It seems like there's another master still running somewhere. If
that's not correct, a simple fix is to restart zookeeper when you
restart Accumulo (stop Accumulo, restart ZK, start Accumulo).
If it happens again, it might be worth checking the change
logs/release notes to see if this might be caused by one of the bugs
fixed by ZK 3.4.4 or 3.4.5, or Accumulo 1.4.3.
Christopher L Tubbs II
On Sat, Jun 29, 2013 at 11:25 AM, Thomas Jackson <[EMAIL PROTECTED]> wrote:
> I am trying to relocate the Accumulo Master process to another server. I
> have already brought down the Accumulo cluster (restarted zookeepers as
> well) and have updated all references that pointed to the original master
> server to point to the new master server.
> However, after restarting Accumulo the master server is not starting. In
> the master log I see the following error:
> INFO: trying to get master lock
> INFO: Failed to get master lock, even after waiting for session timeout,
> becoming back-up server
> I'm hoping that someone has encountered this before and it's simple tweak I
> I'm running Accumulo 1.4.2, Zookeeper 3.4.3 Hadoop 1.0.3