corrupted index db . Deleting the index db to avoid errors ..

[merryHunter]

❓Question

Hi, I am following the instructions in a fresh python environment from the page https://aimstack.readthedocs.io/en/latest/using/remote_tracking.html#server-side-setup .However, when I click on the ui link, a strange exception is caught and I see the following print:

ubuntu@ip-172-31-42-100:~/server-tracking$ aim up
Running Aim UI on repo `<Repo#-2658956876120156914 path=/home/ubuntu/server-tracking/.aim read_only=None>`
Open http://127.0.0.1:43800
Press Ctrl+C to exit
Corrupted index db. Deleting the index db to avoid errors. Please run `aim storage reindex command to restore optimal performance.`

After I run aim storage reindex and restart the ui, nothing changes, same error when I open UI in browser.

I doubt it is an expected behavior. Could it be related to some known bug?

Env info

Python: 3.10.12
aim: 3.25.0

[mihran113]

Hey @merryHunter!
Thanks for pointing this out, there was an additional change in this part which resulted in this behavior: marking empty index db as corrupted, for now you can just ignore that. I've opened a fix for that, we'll ship it with the next patch release.

[merryHunter]

great, thank you!

@mihran113 Could you please share discord invite link? I have a few questions on how to run aim in production setup with nginx, probably I can find answer in discord or ask around there..

[mihran113]

Hey @merryHunter!
This is the discord invite link: https://discord.com/invite/zXq2NfVdtF

[ic-dc]

@mihran113 seems it expires very soon, I just clicked and it says invalid again :(