Upgrading to Graylog 2.3.x

Graylog switches to Elasticsearch HTTP client

In all prior versions, Graylog used the Elasticsearch node client to connect to an Elasticsearch cluster, which was acting as a client-only Elasticsearch node. For compatibility reasons of the used binary transfer protocol, the range of Elasticsearch versions Graylog could connect to was limited. For more information and differences between the different ways to connect to Elasticsearch, you can check the Elasticsearch documentation.

Starting with version 2.3.0, we are switching over to using a lightweight HTTP client, which is almost version-agnostic. The biggest change is that it does not connect to the Elasticsearch native protocol port (defaulting to 9300/tcp), but the Elasticsearch HTTP port (defaulting to 9200/tcp).

Due to the differences in connecting to the Elasticsearch cluster, configuring Graylog has changed. These configuration settings have been removed:

elasticsearch_cluster_discovery_timeout
elasticsearch_cluster_name
elasticsearch_config_file
elasticsearch_discovery_initial_state_timeout
elasticsearch_discovery_zen_ping_unicast_hosts
elasticsearch_http_enabled
elasticsearch_network_bind_host
elasticsearch_network_host
elasticsearch_network_publish_host
elasticsearch_node_data
elasticsearch_node_master
elasticsearch_node_name_prefix
elasticsearch_path_data
elasticsearch_path_home
elasticsearch_transport_tcp_port

The following configuration options are now being used to configure connectivity to Elasticsearch:

Config Setting Type Comments Default
elasticsearch_connect_timeout Duration Timeout when connection to individual Elasticsearch hosts 10s (10 Seconds)
elasticsearch_hosts List<URI> Comma-separated list of URIs of Elasticsearch hosts http://127.0.0.1:9200
elasticsearch_idle_timeout Duration Timeout after which idle connections are terminated -1s (Never)
elasticsearch_max_total_connections int Maximum number of total Elasticsearch connections 20
elasticsearch_max_total_connections_per_route int Maximum number of Elasticsearch connections per route/host 2
elasticsearch_max_retries int Maximum number of retries for requests to Elasticsearch 2
elasticsearch_socket_timeout Duration Timeout when sending/receiving from Elasticsearch connection 60s (60 Seconds)
elasticsearch_discovery_enabled boolean Enable automatic Elasticsearch node discovery false
elasticsearch_discovery_filter String Filter by node attributes for the discovered nodes empty (use all nodes)
elasticsearch_discovery_frequency Duration Frequency of the Elasticsearch node discovery 30s (30 Seconds)

In most cases, the only configuration setting that needs to be set explicitly is elasticsearch_hosts. All other configuration settings should be tweaked only in case of errors.

Warning

The automatic node discovery does not work if Elasticsearch requires authentication, e. g. when using Shield (X-Pack).

Caution

Graylog does not react to externally triggered index changes (creating/closing/reopening/deleting an index) anymore. All of these actions need to be performed through the Graylog REST API in order to retain index consistency.

Special note for upgrading from an existing Graylog setup with a new Elasticsearch cluster

If you are upgrading the Elasticsearch cluster of an existing Graylog setup without migrating the indices, your Graylog setup contains stale index ranges causing nonexisting index errors upon search/alerting. To remediate this, you need to manually trigger an index range recalculation for all index sets once. This is possible using the web interface using the System->Indices functionality or by using the REST API using the /system/indices/ranges/<index set id>/rebuild endpoint.

Graylog REST API

Rotation and Retention strategies

The deprecated HTTP resources at /system/indices/rotation/config and /system/indices/retention/config, which didn’t work since Graylog 2.2.0, have been removed.

These settings are part of the index set configuration and can be configured under /system/indices/index_sets.

Stream List Response structure does not include in_grace field anymore

The response to GET /streams, GET /streams/<id> & PUT /streams/<id> does not contain the in_grace field for configured alert conditions anymore.

The value of this flag can be retrieved using the GET /alerts/conditions endpoint, or per stream using the GET /streams/<streamId>/alerts/conditions endpoint.