You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
To ensure that users are using the correct python versions, install the repository with `python3 -m pip install -e .` and run `which opensearch-benchmark` to get the path. Pre-append this path to each of the three commands above and re-run them in the command line.
@@ -46,12 +46,12 @@ Keep in mind the file path outputted differs for each operating system and might
46
46
47
47
- For example: When running `which opensearch-benchmark` on an Ubuntu environment, the commad line outputs `/home/ubuntu/.pyenv/shims/opensearch-benchmark`. On closer inspection, the path points to a shell script. Thus, to invoke OpenSearch Benchmark, pre-=append the OpenSearch Benchmark command with `bash` and the path outputted earlier:
- Another example: When running `which opensearch-benchmark` on an Amazon Linux 2 environment, the command line outputs `~/.local/bin/opensearch-benchmark`. On closer inspection, the path points to a Python script. Thus, to invoke OpenSearch Benchmark, pre-append the OpenSearch Benchmark command with `python3` and the path outputted earlier:
This will download OpenSearch 1.0.0 and run one of OpenSearch Benchmark's official workloads - the [geonames workload](<https://github.com/opensearch-project/opensearch-benchmark-workloads/tree/main/geonames>) - against it.
45
45
Note that this uses the `--test-mode` argument to only run a single instance of each operation in order to reduce the time needed for a test execution. This argument is used as a sanity check and should be removed in an actual benchmarking scenario.
Execute the `http_logs` workload against an existing OpenSearch cluster but only use one client for indexing and only ingest 10% of the total data corpus.
48
+
Execute the `http_logs` workload against an existing OpenSearch cluster but only use one client for indexing and only ingest 10% of the total data corpus.
0 commit comments