Hey @Cyber_punk. This is a valid question, and one that I had wondered about myself a little while ago.
The answer is that each release of the timescaledb-ha docker image gets a new very-specific (and immutable) tag. An example is the recent pg14.4-ts2.7.1-p0 tag. That tag contains Postgres version 14.4, TimescaleDB version 2.7.1, and is the 0th build of that combination of tools. That image is also tagged with the “mutable” tags
pg14.4-ts2.7.1-latest. These mutable tags can (and probably will) be updated in future to point to a new version of the docker image.
As an example, if we update some tools in the docker image, but keep the same Postgres, and TimescaleDB versions, then that will be published as
pg14.4-ts2.7.1-p1. This is the “immutable” tag which will never point to a different docker image. When that version is published, the mutable tags (
pg14-latest, …) will be updated to point to the that docker image.
In general, for production use-cases, it would be safest to use an immutable tag (ending with
X is some number).
If you’re just trying out an image, feel free to use
As for the differences in prompt (
tsdb=>), the name shown on the prompt is the name of the database that you’re connected to. I’m assuming that the tutorial is based on timescale cloud, in which the default database name is
tsdb. The default database in the docker images is