The instance has been updated to version 0.19.4!
You can see the official release post and notes here: https://lemmy.ml/post/16585359?scrollToComments=true
Since this update required some manual intervention anyway to update the database version, I took the opportunity to perform some overdue server maintenance and upgrades.
I have migrated the server to a more powerful configuration, and changed the way image files are being stored and served (object storage as opposed to filesystem storage).
I am also aware of some federation issues that have affected this and other instances. I am not sure whether this upgrade will deal with these problems automatically. I will be looking into it and hopefully will implement some solution soon.
Sorry for the downtime experienced in the past few days, but hopefully with the more powerful server the experience will be smoother and will remain so for a while.
For other admins and users interested, this is a bit of information about the server
The previous configuration:
Hosting provider: serverspace.io
8 GB RAM
4 Core CPU
50 Mbps network
400 GB SSD
Cost: $55 / month
The 400 GB SSD was being used to store both the server's database and the image files, and this component in isolation had a monthly cost of $36. I have migrated the images to an object storage provider ($6/month for 500 GB), and used the resources previously used for disk space to increase the other parameters. The new configuration is a lot better and less expensive.
New configuration:
Hosting provider: serverspace.io
Object storage provider: contabo.com
16 GB RAM
6 core CPU
50 Mbps network
100 GB SSD
500 GB Object storage
Cost: $47/month
Thanks a lot to the lemmy and pict-rs devs for the hard work they put into developing the infrastructure!
Thanks!
Since the upgrade, I can see (here) that activities from lemmy.world are being fetched significantly faster. I will let it run for some time and hopefully this speed-up is enough to catch up.
Great, looks promising!
You seem also to be catching up with the backlog somehow? https://grafana.lem.rocks/d/cdfzs0dwal3pca/federation-health-time-behind?orgId=1&var-instance=All&var-remote_instance=mander.xyz&var-remote_software=All&from=now-12h&to=now
Hopefully the trend will continue, let's wait a few hours and see
FYI, I just tried posting from a LW account 4 minutes, the comments still don't show up: https://lemmy.world/comment/10702133
Ah, thanks!
Hmm, I am not sure about how the queue gets filled. It might be that the instance needs to process all of the older content first.
I will give it some hours and if it does not look like it will catch up, I will look into other methods.
The instance is catching up at a rate of ~15 activities/s, and we are 2.5M activities behind. So it should take about 2 days to catch up.
I have tried fiddling with some parameters to see if this could be improved, but I did not find the right knob to tune. Best is probably to just let it run 😄
The trend seems to go down at a nice pace, there's something quite satisfying to see it go down like that ha ha
Does this have anything to do with the front page giving an error message? I get a lemmy server error, however I was able to log out and log back in again, and I can get to the communities page which got me back to this discussion, but even forcing a browser refresh of the front page continues to give an error message for me. FYI, ctrl-shift-R also fails, I'm not sure what else to try.
Hm, it is here ha ha https://mander.xyz/post/14329054
https://join-lemmy.org/news/2024-06-19_-Lemmy_Release_v0.19.5-_A_Few_Bugfixes
This was a small update so it only required me to get to my computer. I just had to change 0.19.4 to 0.19.5 and restart the docker containers :)
The 0.19.3 -> 0.19.4 was lot more involved because of the database upgrade and all of the extra server stuff that I needed to catch up on.
Sal, I got some error here:
couldnt_get_posts. That appears every time that I try to visualise the instance through the "Subscribed" communities.The error does not appear if I browse the instance through "Local" or "All", nor if I check the list of communities that I'm subscribed to. The error happens across devices and front-ends (tested with Linux Firefox, Android Firefox, Jerboa).
EDIT/UPDATE: I unsubscribed manually to a few communities each time, checking if the front page loaded fine. Eventually, it did.
Re-subscribing to either !technology@lemmy.world or !fediverse@lemmy.world seems to trigger the error again.
(@Shdwdrgn@mander.xyz ) I think it is related with the processing of so many activities per second while syncing with Lemmy.World.
I have tried changing some rate limits and the number of maximum database connections, but I get similar errors while moving around the site. The server is running well below 100% on all metrics, and the logs don't produce an obvious error. I will wait until we catch up with Lemmy.World and see if that fixes the issue. If it doesn't, then I will trouble shoot more in-depth.
There's some potential solutions to this mentioned here https://github.com/LemmyNet/lemmy/issues/4277
Thanks! The database is currently fetching lemmy world data as fast as possible, so this might be saturating the database. Once the syncronization is tested, I will try to tune postgres if it still happens.
The issue seems to be gone for me. And the instance "feels" considerably faster than it did before :-D
Great, thanks for letting me know!! I did a bit of research and I think that the next time I need to upgrade the instance I will move it to a dedicated server instead of a virtual private server. But hopefully it runs smoothly for some time! :)
Hey @RobotToaster@mander.xyz , @lvxferre@lemmy.ml , and @Shdwdrgn@mander.xyz - we are now in sync with lemmy.world and the database under much less stress. This seems to have fixed the problems that I observed on my end. Are your problems also resolved?
I just resubbed to the communities that were causing issues and it looks like everything is working well now, thanks.
Ah! It looks like it JUST fixed it, I had checked about 15 minutes ago and still couldn't get my subscribed list to load, but it does appears to be loading up now. I'll reset my preferences and keep an eye on it through the day -- hoping for the best!
It should have been fixed about 13 minutes ago! Thanks for checking! If problems persist I will try to fine-tune the database settings as suggested above.
There was a brief outage soon after I left that last reply, assume you were doing another round of maintenance at the time, but otherwise it's still looking good from here.
Ah, thanks. I think that was when I reboot the instance to change a setting that controls how post thumbnails are generated, as the new setting was failing to generate thumbnails for many posts.
That's the same error I was asking about below, and wondering if it had to do with the current status of collecting all the content. If so, it sounds like it could be another day or two before we can view our front pages again?
Sorry for the double reply.
Check if you're subscribed to any of those communities: !technology@lemmy.world , !comicstrips@lemmy.world or !memes@lemmy.world. If you are, try to unsubscribe and see if your front page works again. For me it did the trick.
!fediverse@lemmy.world was also giving me problems, but apparently it's fine now, so odds are that the problem will eventually go away on its own.
Unsubscribing from comic strips did the trick a couple days ago, thanks! And just now I’ve subscribed back and everything is normal. So feel free to subscribe again.
All of this sounds like the issue is related to the new server still loading up the communities. I'm subbed to a pretty good list so it's probably best just to wait for the server to catch up. Thanks for the info though!
For now, one option is to temp change your settings to browse by "Local" or "All". Or alternatively temporarily unsubscribe off a few comms; if you're subscribed to a small number of comms, some trial and error might help you to find which ones are causing the issue. (For me it was two LW comms).
