Extremely cursed behavior with migrations and suspension of migrated accounts
What happened? (Please give us a brief description of what happened.)
Where do I begin? I am going to strip out the context for this because that is irrelevant for the bug report, all that I am going to say is that due to a particular set of circumstances I have ended up in a predicament.
- I have an account on transfem.social, it is named puppygirlhornypost@transfem.social.
- This account was migrated to kitsunes.club, the migration completed successfully. The account puppygirlhornypost@kitsunes.club still holds onto the account alias for puppygirlhornypost@transfem.social.
- After the migration was complete, puppygirlhornypost@transfem.social was suspended. All I know is, an account suspension deactivates the actor, Akkoma and Pleroma are unable to reactivate it. Misskey (and by extension sharkey) instances are able to handle unsuspension fine, showing the now unsuspended user and its prior note history. Except for the destination instance.
- The destination instance (earlier referred to as kitsunes.club) is unable to retrieve ANY notes from the unsuspended account originally located on transfem.social. EXCEPT for pinned notes, it is able to see those perfectly fine.
What did you expect to happen? (Please give us a brief description of what you expected to happen.)
How do I say this with the least amount of snark... I expected that when I unsuspended my account after regaining control that I would be facing serious issues. I went ahead and made a public post detailing to Akkoma and Pleroma instances that they need to manually reactivate my actor in order for me to be able to reference my prior post history. I am treating that account as an archive, and it is crucial for me to be able to federate those notes because there are over 20,000 of them on a variety of topics from mental health, server administration, disability awareness and so forth. It doesn't really matter, I would still write this issue even if I had 20k notes of "meow", because in my opinion everyone deserves to have a functioning account.
Version (What version of Sharkey is your instance running? You can find this by clicking your instance's logo at the top left and then clicking instance information.)
The originating instance (transfem.social) is/was running a custom fork 2024.6.0-transfem. The destination instance (kitsunes.club) is also running a custom fork (or I believe it to be) based around 2024.6.0-dev.
Instance (What instance of Sharkey are you using?)
Instances* (i do not want to modify the template further)
https://transfem.social/ https://kitsunes.club/
What type of issue is this? (If this happens on your device and has to do with the user interface, it's client-side. If this happens on either with the API or the backend, or you got a server-side error in the client, it's server-side.)
I am sorry, but this is probably one of the rare occasions that we have a bug on multiple layers. The fact that I can retrieve pinned notes makes me believe it's partially the front end failing (although I could isolate this by trying to fetch the note via ap/get directly on the api - something I am willing to try if someone picks up this case)
What browser are you using? (Client-side issues only)
Firefox, but it shouldn't matter
What operating system are you using? (Client-side issues only)
Fedora 40
How do you deploy Sharkey on your server? (Server-side issues only)
N/A, nothing to my knowledge should be breaking sharkey internally like this. I will say, both instances run Varnish as a cache/http accelerator. Varnish can be ruled out by one thing, and that is these are POST requests being made and unless there is a bug in varnish (if there was everything else would be broken) our caching middleware is not going to intercept said requests. Traefik (the reverse proxy of kitsunes.club) is also not going to be modifying transit (all it does is tls termination, and acting as a reverse proxy) on kitsunes.club. On transfem.social, I do not see NGINX+HAPROXY modifying the requests (tls termination, reverse proxying for nginx; load balancing websockets for HAProxy)
What operating system are you using? (Server-side issues only)
Ubuntu, or debian. I don't remember what my hosts run. I do not believe this could be an issue related to NETFILTER (the linux kernel's firewall typically paired with a userspace component such as ipchains, iptables, and now nftables) or the linux kernel's networking stack.
Relevant log output (Please copy and paste any relevant log output. You can find your log by inspecting the page, and going to the "console" tab. This will be automatically formatted into code, so no need for backticks.)
N/A (logs are shredded)
Contribution Guidelines By submitting this issue, you agree to follow our Contribution Guidelines
- I agree to follow this project's Contribution Guidelines
- I have searched the issue tracker for similar issues, and this is not a duplicate.