lunch/time triggered issue? Has anybody seen this > one before if so what's the workaround? It looks a bit like it could > connection/file descriptor leak too from my limited experience in > these things. See the dump of log below. > > I have two other minor comments on the excellent SG interface > > - SG base64 encodes all the output which is good, but it'd be great if > it was made a painfully plain stating the obvious fact on the wiki for > boneheads like myself > - I saw that SG is probably going to have some authentication features > and caching added, will it also be possible to explicitly disable > them, because I would rather use a dedicated reverse proxy for that > > 09/12/23 13:41:05 WARN zookeeper.ClientCnxn: Exception closing session > 0x125bafce85d0006 to sun.nio.ch.selectionkeyi...@16d46d7 > java.io.IOException: TIMED OUT > at org.apache.zookeeper.ClientCnxn$SendThread.run(ClientCnxn.java:906) > 09/12/23 13:41:43 DEBUG client.HConnectionManager$ClientZKWatcher: > Got ZooKeeper event, state: Disconnected, type: None, path: null > 09/12/23 13:42:50 WARN zookeeper.ClientCnxn: Exception closing session > 0x125bafce85d0006 to sun.nio.ch.selectionkeyi...@227d4e > java.io.IOException: TIMED OUT > at org.apache.zookeeper.ClientCnxn$SendThread.run(ClientCnxn.java:906) > 09/12/23 13:43:47 WARN zookeeper.ClientCnxn: Exception closing session > 0x125bafce85d0006 to sun.nio.ch.selectionkeyi...@8d8f67 > java.io.IOException: TIMED OUT > at org.apache.zookeeper.ClientCnxn$SendThread.run(ClientCnxn.java:906) > 09/12/23 13:44:39 WARN zookeeper.ClientCnxn: Exception closing session > 0x125bafce85d0006 to sun.nio.ch.selectionkeyi...@1c1eceb > java.io.IOException: TIMED OUT > at org.apache.zookeeper.ClientCnxn$SendThread.run(ClientCnxn.java:906) > 09/12/23 13:44:51 WARN zookeeper.ClientCnxn: Exception closing session > 0x125bafce85d0006 to sun.nio.ch.selectionkeyi...@a96863 > java.io.IOException: Session Expired > at org.apache.zookeeper.ClientCnxn$SendThread.readConnectResult > (ClientCnxn.java:589) > at org.apache.zookeeper.ClientCnxn$SendThread.doIO(ClientCnxn.java: > 709) > at org.apache.zookeeper.ClientCnxn$SendThread.run(ClientCnxn.java:945) > 09/12/23 13:44:51 DEBUG client.HConnectionManager$ClientZKWatcher: > Got ZooKeeper event, state: Expired, type: None, path: null > 09/12/23 13:44:56 DEBUG zookeeper.ZooKeeperWrapper: Closed connection > with ZooKeeper > > > Robert Gibbon > Solutions Architect > Integration Design & Solution Engineering > > [email protected] > +32 474 53 52 51 > > Vodafone Group Service GmbH > Mannesmannufer 2, D-40213 Düsseldorf, Germany Amtsgericht Düsseldorf, > HRB 53554 > Geschäftsführung: Helmut Hoffmann, Dr. Joachim Peters > >
