Im using ElasticHQ to monitor our cluster and I am noticing that our Search
Fetch times are about 5X more then query times, which seems unreasonable
(20ms vs ~4ms)
Im also noticing a 1.44 mb swap space on one of the nodes (the other node
is at 0)
We're using a 2 node cluster, 2 shards 1 replica on EC2 m1.small instances
Our index contains about 27K docs, total size of ~ 60mb (pretty small -
hence the small instance)
Im using ElasticHQ to monitor our cluster and I am noticing that our
Search Fetch times are about 5X more then query times, which seems
unreasonable (20ms vs ~4ms)
Im also noticing a 1.44 mb swap space on one of the nodes (the other node
is at 0)
We're using a 2 node cluster, 2 shards 1 replica on EC2 m1.small instances
Our index contains about 27K docs, total size of ~ 60mb (pretty small -
hence the small instance)
On Wed, Jan 15, 2014 at 4:45 PM, Arik Gaisler <arikg...@gmail.com<javascript:>
wrote:
Im using ElasticHQ to monitor our cluster and I am noticing that our
Search Fetch times are about 5X more then query times, which seems
unreasonable (20ms vs ~4ms)
Im also noticing a 1.44 mb swap space on one of the nodes (the other node
is at 0)
We're using a 2 node cluster, 2 shards 1 replica on EC2 m1.small instances
Our index contains about 27K docs, total size of ~ 60mb (pretty small -
hence the small instance)
I believe this makes sense then - search is entirely in-memory operation
(after a certain warm up), and fetch involves disk IO. Try keep your docs
as small as possible and don't disable _source, but other than that you're
bounded by the speed of your disks.
Im using ElasticHQ to monitor our cluster and I am noticing that our
Search Fetch times are about 5X more then query times, which seems
unreasonable (20ms vs ~4ms)
Im also noticing a 1.44 mb swap space on one of the nodes (the other
node is at 0)
We're using a 2 node cluster, 2 shards 1 replica on EC2 m1.small
instances
Our index contains about 27K docs, total size of ~ 60mb (pretty small -
hence the small instance)
Any suggestions to why?
Arik
--
You received this message because you are subscribed to the Google
Groups "elasticsearch" group.
To unsubscribe from this group and stop receiving emails from it, send
an email to elasticsearc...@googlegroups.com.
Im also seeing that the refresh is pretty high137.4ms
Another factor is that the fetch times are constantly growing - while the
query times are cinctantly the same
Arik
On Wednesday, January 15, 2014 5:33:51 PM UTC+2, Itamar Syn-Hershko wrote:
I believe this makes sense then - search is entirely in-memory operation
(after a certain warm up), and fetch involves disk IO. Try keep your docs
as small as possible and don't disable _source, but other than that you're
bounded by the speed of your disks.
Im using ElasticHQ to monitor our cluster and I am noticing that our
Search Fetch times are about 5X more then query times, which seems
unreasonable (20ms vs ~4ms)
Im also noticing a 1.44 mb swap space on one of the nodes (the other
node is at 0)
We're using a 2 node cluster, 2 shards 1 replica on EC2 m1.small
instances
Our index contains about 27K docs, total size of ~ 60mb (pretty small -
hence the small instance)
Any suggestions to why?
Arik
--
You received this message because you are subscribed to the Google
Groups "elasticsearch" group.
To unsubscribe from this group and stop receiving emails from it, send
an email to elasticsearc...@googlegroups.com.
If such latency worries you, use in-memory structures to pull the original
documents based on the IDs of the results. Your dataset is small enough to
do that, really. But I suspect this is premature optimization that you are
trying to do.
Im also seeing that the refresh is pretty high137.4ms
Another factor is that the fetch times are constantly growing - while the
query times are cinctantly the same
Arik
On Wednesday, January 15, 2014 5:33:51 PM UTC+2, Itamar Syn-Hershko wrote:
I believe this makes sense then - search is entirely in-memory operation
(after a certain warm up), and fetch involves disk IO. Try keep your docs
as small as possible and don't disable _source, but other than that you're
bounded by the speed of your disks.
Im using ElasticHQ to monitor our cluster and I am noticing that our
Search Fetch times are about 5X more then query times, which seems
unreasonable (20ms vs ~4ms)
Im also noticing a 1.44 mb swap space on one of the nodes (the other
node is at 0)
We're using a 2 node cluster, 2 shards 1 replica on EC2 m1.small
instances
Our index contains about 27K docs, total size of ~ 60mb (pretty small
hence the small instance)
Any suggestions to why?
Arik
--
You received this message because you are subscribed to the Google
Groups "elasticsearch" group.
To unsubscribe from this group and stop receiving emails from it, send
an email to elasticsearc...@googlegroups.com.
--
You received this message because you are subscribed to the Google
Groups "elasticsearch" group.
To unsubscribe from this group and stop receiving emails from it, send
an email to elasticsearc...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/
msgid/elasticsearch/fc28fd75-67c9-46f2-99b2-03cb48bac34b% 40googlegroups.com.
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.