You probably already knew that you can download Kibana objects—dashboards, visualizations, and saved searches—as JSON files. JSON files are plain text, making them ideal to store in a version control tool like Git.

Using the command line and the Logz.io API, you can make an automated, recurring backup of your Kibana objects to GitHub. This way, you’ll have a complete history of your backups. Your Kibana objects will be accessible from GitHub if you ever need to migrate content to a new Logz.io account or self-hosted Kibana.

Even though this tutorial covers backing up to GitHub, you can use these methods to back up to other Git hosting providers—like GitLab, BitBucket, or your own self-hosted Git.

Variables in this article

In the code blocks in this article, you’ll see <<API-TOKEN>> and <<API-URL>> variables. These need to be replaced with information specific to your account:

  • Replace <<API-TOKEN>> with an API token from the account you want to use
  • Replace <<API-URL>> with your region’s base API URL. For more information on finding your account’s region, see Account region.
In this tutorial

Making sure you have everything you need

Let’s set up the prerequisites now so that we can use the command line for the rest of the tutorial.

Create a GitHub repo

Make sure you’re signed in to GitHub, and create a new repository.

Give a Repository name of “kibana-backup” and choose whether you want the repo to be Public or Private. Select Initialize this repository with a README, and then click Create repository.

Your new repo is created and you’re taken to its main page.

Clone the repo to your local machine

If Git isn’t already installed on your local machine, install it now. See Install Git from Atlassian if you need help with this.

Clone Github repo

Choose whether you want to clone using HTTPS or SSH, and then copy the URL.

If you’re not sure whether to use HTTPS or SSH to clone your repo, see Which remote URL should I use? from GitHub.

In the command line, clone the repo into a new folder named “kibana-backup” and cd into the kibana-backup folder:

git clone <<REPO-URL>> kibana-backup
cd kibana-backup

Downloading Kibana objects

You can use the Logz.io API to download all Kibana objects of each type (dashboards, visualizations, and saved searches), one at a time. This means that you’ll need to make three API requests to Logz.io.

Sample request 1: Write saved searches to a file
curl -X POST \
  https://<<API-URL>>/kibana/export \
  -H 'Content-Type: application/json' \
  -H 'X-API-TOKEN: <<API-TOKEN>>' \
  -d '{"type": "search"}'
  -o kibana-search.json

This API endpoint is documented in Export Kibana objects in the API docs.

...and the response

The -o flag in the request tells cURL to output the response to a file (in this case, kibana-search.json). Because of this, you won’t see a response in the command line.

To check that the request worked, run ls. The contents of the folder should match this:

kibana-backup/
  └╴README.md
  └╴kibana-search.json

You can also check the contents of kibana-search.json. If you have no saved searches in the account, you should see something like this:

{"kibanaVersion":"6.3.2","hits":[]}

If you do have saved searches, they’re stored as objects in the hits array, like this:

{
  "kibanaVersion": "6.3.2",
  "hits": [
    {
      "_index": "logzioKibanaIndex",
      "_type": "doc",
      "_id": "search:All-logs",
      "_score": 2.295064,
      "_source": {
        "type": "search",
        "search": {
          "title": "All logs",
          "description": "",
          "hits": 0,
          "columns": [
            "message"
          ],
          "sort": [
            "@timestamp",
            "desc"
          ],
          "version": 1,
          "kibanaSavedObjectMeta": {
            "searchSourceJSON": "{...}"
          }
        }
      }
    }
  ]
}
Sample request 2: Write visualizations to a file
curl -X POST \
  https://<<API-URL>>/kibana/export \
  -H 'Content-Type: application/json' \
  -H 'X-API-TOKEN: <<API-TOKEN>>' \
  -d '{"type": "visualization"}'
  -o kibana-visualization.json
...and the response

This time, you used -o to output the response to kibana-visualization.json.

The visualizations are stored as objects in the hits array, like this:

{
  "kibanaVersion": "6.3.2",
  "hits": [
    {
      "_index": "logzioKibanaIndex",
      "_type": "doc",
      "_id": "visualization:E-commerce-App-Transactions-overtime",
      "_score": 0.93727,
      "_source": {
        "type": "visualization",
        "visualization": {
          "title": "E-commerce App - Transactions overtime",
          "visState": "{...}",
          "description": "",
          "savedSearchId": "Webapp-Transactions",
          "version": 1,
          "kibanaSavedObjectMeta": {
            "searchSourceJSON": "{...}"
          }
        }
      }
    }
  ]
}

To check that the request worked, run ls. The contents of the folder should match this:

kibana-backup/
  └╴README.md
  └╴kibana-search.json
  └╴kibana-visualization.json
Sample request 3: Write dashboards to a file
curl -X POST \
  https://<<API-URL>>/kibana/export \
  -H 'Content-Type: application/json' \
  -H 'X-API-TOKEN: <<API-TOKEN>>' \
  -d '{"type": "dashboard"}'
  -o kibana-dashboard.json
...and the response

This time, you used -o to output the response to kibana-dashboard.json.

The dashboards are stored as objects in the hits array, like this:

{
  "kibanaVersion": "6.3.2",
  "hits": [
    {
      "_index": "logzioKibanaIndex",
      "_type": "doc",
      "_id": "dashboard:ELB-Access-Logs",
      "_score": 2.9026418,
      "_source": {
        "type": "dashboard",
        "dashboard": {
          "hits": 0,
          "description": "",
          "timeRestore": false,
          "version": 1,
          "panelsJSON": "[...]",
          "kibanaSavedObjectMeta": {
            "searchSourceJSON": "{...}"
          },
          "title": "ELB Access Logs",
          "_logzioOriginalAppId": 202
        }
      }
    }
  ]
}

To check that the request worked, run ls. The contents of the folder should match this:

kibana-backup/
  └╴README.md
  └╴kibana-dashboard.json
  └╴kibana-search.json
  └╴kibana-visualization.json

Committing and pushing to GitHub

Once you’ve run all three requests and retrieved saved searches, visualizations, and dashboards, you’re ready to send everything to GitHub.

Code sample
git add .
git commit -m 'Think of a helpful commit message and write it here'
git push

The -m flag in git commit gives a message for the Git commit.

Navigate to the repo in GitHub in your browser. You’ll know everything worked as expected if you see each of the three files listed in the repo (kibana-dashboard.json, kibana-search.json, kibana-visualization.json).

Boom! 💥 You did it.

Now you can automate the whole process and create a script to back up your Kibana objects at regular intervals. Or if you’re feeling really fancy, you can commit and push to GitHub only when there are changes to your saved objects.

Have at it!