Upgrade Postgres - Unable to connect to server

Hi all,
I’m running Speckle on our own deployed instance on AWS with https enabled etc. I followed the manual deployment guide here. So far so good and it was running all fine.

I recently tried to upgrade the server and ran into some trouble upgrading the postgres database. I managed to save a back-up of the database (we have a few projects in there that are worth keeping), which I would now like to attach to an entirely new server setup - just spun up v2.15.0 successfully. But…, I can’t access the database in pgAdmin so any help would be greatly appreciated.

I followed the descriptions here, spinnigng up a pgAdmin4 docker container with the default yml file on that page, but when connecting to the pgAdmin dashboard and trying to setup a new server, I get an error ‘Unable to connect to server: connection is bad: Try again’. Google and Stack Overflow are not of any great help there unfortunately.

Anyone experience with this, or any ideas what I’m missing perhaps? Thanks!

Hi @JJanssen

This seems like it is unable to connect to the database at all.

It is possible Docker Compose has placed pgAdmin dashboard in a distinct and separate virtual network from the Postgres database.
In the docker compose yaml file you used to deploy the database there may be a statement that defines the network name, e.g.:

    name: speckle-server

You could try adding the same statement to the file you used to deploy pgAdmin and then re-run it. This should then ensure it is in the same virtual network and able to contact the database. Alternatively, copy and paste the definition for the pgAdmin and place it in the same file in which you defined the database, and then run that file.


Life went easier after adding pgadmin into the same docker-compose.yml. Or you are creating the network like described by @iainsproat. Good thing … one compose-file for the update routine, in case you don’t have it locally anymore.


yes! thanks both! that was indeed the case… all it was is to indeed add those additional lines with the network name. Might be worth adding that to the vanilla docker compose yaml for the pgAdmin in the documentation.
Had to do the maintenance trick at the bottom of the page though, but all good now.

I do like the idea of combining them in one yaml file. It’s all running fine now (with the old database and users restored! :smiley:) , so will leave that for later.

Thanks again for your help!

1 Like

Glad you got it resolved. Thank you for the feedback, I’ve now updated the documentation :+1: