Logo

A powerful, easily deployable network traffic analysis tool suite for network security monitoring

Quick Start

Documentation

Components

Supported Protocols

Configuring

Arkime

Dashboards

Hedgehog Linux

Contribution Guide

OpenSearch Dashboards

OpenSearch Dashboards is an open-source fork of Kibana, which is no longer open-source software.

Adding new visualizations and dashboards

Visualizations and dashboards can be easily created in OpenSearch Dashboards using its drag-and-drop WYSIWIG tools. Assuming users have created a new dashboard to package with Malcolm, the dashboard and its visualization components can be exported either of two ways.

The easier (and preferred) method is to use the dashboard export API, as it handles the replacers (described below in the more complicated method) automatically.:

  1. Identify the ID of the dashboard (found in the URL: e.g., for /dashboards/app/dashboards#/view/xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx the ID would be xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx)

  2. Using a web browser, enter the URL https://localhost/mapi/dashboard-export/xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx, replacing localhost with the IP address or hostname of your Malcolm instance and the placeholder dashboard ID with the ID you identified in the previous step. Save the raw JSON document returned as ./dashboards/dashboards/xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx.json (using the actual ID) under your Malcolm directory.

OR

  1. Using the command line, export the dashboard with that ID and save it in the ./dashboards/dashboards/ directory with the following command:
export DASHID=xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx && \
  docker compose exec api curl -sSL -XGET "http://localhost:5000/mapi/dashboard-export/$DASHID" > \
  ./dashboards/dashboards/$DASHID.json
  1. Include the new dashboard either by using a bind mount for the ./dashboards/dashboards/ directory or by rebuilding the dashboards-helper image. Dashboards are imported the first time Malcolm starts up.

The manual, more complicated way, consists of the following steps:

  1. Identify the ID of the dashboard (found in the URL: e.g., for /dashboards/app/dashboards#/view/xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx the ID would be xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx)

  2. Using the command line, export the dashboard with that ID and save it in the ./dashboards/dashboards/ directory with the following command:

export DASHID=xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx && \
  docker compose exec dashboards curl -XGET \
  "http://localhost:5601/dashboards/api/opensearch-dashboards/dashboards/export?dashboard=$DASHID" > \
  ./dashboards/dashboards/$DASHID.json
  1. It is preferrable for Malcolm to dynamically create the arkime_sessions3-* index template rather than including it in imported dashboards, so edit the xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx.json that was generated, carefully locating and removing the section with the id of arkime_sessions3-* and the type of index-pattern (including the comma preceding it):
    ,
    {
      "id": "arkime_sessions3-*",
      "type": "index-pattern",
      "namespaces": [
        "default"
      ],
      "updated_at": "2021-12-13T18:21:42.973Z",
      "version": "Wzk3MSwxXQ==",
      …
      "references": [],
      "migrationVersion": {
        "index-pattern": "7.6.0"
      }
    }
  1. In your text editor, perform a global-search and replace, replacing the string arkime_sessions3-* with MALCOLM_NETWORK_INDEX_PATTERN_REPLACER and malcolm_beats_* with MALCOLM_OTHER_INDEX_PATTERN_REPLACER. These replacers are used to allow customizing indexes for logs written to OpenSearch or Elasticsearch.
  2. Include the new dashboard either by using a bind mount for the ./dashboards/dashboards/ directory or by rebuilding the dashboards-helper image. Dashboards are imported the first time Malcolm starts up.

OpenSearch Dashboards plugins

The dashboards.Dockerfile installs the OpenSearch Dashboards plugins used by Malcolm (search for opensearch-dashboards-plugin install in that file). Additional Dashboards plugins could be installed by modifying this Dockerfile and rebuilding the dashboards image.