lgostash-bench is a utility that performance tests logstash filters. This is primarily used to aid testing filters before they are pushed to production.
- Obtain a sample of the ingest data.
- There are many ways to obtain that data but the simplest is to remove any filters and set the output of logstash to a file and let it run.
- The larger the sample file the better the benchmark data will be.
- If possible locate the file on a memory backed location such as /dev/shm
- Create a temporary directory to use.
- The temporary directory will store the output data as well as the sincedb file.
- Recommend placing it in a location like /dev/shm or another memory backed disk.
- Select your filter file
- The filter file should be only the filter {} section.
- Run!
logstash-bench -f file.conf -s source.logs
| long name | short name | description | default |
|---|---|---|---|
| input-codec | c | The codec that logstash should use for input. | json |
| source-file-path | s | The path to the source log file. | ./input.log |
| temp-dir-path | t | The path to the temporary directory. | /dev/shm/ |
| number-of-iterations | i | The number of times the test will run. | 1 |
| logstash-executable-path | l | The path to the logstash executable. | /usr/share/logstash/bin/logstash |
| filter-file-path | f | The path to the filter file to test. | ./filter.conf |
| details | d | Show the details of the process via stdout. | false |