I myself prefer TransportClient as it has absolutely no impact on the cluster
state.
A new node joining the cluster generates a cluster state update which is
propagated to all nodes.
I prefer having a very stable cluster state than continuous updates. Think
about a Java batch which starts every
Thanks David once again. Can you please help in deciding which one should I
go for NodeClient or TransportClient for production deployments.
>From the elastic search documentation what I understand is that Node client
eliminates the double hop that is present in Transport Client. Shall I opt
fo
If you don’t want to define transport addresses, you’d better use a NodeClient.
A NodeClient is actually a node running inside the cluster (with auto
discovery…) but without any data.
You just need to provide the cluster name with a NodeClient. See
http://www.elasticsearch.org/guide/en/elasticse
I am trying to connect remotely to elastic search cluster having 2 nodes
My question is I am able to query successfully only when we configure both
the cluster name and host name of the individual nodes? Why it doesn't work
if I only configure the cluster name in my client code
Settings settings
