We have an Elasticsearch instance, containing 100s of indices in it with different, statically-defined mappings. Is there a way to prioritize some of them for read/write operations not to be affected by low-priority ones?
i.e. we want the high-priority indices to have a high performance in search with keeping them in the memory and not to be thrown out of the memory by the low-priority indices
No, that is not possible within an instance. You could however put search-heavy indices on specific nodes (that hold less data) using shard allocation filtering.
1 Like
You can create a dedicated Cluster for searching and leverage CCR from you current cluster to filter only high value index there
1 Like
This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.