Data stream with hot/warm/cold tiers read performance

Hi all,

I have a question regarding the query/read performance on data steam.

As the document shows, for read requests data stream will route them to all its backing indices.

If I use hot/warm/cold tiers with data steam and there would be backing indices residing on warm/cold tier nodes, then for query on a specific data stream, the backing indices on warm/cold tiers will get the same frequency of read requests as hot tier, it looks to me this will impact the query performance because in warm/cold tier we usually use less powerful machine?

Please correct me if I am wrong, thanks.

Nodes can quite efficiently determine whether a shard may hold relevant data or not based on the timestamp, so if you are only querying newer data the load on warm and cold nodes should be low even though they are queried.

Thanks, that makes sense.

This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.