Add an option for cleaning/dropping database before restoring a backup

[GuillaumeSavin]

Hello

I will first thanks you all portabase team for your project, this is really nice, I tested backup and restoration and I am pretty pleased with the result. But I think something that will add value would be able to choose to DROP database before restauration because as for now restauration partially failed when on an already filled database (error with constraint, keys table, etc... already existing). Right now I need to stop my bdd container and delete its volume before restarting it so the restauration process can complete without any problem. But that would be nice to be able to do it directly from portabase by deleting database before restoring it (as an option of course to prevent bad manipulation). But in any case I want to thank you again for your project !

[RambokDev]

Hi @GuillaumeSavin,

Which database provider are you using? Normally, Portabase should drop the database before restoring it. This is what we do for PostgreSQL.

Thanks for you support,
Regards

[RambokDev]

Sorry, let me correct what I said. Depending on the provider, we only drop the objects that are recreated during the restore. For example, with PostgreSQL, we use the --clean option.

We could add a command argument to fully drop the database before restoring it for users who need that behavior. At the moment, this is not implemented for all providers.

What is your exact use case? Which database provider are you using, and what restore workflow do you expect?

Regards

[GuillaumeSavin]

First thanks for your quick answer !
The provider I use is postgres. My use case would be to either be able to restore on fresh instance in this case no need to drop database. But if i want to restore on an already full of data instance, something that will be great would to have an option to drop database before actually restoring, maybe a checkbox in modal after clicking on restoring button ?

In any case here are the logs when the restoration failed (errors between lines 3709-3711, 3806-4772, 4935-4937, 5041-6007)

portabase_restoration.log

[RambokDev]

Did you try with the keep_ownership true option ? https://portabase.io/docs/agent/db/postgresql#options

[RambokDev]

If the issue persists, could you join us on Discord so we can investigate it further? We’ll find a solution and, if necessary, release a fix or add a new feature.

Regards,

[GuillaumeSavin]

Re @RambokDev I tested to add keep_ownership to true in databases.json file of portabase agent force-recreate the container to make sure it restart with new configuration but it still failled with pretty much the same problem. I will join you on discord then ! Than again for your help !

[RambokDev]

Portabase/agent#88

[GuillaumeSavin]

Thanks @RambokDev for the news ! I will add it to discord channel about this matter.

[RambokDev]

Thanks @RambokDev for the news ! I will add it to discord channel about this matter.

I already wrote you on discord about that