@TheracAriane@thebrainbin.org Because the forms for those got merged. Just click "Create Thread" and then insert the URL :) I switched to the 1.9.0 release candidate. Have a look at all the major changes here: https://kbin.melroy.org/m/mbinReleases@gehirneimer.de/t/1321152
I think what happened here is the following:
- You tried sending a post to an inactive and therefore not fetched for a while magazine
- The server updates the remote magazine actor so it has a fresh copy
- Notices the magazine is gone on the other server and therefore deleting it locally
- As a result your post was also deleted
I don't know if this is the best behavior... But it is built this way at the moment.... Maybe @melroy@kbin.melroy.org has an opinion as well? Maybe also post this to !mbinmeta@gehirneimer.de
@TheracAriane@thebrainbin.org I am sorry, but I haven't even heard of them 😅
Thank you :)
It is linked in the about page :) Sure thing :)
Yeah I have thought about that actually. The problem is simply with cost. Image hosting is expensive and minio is quite resource intensive... I would need to buy a set of enterprise grade ssds and should do so for the current server as well. It hosts the image storage for all 3 fediverse servers I host combined and takes ~1.5 TB of storage at the moment, so I would need at the very least 2 drives with 4TB each to have a lot of room to spare. If I go with WD Red nvme SSDs that is 341€ per drive... I don't have to do that and I will investigate the options further, because yes I know, having a single point of failure is not good. I just didn't expect I had to do it this soon, as the machine is not even 1.5 years old...
Here is a comarision of the CPU utilization. I restricted the access around 9:30 and you can see a huge drop in utilization at that time
@TheracAriane@thebrainbin.org If I am supposed to be able to fix it, I need more information than that :D