How to get large response to query fast?

I use AWS Elasticsearch. I have stored ~40 Mil documents and I want to fetch large amount of query results (e.g. one milion most relevant document). I am fetching document titles only but searching is executed through docs contents.
I have problem that fetching is quite slow and its speed is decreasing. I use parallel fetching using sliced scroll api.

How to make elastic to fetch results faster?

More nodes, likely :slight_smile:

If you're using sliced scrolling, you're using the best/fastest way to retrieve documents in bulk from Elasticsearch. At this point, you're likely bottlenecked on how quickly documents can physically be retrieved from disk.

The only real option is to add more nodes so there are more disk IO to service requests, or upgrade the speed of your IO.

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