c33f85bc37
With this commit we add a benchmark for the new REST client and the existing transport client. Closes #19281 |
||
|---|---|---|
| .. | ||
| src/main | ||
| README.md | ||
| build.gradle | ||
README.md
Steps to execute the benchmark:
- Start Elasticsearch on the target host (ideally not on the same machine)
- Create an empty index with the mapping you want to benchmark
- Start either the RestClientBenchmark class or the TransportClientBenchmark
- Delete the index
- Repeat steps 2. - 4. for multiple iterations. The first iterations are intended as warmup for Elasticsearch itself. Always start the same benchmark in step 3!
- After the benchmark: Shutdown Elasticsearch and delete the data directory
Repeat all steps above for the other benchmark candidate.
Example benchmark:
- Download benchmark data from http://benchmarks.elastic.co/corpora/geonames/documents.json.bz2 and decompress
- Use the mapping file https://github.com/elastic/rally-tracks/blob/master/geonames/mappings.json to create the index
Example command line parameter list:
192.168.2.2 /home/your_user_name/.rally/benchmarks/data/geonames/documents.json geonames type 8647880 5000 "{ \"query\": { \"match_phrase\": { \"name\": \"Sankt Georgen\" } } }\""
The parameters are in order:
- Benchmark target host IP (the host where Elasticsearch is running)
- full path to the file that should be bulk indexed
- name of the index
- name of the (sole) type in the index
- number of documents in the file
- bulk size
- a search request body (remember to escape double quotes). The
TransportClientBenchmarkusesQueryBuilders.wrapperQuery()internally which automatically adds a root keyquery, so it must not be present in the command line parameter.
You should also define a few GC-related settings -Xms4096M -Xmx4096M -XX:+UseConcMarkSweepGC -verbose:gc -XX:+PrintGCDetails and keep an eye on GC activity. You can also define -XX:+PrintCompilation to see JIT activity.