Disable Fields from Indexing.

When any search engine has to index large blocks of data, the process is more CPU intensive than smaller blocks of data. So if you have objects with large amounts of text that are not important for search results, you can provide a custom mapping and omit those fields from indexing using the setting enabled. For more detailed information, see Elastic Search Guide, Object Type.

Add Elasticsearch Nodes.

If your Couchbase Server cluster experiences a backlog of items in the replication queue, you may want to consider adding additional Elasticsearch nodes. This should increase how quickly items can be indexed by the search engine.

Adjusting Concurrent Replication

If you are running your Couchbase cluster and Elasticsearch cluster on hardware with high-performance CPUs, you can increase this setting to improve replication speed between the two clusters. In the case of Elasticsearch, there are also scenarios where you may want to decrease this setting from the default so you do not overwhelm an Elasticsearch node. This is discussed below; for more information about this XDCR parameter, see Couchbase Server Manual, Changing Internal XDCR Settings.

One of the key parameters you can use to adjust XDCR performance is xdcrMaxConcurrentReps; this will increase or decrease the maximum concurrent replication by a Couchbase node. The default number of concurrent replications via XDCR is 32. For instance, if you have five nodes in a Couchbase cluster and you have one Elasticsearch node, Couchbase Server can generate up to 160 concurrent replications targeting the single Elasticsearch node. Each replication can require multiple TCP connections and both the concurrent replications and the number of connections may overwhelm the Elasticsearch node. If this does occur you can see the following types of errors in Couchbase Web Console | XDCR | Ongoing XDCR section:

Error replicating vbucket 7:
{badmatch, {error,all_nodes_failed,
<<"Failed to grab remote bucket info from any of known nodes">>}}

Error replicating vbucket 7:
{error,{error,timeout}}}

This means that Couchbase Server cannot communicate with Elasticsearch in the time that it expects. Couchbase Server can recover from these types of errors and retry replication, however your replication may take longer to complete or operate with higher latency because the operations must be later retried. If you encounter this scenario, you should lower the default xdcrMaxConcurrentReps setting to 8 or less so that the total number of concurrent replications can be handled by your Elasticsearch node. For more information on changing this setting, see Couchbase Server Manual, Changing Internal XDCR Settings.