redirected to /maintenance after update about wishthis

[grandeljay]

Hey @cryptogeek13,

thank you for reporting this!

Have you tried logging in? Usually there are some changes that need to be done to the database. Once you're in, there should be a migration button for the database.

Does that work?

[cryptogeek13]

That triggers navigation but I end up at /maintenance again. I guess I get redirected, but can' check right now because I'm on mobile.

[cryptogeek13]

Can I trigger database migration directly calling the API?

[grandeljay]

That triggers navigation but I end up at /maintenance again. I guess I get redirected, but can' check right now because I'm on mobile.

Are you the instance admin? The redirect to the maintenance page is intended behaviour. The administrator should upgrade the database after logging in. It should be visible on the maintenance page if you are logged in as administrator.

In case it isn't (for whatever reason):

1. Check if your power is set to 100 in the users table or
2. Run all of the SQL files manually which are higher than your current version: in /src/update.

Let me know if any of this helps you!

[cryptogeek13]

Well, I think I am the instance admin but since I can't get to the login page I can't tell if I could start the migration. I''ll try and do the manual steps tomorrow and check my permissions in database while I'm at it. But in my head my user rights could only matter if I managed to log in.

[cryptogeek13]

I tried again today and funnily enough this time the login page opened when I clicked the login button, even though nothing changed (same machine, browser, container than yesterday). After logging in, I still got redirected to /maintenance without being logged in. But second try worked and got me to /update.

There again I had to click the migration button to V1.1.0 two times, even though I got a green toast message after the first time.

Now I'm back to normal, but the first couple changes to a wish didn't get reflected, even though I got a success toast message. I had to force reload without caching a few times (strg+F5) to make it work.

There is definitely something wrong with my setup, could it be some kind of caching issue?

[grandeljay]

I've noticed caching issues on the stable branch but it should work on the release-candidate and develop branch. Which one are you currently using?

[cryptogeek13]

I'm using the release-candidate docker image. If your changes on release-candidate and develop are recent, it might be a little bit behind? Docher Hub says all images are from 11 days ago

[grandeljay]

Unfortunately, I don't know how frequently the images update but I removed caching about three weeks ago so it should already be included in the RC image.

Would you mind trying it in a non chromium based browser (such as Firefox)? I've noticed that Google Chrome in particular has very aggressive caching and even ignores some HTTP headers, which state that the document has changed and needs updating.

[cryptogeek13]

Actually I am already using Firefox. But I just checked and can confirm that those lines you removed in the linked commit are still present in my image, so I guess that's the problem. Should I open an issue in the docker repo?

[grandeljay]

Yes, please!

[cryptogeek13]

Alright, this can be closed then. Thanks for your help analyzing this