How will 'failed to merge' hurt us?


(arta) #1

Hi,
I've been seeing 'failed to merge' errors both on the primary shard node and the secondary shard node.
(I have 1 replica)
The error is logged like every 1 to 30 sec, and always the same index name (i36) and shard # (21).

[2012-09-28 00:01:31,073][WARN ][index.merge.scheduler ] [Titanium Man] [i36][21] failed to merge

My first question is whether we can go with it, or it will hurt us, or even, whether it is hurting us in some way.

Further questions are:
Is there any way to further investigate what is wrong?
Is there any way to fix this other than recreate index from scratch?
Is there any way to know which routing value is mapped to this particular shard?
(I'm using explicit routing)

Thanks for your help.
I love this users group as members are very active and helpful!


(arta) #2

The failure reason in the log:

[2012-09-28 00:01:31,073][WARN ][index.merge.scheduler ] [Titanium Man] [i36][21] failed to merge
java.lang.ArrayIndexOutOfBoundsException


(system) #3