Can you give detailed information about documents.json?

What is the purpose of the documents.json file when creating custom track?

What is the difference between documents.json with 1gb and documents.json with 10gb when running the benchmark?


please see the Rally docs for full tutorial how to create custom tracks. The file you're mentioning contains the documents that you want to bulk-index to Elasticsearch. It is the document corpus you're operating on in this benchmark and different index sizes cause all sorts of different system behavior, e.g. with larger indices, querying might take longer, page cache behavior might be different (causing less or more load on I/O), you will see different background merge activity in Elasticsearch etc. etc. .

For a general introduction of what to take care of when benchmarking (also from a methodological perspective) I recommend to read the blog post Seven Tips for Better Elasticsearch Benchmarks.


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