I work with Tryton for 2 years (as Project Manager on Tryton projects) and we are experiencing performance problems (several minutes to update Tryton when we modify a model) with our development environment.
I think the problem comes the way we develop with Tryton.
I’d like to ensure that the way we define our development environment is ok.
We use git for code managment, visual studio code (on windows) as IDE.
What are the best practices to create a development environment with constraints above ?
Currently we have a docker container we are connected to. We modify Tryton code directly into the container and restarts tryton server after each code modification. The postgreSQL server is an other container on a VM (not on the developer computer)
You woul usually only need to update the module you modified, and then only if you actually added a new field / xml record.
I am working with over 200 modules for a development instance, and yes a full upgrade takes time, but I usually avoid those when working on a feature in a given module, and only upgrade this module.
You should use the -v parameter for increased verbosity.
I have also very rarely noticed temporary problems on specific queries on Postgres’ information_schema (which Tryton uses a lot during the upgrade process). When those happen (they disappear on their own after a time), the upgrade process can literally take hours, and I strongly suspect that occurs more frequently with Postgres running in a docker, so maybe switching to a local postgres install / Linux environments may help.