How to Upgrade (Legacy Images)

Important: This instructions are for those who use our older instace that did not use Docker. If you use our new Docker based instance, check the updated guide.

Important: Starting from the DataReporter V8 release we’re no longer building the tarball version of a release – only Docker images. This of course doesn’t prevent you from deploying DataReporter anyway you want, but the bin/upgrade script won’t work with the new releases.

It’s recommended to upgrade your DataReporter instance once there are new releases, to benefit from new features and bug fixes. The upgrade process is relatively simple, and assuming you used one of the base images we provide, you can simply run the upgrade script. If you have a custom deployment, you can use the upgrade script as reference to create your own process.

Danger: Bitnami instances are not compatible with this script.

How to upgrade (v1.0.0 and newer)

Starting from v1.0.0, the upgrade script is part of the codebase, and running the script is as simple as:

cd /opt/redash/current
sudo bin/upgrade

Info: If you’re upgrading to a beta version, you need to specify the channel: sudo bin/upgrade --channel beta.

Docker

If you’re using Docker to run DataReporter, don’t use the upgrade script. Instead you need to:

1. Update the Docker image you’re using.
2. Apply migrations (if needed). To apply migrations you need to invoke the image with manage db upgrade as the command. The way to do this depends on the way you deploy DataReporter.

How to upgrade (pre v1.0.0 versions)

In v1.0.0 we added the script to the repository, but if you have an earlier version, you will need to download the script first (note that you need to download it to your DataReporter server):

wget https://raw.githubusercontent.com/getredash/redash/master/bin/upgrade
chmod +x upgrade

Run it:

sudo ./upgrade

Before doing the upgrade, please make sure to do the following changes to your /opt/redash/.env file:

1. If you have local PostreSQL database, you will need to update the URL from postgresql://redash to postgresql:///redash.
2. Remove the REDASH_STATIC_ASSETS_PATH definition.

Note: You can upgrade to v1.0.0 and later only from v0.12.0, so if you have an older version, run: sudo upgrade --channel legacy first and then sudo upgrade.

Troubleshooting

Upgrade failing with “sudo: /etc/init.d/redash_supervisord: command not found”

You need to manually run the restart command:

sudo service supervisor restart

In v2.0.0 we fixed the upgrade script to use the correct command.

Some (or all) data sources disappeared after upgrading

The upgrade process updates the code, applies migrations and upgrade Python requirements. But it does not upgrade Python requirements for the data sources (described in /opt/redash/current/requirements_all_ds.txt).

In some cases, old packages will prevent the data source from loading. Make sure to manually update the requirements relevant to the data sources you need.

I don’t see BigQuery in the data sources list after upgrading to v2.0.0

You need to upgrade the following Python packages:

google-api-python-client==1.5.1
oauth2client==3.0.0

(sudo pip install -U google-api-python-client==1.5.1 oauth2client==3.0.0)

I don’t see Athena in the data sources list after upgrading to v2.0.0

Make sure to install PyAthena:

sudo pip install PyAthena>=1.0.0

Getting “AttributeError: ‘module’ object has no attribute ‘SSL_ST_INIT’” error during upgrade

Make sure to upgrade the PyOpenSSL package (see the version specified in requirements_all_ds.txt).