Cluster vs single node


(Geoff) #1

Hi,

First I will apologize for most likely a naive question. I am currently building out our elasticsearch implementation from a single node, to a 3 node cluster. I have followed most of the tutorials I have found on the internet, but it seems my cluster is slower by up to 3x versus the single node we are currently using. When I do a search on our cluster, I see all the drives get pegged to 100% utilization, and the iowait gets up to around 70 on each node. When I do the same query on the single node, I don't even see a hit to the disk, and very little cpu. Here is our current setup in AWS:

single node:

m3.large
1 750GB 200 PIOPS drive
elasticserch 0.90.7

Cluster:

3 c3.xlarge
1 200GB 400 PIOPS drive
elasticsearch 1.2.2

The data for the cluster is spread across all 3 nodes, and I mean each node has the complete set of the data. The single node is using all default configuration, but the cluster nodes are just using these added configs:

'cluster.routing.allocation.node_initial_primaries_recoveries' => '4'
'cluster.routing.allocation.node_concurrent_recoveries' => '2'
'indices.recovery.max_bytes_per_sec' => '20mb'
'indices.recovery.concurrent_streams' => '5'
'index.number_of_replicas' => '2'

Are there any other configuration options I should add to get better search performance?

Thank you for your time.

Geoff


(Mark Walkom) #2

What sort of data are you searching and what sort of searches are they
against it?

Regards,
Mark Walkom

Infrastructure Engineer
Campaign Monitor
email: markw@campaignmonitor.com
web: www.campaignmonitor.com

On 7 August 2014 04:20, Geoff geoffreyowen@gmail.com wrote:

Hi,

First I will apologize for most likely a naive question. I am currently
building out our elasticsearch implementation from a single node, to a 3
node cluster. I have followed most of the tutorials I have found on the
internet, but it seems my cluster is slower by up to 3x versus the single
node we are currently using. When I do a search on our cluster, I see all
the drives get pegged to 100% utilization, and the iowait gets up to around
70 on each node. When I do the same query on the single node, I don't even
see a hit to the disk, and very little cpu. Here is our current setup in
AWS:

single node:

m3.large
1 750GB 200 PIOPS drive
elasticserch 0.90.7

Cluster:

3 c3.xlarge
1 200GB 400 PIOPS drive
elasticsearch 1.2.2

The data for the cluster is spread across all 3 nodes, and I mean each node
has the complete set of the data. The single node is using all default
configuration, but the cluster nodes are just using these added configs:

'cluster.routing.allocation.node_initial_primaries_recoveries' => '4'
'cluster.routing.allocation.node_concurrent_recoveries' => '2'
'indices.recovery.max_bytes_per_sec' => '20mb'
'indices.recovery.concurrent_streams' => '5'
'index.number_of_replicas' => '2'

Are there any other configuration options I should add to get better search
performance?

Thank you for your time.

Geoff

--
View this message in context:
http://elasticsearch-users.115913.n3.nabble.com/Cluster-vs-single-node-tp4061375.html
Sent from the ElasticSearch Users mailing list archive at Nabble.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 elasticsearch+unsubscribe@googlegroups.com.
To view this discussion on the web visit
https://groups.google.com/d/msgid/elasticsearch/1407349203339-4061375.post%40n3.nabble.com
.
For more options, visit https://groups.google.com/d/optout.

--
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 elasticsearch+unsubscribe@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/elasticsearch/CAEM624a3Ef%2B7Ax9anUGaDdr1aZKGsRxccsJBiJskECWUwWJcHA%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.


(system) #3