More on my HBase problem here just from looking at the logs: The regionservers aren't able to connect to the master because what's being stored in the ZNode /hbase/master is "127.0.1.1" where I want it to be the IP address of the machine. How can I set the master? That is probably the root of all of my problems.
-----Original Message----- From: Mark Vigeant [mailto:[email protected]] Sent: Monday, November 23, 2009 12:17 PM To: [email protected] Subject: RE: Zookeeper Error: So the master log indicates that in fact the master and the regions don't know about eachother, and I think I know why. The username on all my machines right now is hadoop (just for simplicity) and in all of my /etc/hosts files the default list was 127.0.0.1 localhost 127.0.1.1 hadoop Which was confusing to me (I just set up these machines with Karmic) because I hadn't seen that before. Now, as I said before when I try to run "list" in the hbase shell what returns is a ZooKeeper error. But because it filled up the terminal with error messages so quickly, I had missed this line that is printed immediately: "DEBUG zookeeper.ZooKeeperWrapper: Read ZNode /hbase/master got 127.0.1.1:60000" Which indicates that for some reason my master is running on "hadoop" when all the regionservers and configuration files think it's "fry." I don't know how to fiz this, but I'm sure this is the main problem. http://pastebin.com/m142bd3d5 here's the master log in its entirety. -----Original Message----- From: [email protected] [mailto:[email protected]] On Behalf Of stack Sent: Monday, November 23, 2009 12:06 PM To: [email protected] Subject: Re: Zookeeper Error: On Mon, Nov 23, 2009 at 7:50 AM, Mark Vigeant <[email protected]>wrote: > > I just started HBase up again and I got an IOException: NoNode for > /hbase/master but then it started up anyways. However, when I run the shell > and type "list" it gives me a "NoServerForRegionException: Timed out trying > to locate root region" and then a huge trace. It then tries to debug and > sleeps for intervals of 2000 ms. So it seems to me like the zookeeper tree > doesn't have the information it's supposed to: am I right here? Is there a > way to format and restart everything here/ will that help? > > Another interesting tidbit is that when I run "bin/stop-hbase" from my > master, the HMaster gets shut down, but none of the regionservers do. The nonode for /hbase/master can be ignored usually. If you look at your log, it succeeds eventually IIRC. The NoServerForRegionException would seem to indicate no regionservers are connecting to your master. Pastebin your master log so we can take a look-see. St.Ack This email message and any attachments are for the sole use of the intended recipients and may contain proprietary and/or confidential information which may be privileged or otherwise protected from disclosure. Any unauthorized review, use, disclosure or distribution is prohibited. If you are not an intended recipient, please contact the sender by reply email and destroy the original message and any copies of the message as well as any attachments to the original message. This email message and any attachments are for the sole use of the intended recipients and may contain proprietary and/or confidential information which may be privileged or otherwise protected from disclosure. Any unauthorized review, use, disclosure or distribution is prohibited. If you are not an intended recipient, please contact the sender by reply email and destroy the original message and any copies of the message as well as any attachments to the original message.
