This is a pytest plugin that enables you to test your code that relies on a running Elasticsearch search engine. It allows you to specify fixtures for Elasticsearch process and client.

How to use


This plugin requires at least version 5.0 of elasticsearch to work.

The plugin contains two fixtures:

  • elasticsearch - a client fixture that has functional scope, and which cleans Elasticsearch at the end of each test.
  • elasticsearch_proc - a session scoped fixture, that starts Elasticsearch instance at its first use and stops at the end of the tests.
  • elasticsearch_nooproc - a nooprocess fixture, that's holds connection data to already running elasticsearch

Simply include one of these fixtures into your tests fixture list.

You can also create additional elasticsearch client and process fixtures if you'd need to:

from pytest_elasticsearch import factories

elasticsearch_my_proc = factories.elasticsearch_proc(
    port=None, logsdir='/tmp')
elasticsearch_my = factories.elasticsearch('elasticsearch_my_proc')


Each elasticsearch process fixture can be configured in a different way than the others through the fixture factory arguments.

Connecting to already existing Elasticsearch service

Some projects are using already running Elasticsearch servers (ie on docker instances). In order to connect to them, one would be using the elasticsearch_nooproc fixture.

es_external = factories.elasticsearch('elasticsearch_nooproc')

By default the elasticsearch_nooproc fixture would connect to elasticsearch instance using 9300 port.


You can define your settings in three ways, it's fixture factory argument, command line option and pytest.ini configuration option. You can pick which you prefer, but remember that these settings are handled in the following order:

  1. Fixture factory argument
  2. Command line option
  3. Configuration option in your pytest.ini file

Configuration options

ElasticSearch option Fixture factory argument Command line option pytest.ini option Noop process fixture Default
Elasticsearch executable executable --elasticsearch-executable elasticsearch_executable /usr/share/elasticsearch/bin/elasticsearch
logs directory logsdir --elasticsearch-logsdir elasticsearch_logsdir $TMPDIR
host host --elasticsearch-host elasticsearch_host host
port port -elasticsearch-port elasticsearch_port 6300 random
Elasticsearch cluster name cluster_name --elasticsearch-cluster-name elasticsearch_cluster_name elasticsearch_cluster_<port>
index storage type index_store_type --elasticsearch-index-store-type elasticsearch_index_store_type mmapfs
network publish host network_publish_host --elasticsearch-network-publish-host elasticsearch_network_publish_host
logs prefix logs_prefix --elasticsearch-logs-prefix elasticsearch_logs_prefix
transport tcp port transport_tcp_port --elasticsearch-transport-tcp-port elasticsearch_transport_tcp_port random

Example usage:

  • pass it as an argument in your own fixture

    elasticsearch_proc = factories.elasticsearch_proc(
  • use --elasticsearch-logsdir command line option when you run your tests

    py.test tests --elasticsearch-cluster-name=awsome_cluster
  • specify your directory as elasticsearch_cluster_name in your pytest.ini file.

    To do so, put a line like the following under the [pytest] section of your pytest.ini:

    elasticsearch_cluster_name = awsome_cluster

Known issues

It might happen, that the process can't be started due to lack of permissions. The files that user running tests has to have access to are:

  • /etc/default/elasticsearch

Make sure that you either run tests as a user that has access to these files, or you give user proper permissions or add it to proper user groups.

In CI at the moment, we install elasticsearch from tar/zip archives, which do not set up additional permission restrictions, so it's not a problem on the CI/CD.