Upgrading

Staying up to date.

If you're upgrading to a new major release, it's always recommended to start by generating a new configuration file (using the new version of Dispatch). This will ensure that any new settings which may have been added are clearly visable and get configurated correctly.

Beyond that, upgrades are simple as bumping the version of Dispatch (which will cause any changed dependencies to upgrade), running data migrations, and restarting all related services.

In some cases you may want to stop services before doing the upgrade process or avoid intermittent errors.

Upgrading Dispatch

Upgrading the package

The easiest way to upgrade the Dispatch package usingpip:

pip install --upgrade dispatch

You may prefer to install a fixed version rather than just assuming the latest, as it will allow you to better understand what is changing.

If you're installing from source, you may have additional requirements that are unfulfilled, so take the necessary precautions of testing your environment before committing to the upgrade.

Running Migrations

Just as during the initial setup, migrations are applied with the upgrade command.

dispatch database upgrade

Restarting services

You'll need to ensure that all services running Dispatch code are restarted after an upgrade. This is important as Python loads modules in memory and code changes will not be reflected until a restart.

These services include:

  • server -- dispatch server start

  • scheduler -- dispatch scheduler start