Now we are using Elasticsearch 1.3.7 (3 TB data) with one index and many types (more than 250 ). we are facing mapping conflicts as big blow for upgradation.
so we planning to change our ES architecture as creating many index (more than 250 indices) with one type in each one index (which recommended in further versions).
While testing these in preview stage servers, we are facing more issues like high openfiles count , server load, Hgh heap utilization (resource utilization).
Please suggest steps can we do in configuration to overcome these issues (HIGH openfiles issues, resource utilization)
Please don't say upgrade, because we are working hard for it.