Hi,
I applied the suggested timeout changes and it lasted longer before it started
failing, but its now just throwing the same errors on "procedure.Subprocedure:
Subprocedure pool is full!"
Looking through the logs, it looks like the nodes may not correctly updating
they completion state after
Hi,
I got a change finally to apply the timeout changes and it lasted longer before
it started failing, but its now just throwing the same errors on
"procedure.Subprocedure: Subprocedure pool is full!"
Looking through the logs, it looks like the nodes may not correctly updating
they completion
Hi,
I got a change finally to apply the timeout changes and it lasted longer before
it started failing, but its now just throwing the same errors on
"procedure.Subprocedure: Subprocedure pool is full!"
Looking through the logs, it looks like the nodes may not correctly updating
they completion
Here're the config parameters related to controlling snapshot timeout:
hbase.snapshot.master.timeoutMillis
60
This is the time HBase master waits for the snapshot operation to
complete.
Do not confuse this hbase.snapshot.master.timeout.millis, which
although
Hi,
I'm running snapshots on 7 tables every hour, the last run managed 15 rounds of
snapshots before starting to fail.
I did notice the following while getting the success count
2014-09-10 11:01:55,453 DEBUG
[rs(##,16020,1410126483043)-snapshot-pool151-thread-1]
snapshot.FlushSnap
Subprocedure pool is full.
How many snapshot requests did you submit before seeing the following ?
Cheers
On Sep 23, 2014, at 2:28 AM, Ian Brooks wrote:
> Hi,
>
> I'm seeing an issue on our hbase cluster which is preventing snapshots from
> working. So far the only way i can get it working
Hi,
I'm seeing an issue on our hbase cluster which is preventing snapshots from
working. So far the only way i can get it working again is to restart all the
regionservers which is not ideal.
I'm running hbase 0.98.3 on hadoop 2.4.0, the stacktraces are
>From host given the snapshot command vi
