Table of Contents

Standalone Elastic Backup tool

Note

Elasticsearch is only supported up to version 6.8, which is no longer supported by Elastic. We therefore recommend using Storage as a Service instead, or if you do want to continue using self-hosted storage, using OpenSearch.

This tool allows you to back up and restore Elasticsearch database clusters. By default, it is available on a DMA in the folder C:\Skyline DataMiner\Tools. To get a full backup of the Elasticsearch cluster, you only need to execute the tool on one of the Elasticsearch nodes.

Important

The standalone Elastic Backup tool is being deprecated in favor of Elasticsearch's own snapshot functionality. We therefore recommend that you take database backups using the Elasticsearch snapshot functionality instead.

Syntax

Use the following general syntax when you run this tool:

StandaloneElasticBackup.exe <action> <arguments>

Example:

StandaloneElasticBackup.exe backup --host 127.0.0.1 -u elastic --pw mypw123 -r reponame

Actions

The following actions can be specified in the syntax mentioned above:

  • init: Initialize a repository.
  • backup: Take a backup/snapshot.
  • restore: Restore a backup/snapshot.

Arguments

Depending on the specified action, different arguments are required.

General arguments

The following arguments should be used regardless of the specified action:

  • --host or -h: The host on which the command has to be run.

    Default: 127.0.0.1

  • --port or -p: The port on which Elasticsearch will be contacted.

    Default: 9200

  • --username or -u: The username that has to be used to connect to Elasticsearch.

    Only use this argument when credentials are required.

  • -- pw: The password that has to be used to connect to Elasticsearch.

    Only use this argument when credentials are required.

Arguments when initializing the repository

To initialize a repository for a snapshot, use the following arguments:

  • --repo or -r: The name of the repository to be created.

  • --path: The path of the repository as defined in the yaml.xml file of the Elasticsearch cluster (i.e. path.repo), enclosed by "". Add an extra backslash ("") in front of backslash characters to make sure these are interpreted correctly. This is the location where the snapshot will be placed.

Note

Snapshots are incremental. Do not delete any of them. For more information, see Register a snapshot repository.

Arguments when taking a backup/snapshot

To take a backup/snapshot, use the following arguments:

  • --repo or -r: The repository in which to take the backup.

    • If none is defined and only one repository is found in the Elasticsearch cluster, then that one will be used.
    • If none is defined and none can be found, no backup will be taken.

  • --SnapshotName or -n: The (lowercase) name of the snapshot to be taken.

    Default: DateTime.Now.ToString("yyyy-MM-dd-HH-mm-ss");

Arguments when restoring a backup/snapshot

To restore a backup/snapshot, use the following arguments:

  • --repo or -r: The repository containing the backup to be restored. If none is defined and only one repository is found in the Elasticsearch cluster, then that one will be used. If none is defined and none can be found, an error will be thrown.

  • --SnapshotName or -n: The name of the snapshot to be restored.

Note
  • If all data was deleted from the database, first initialize the repository again before you restore a backup.
  • We recommend that you disable security when restoring a backup with security enabled. To do so, comment the security setting in the .yaml file.
  • It is only possible to restore indices that do not exist yet. Therefore, in most cases, you will have to delete all data from all nodes before restoring a backup.
  • Do not restore a backup while DataMiner is online.
  • You can get a list of existing snapshots to help find the right snapshot name by running a query on Kibana DevTools, e.g. GET /snapshot/my_repository. For more information, see Get a list of available snapshots.
Propose changes