Hello, anyone.
I used to setup some non-data nodes to accept bulk indexing from my
logstash/outputs/elasticsearch(http protocol). But some day I found one
article suggested use client node for searching but data node for bulk
indexing.
So, what's the best practical? Which node should we send bulk to,
It does not matter, you can send search and bulk requests to all nodes. ES
will do the routing and automatically forward the requests to the nodes
where they can be executed.
Hello, anyone.
I used to setup some non-data nodes to accept bulk indexing from my
logstash/outputs/elasticsearch(http protocol). But some day I found one
article suggested use client node for searching but data node for bulk
indexing.
So, what's the best practical? Which node should we send bulk to,
It’s completely true that you can send requests to any node, however in practice one does not typically send requests directly to master nodes. It’s best to leave the master nodes alone so they can focus on maintaining global cluster state.
You can safely send bulk requests to either client or data nodes. The only difference is that the node that receives the request will sub-divide the bulk requests into per-node sub-bulk-requests and forward them on, then receive the results before sending back to the caller. In practice many people prefer to create a pool of client nodes and round-robin requests amongst them, never directly talking to the data nodes. Your mileage may vary, but this is a good rule of thumb and I would recommend following that general practice.
It does not matter, you can send search and bulk requests to all nodes. ES will do the routing and automatically forward the requests to the nodes where they can be executed.
Jörg
On Mon, Mar 16, 2015 at 4:40 PM, chenlin rao <rao.chenlin@gmail.commailto:rao.chenlin@gmail.com> wrote:
Hello, anyone.
I used to setup some non-data nodes to accept bulk indexing from my logstash/outputs/elasticsearch(http protocol). But some day I found one article suggested use client node for searching but data node for bulk indexing.
So, what's the best practical? Which node should we send bulk to, master, data, client?
thanks~
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.