A Python-based application to backup Grafana settings using the Grafana API.
The aim of this tool is to:
- Easily backup and restore Grafana.
- Have versioned backups
(date and time in file name)
for restoring and saving to cloud storage providers. Currently supportAmazon S3
Azure Storage
GCP Cloud Storage
(Use service-account's credential file, see here)
- Folder
- Folder Permissions
- Library Elements (doesn't work with Grafana 8.0.0 but 8.4.3)
- Dashboard (contains Alert)
- Datasource
- Alert Channel
- Alert Rules (Supported in version 9.4.0 of grafana and up.)
- Teams
- Team Members (Needs Basic Authentication (username and password, see grafana doc)
- You need to set
Admin's account and password
ingrafanaSettings.json
, or set the base64 encodedadmin account and password
in ENVGRAFANA_BASIC_AUTH
. E.gexport GRAFANA_BASIC_AUTH=YWRtaW46YWRtaW4=
- Or Sets this ENV of the Grafana server
GF_USERS_ALLOW_ORG_CREATE=true
. see grafana doc
- You need to set
- Organization (Needs Basic Authentication (username and password, see grafana doc)
- You need to set
Admin's account and password
ingrafanaSettings.json
, or set the base64 encodedadmin account and password
in ENVGRAFANA_BASIC_AUTH
. E.gexport GRAFANA_BASIC_AUTH=YWRtaW46YWRtaW4=
- Or Sets this ENV of the Grafana server
GF_USERS_ALLOW_ORG_CREATE=true
. see grafana doc
- You need to set
- User (Needs Basic Authentication (username and password, see grafana doc)
- You need to set
Admin's account and password
ingrafanaSettings.json
, or set the base64 encodedadmin account and password
in ENVGRAFANA_BASIC_AUTH
. E.gexport GRAFANA_BASIC_AUTH=YWRtaW46YWRtaW4=
- Grafana's api doesn't provide user's password when backing up, so the
default_user_password
ingrafanaSettings.json
, or in ENVDEFAULT_USER_PASSWORD
, E.gexport DEFAULT_USER_PASSWORD=supersecret
will be used when restoring.
- You need to set
- Snapshots
- Dashboard Versions (only backup, no restore)
- Annotations
NOTE The only supported orgId
right now is 1
, the default organization will be backed up only!
- Bash
- Python 2.7 or Python 3.x
- Access to a Grafana API server.
- A
Token
of anAdmin
role (seeConfiguration
section below for more info)
There are three ways to setup the configuration:
- Use
environment variables
to define the variables for connecting to a Grafana server. - Use
hard-coded settings
inconf/grafanaSettings.json
(this is the default settings file if not specified otherwise). - Use
~/.grafana-backup.json
to define variables in json format.
- Copy grafanaSettings.example.json and modify it for you to use, remove
azure
,aws
,gcp
,influxdb
blocks (but keep the ones you used). - Check out the examples folder for more configuration details.
NOTE If you use environment variables
, you need to add the following to your .bashrc
or execute once before using the tool (please change variables according to your setup):
(GRAFANA_HEADERS
is optional, use it if necessary. please see #45)
### Do not use a trailing slash on GRAFANA_URL
export GRAFANA_URL=http://some.host.org:3000
export GRAFANA_TOKEN=eyJrIjoidUhaU2ZQQndrWFN3RRVkUnVfrT56a1JoaG9KWFFObEgiLCJuIjoiYWRtaW4iLCJpZCI6MX0=
# GRAFANA_HEADERS is optional
export GRAFANA_HEADERS=Host:some.host.org
To create and obtain a Token
for your Grafana server, please refer to the official documentation.
NOTE that you need to generate a Token
with an Admin
role for the backup to succeed, otherwise you will have potential permission issues.
Create a virtualenv, you could using something like pyenv
if you'd prefer
virtualenv -p $(which python3) venv
source venv/bin/activate
pip install grafana-backup
First clone this repo
git clone https://github.com/ysde/grafana-backup-tool.git
cd grafana-backup-tool
Installation works best using pip
pip install .
- First perform the Configuration and Installation sections as described above.
- Use the
grafana-backup save
command to backup all your folders, dashboards, datasources and alert channels to the_OUTPUT_
subdirectory of the current directory.
Example:
$ grafana-backup save
$ tree _OUTPUT_
_OUTPUT_/
└── 202006272027.tar.gz
- Use the
grafana-backup restore <archive_file>
command with a path to a previous backup to restore everything.
NOTE this may result in data loss, by overwriting data on the server.
Example:
$ grafana-backup restore _OUTPUT_/202006272027.tar.gz
Replace variables below to use the Docker version of this tool
{YOUR_GRAFANA_TOKEN}
: Your Grafana siteToken
.{YOUR_GRAFANA_URL}
: Your Grafana siteURL
.{YOUR_BACKUP_FOLDER_ON_THE_HOST}
: Thebackup folder
on the Grafana host machine.
Check out the CronJob in examples for a simple example of how grafana-backup-tool can be ran within a Kubernetes environment
If you decide to use a volume (-v) then you'll need to create the volume first with 1337 uid/gid ownership first, example:
mkdir /tmp/backup
sudo chown 1337:1337 /tmp/backup
docker run --user $(id -u):$(id -g) --rm --name grafana-backup-tool \
-e GRAFANA_TOKEN={YOUR_GRAFANA_TOKEN} \
-e GRAFANA_URL={YOUR_GRAFANA_URL} \
-e GRAFANA_ADMIN_ACCOUNT={YOUR_GRAFANA_ADMIN_ACCOUNT} \
-e GRAFANA_ADMIN_PASSWORD={YOUR_GRAFANA_ADMIN_PASSWORD} \
-e VERIFY_SSL={True/False} \
-v {YOUR_BACKUP_FOLDER_ON_THE_HOST}:/opt/grafana-backup-tool/_OUTPUT_ \
ysde/docker-grafana-backup-tool
Example:
docker run --user $(id -u):$(id -g) --rm --name grafana-backup-tool \
-e GRAFANA_TOKEN="eyJrIjoiNGZqTDEyeXNaY0RsMXNhbkNTSnlKN2M3bE1VeHdqVTEiLCJuIjoiZ3JhZmFuYS1iYWNrdXAiLCJpZCI6MX0=" \
-e GRAFANA_URL=http://192.168.0.79:3000 \
-e GRAFANA_ADMIN_ACCOUNT=admin \
-e GRAFANA_ADMIN_PASSWORD=adminpassword \
-e VERIFY_SSL=False \
-v /tmp/backup/:/opt/grafana-backup-tool/_OUTPUT_ \
ysde/docker-grafana-backup-tool
S3 Example: Set S3 configurations in -e
or grafanaSettings.json
(example)
-e AWS_S3_BUCKET_NAME="my-backups-bucket" \
-e AWS_S3_BUCKET_KEY="grafana-backup-folder" \
-e AWS_DEFAULT_REGION="us-east-1" \
-e AWS_ACCESS_KEY_ID="secret" \
-e AWS_SECRET_ACCESS_KEY="secret" \
Azure Example: Set Azure configurations in -e
or grafanaSettings.json
(example)
-e AZURE_STORAGE_CONTAINER_NAME="azure-storage-container-name" \
-e AZURE_STORAGE_CONNECTION_STRING="azure-storage-connection-string"
GCS Example: Set GCS configurations in -e
or grafanaSettings.json
(example)
-e GCS_BUCKET_NAME="bucket-name" \
-e GCLOUD_PROJECT="gcp-project-name" \
-e GOOGLE_APPLICATION_CREDENTIALS="credential-file-path"
docker run --user $(id -u):$(id -g) --rm --name grafana-backup-tool \
-e GRAFANA_TOKEN={YOUR_GRAFANA_TOKEN} \
-e GRAFANA_URL={YOUR_GRAFANA_URL} \
-e GRAFANA_ADMIN_ACCOUNT={YOUR_GRAFANA_ADMIN_ACCOUNT} \
-e GRAFANA_ADMIN_PASSWORD={YOUR_GRAFANA_ADMIN_PASSWORD} \
-e VERIFY_SSL={True/False} \
-e RESTORE="true" \
-e ARCHIVE_FILE={THE_ARCHIVED_FILE_NAME} \
-v {YOUR_BACKUP_FOLDER_ON_THE_HOST}:/opt/grafana-backup-tool/_OUTPUT_ \
ysde/docker-grafana-backup-tool
Example:
docker run --user $(id -u):$(id -g) --rm --name grafana-backup-tool \
-e GRAFANA_TOKEN="eyJrIjoiNGZqTDEyeXNaY0RsMXNhbkNTSnlKN2M3bE1VeHdqVTEiLCJuIjoiZ3JhZmFuYS1iYWNrdXAiLCJpZCI6MX0=" \
-e GRAFANA_URL=http://192.168.0.79:3000 \
-e GRAFANA_ADMIN_ACCOUNT=admin \
-e GRAFANA_ADMIN_PASSWORD=adminpassword \
-e VERIFY_SSL=False \
-e RESTORE="true" \
-e ARCHIVE_FILE="202006280247.tar.gz" \
-v /tmp/backup/:/opt/grafana-backup-tool/_OUTPUT_ \
ysde/docker-grafana-backup-tool
You can build the docker image simply by executing make
in the root of this repo. The image will get tagged as ysde:grafana-backup
InfluxDB support has been added and Prometheus push gateway support will be added in the future.
In order to monitor successful backups with InfluxDB simply configure grafana-backup
InfluxDB settings using this example configuration.
Or if you prefer to use environment variables you can instead set INFLUXDB_HOST
, INFLUXDB_PORT
, INFLUXDB_MEASUREMENT
, INFLUXDB_USERNAME
and INFLUXDB_PASSWORD
.
Once configured grafana-backup
will automatically enter a 1
in your defined timeseries measurement upon each successful backup.