We have a cluster with more then 1 TB data/100 indices and when we migrated
to a newer version and restarted, the cluster went to yellow state and I
understand it is perfectly normal. I am sure that when in Yellow, all the
search operation can happen with only some performance delay. The question
is, can we also keep on indexing new documents when the cluster is yellow.
Will it slow down the recovery process or it will stop the recovery process
or its not an issue at all? Any answer would be much appreciated.
My experience has been that we usually can index in the yellow state,
though not always at the same rate. I usually only try to index data if
there aren't any relocating/initializing shards in the index I'm trying to
put docs into. That said, I'd imagine that running indexing during recovery
is bound to slow down the process a certain amount, though it's hard to say
how much. :-/
On Wednesday, December 5, 2012 11:20:36 PM UTC-8, Kavin wrote:
Hi all,
We have a cluster with more then 1 TB data/100 indices and when we
migrated to a newer version and restarted, the cluster went to yellow state
and I understand it is perfectly normal. I am sure that when in Yellow, all
the search operation can happen with only some performance delay. The
question is, can we also keep on indexing new documents when the cluster is
yellow. Will it slow down the recovery process or it will stop the recovery
process or its not an issue at all? Any answer would be much appreciated.
Apache, Apache Lucene, Apache Hadoop, Hadoop, HDFS and the yellow elephant
logo are trademarks of the
Apache Software Foundation
in the United States and/or other countries.