Yeah, it’s definitely annoying, but if it’s any consolation, it’s a system-wide problem. Not sure if/when/how they fix, but if this place is going to be a long term Reddit replacement, I feel like they’re going to have to.
this post was submitted on 21 Jun 2023
42 points (100.0% liked)
Selfhosted
40465 readers
544 users here now
A place to share alternatives to popular online services that can be self-hosted without giving up privacy or locking you into a service you don't control.
Rules:
-
Be civil: we're here to support and learn from one another. Insults won't be tolerated. Flame wars are frowned upon.
-
No spam posting.
-
Posts have to be centered around self-hosting. There are other communities for discussing hardware or home computing. If it's not obvious why your post topic revolves around selfhosting, please include details to make it clear.
-
Don't duplicate the full text of your blog or github here. Just post the link for folks to click.
-
Submission headline should match the article title (don’t cherry-pick information from the title to fit your agenda).
-
No trolling.
Resources:
- selfh.st Newsletter and index of selfhosted software and apps
- awesome-selfhosted software
- awesome-sysadmin resources
- Self-Hosted Podcast from Jupiter Broadcasting
Any issues on the community? Report it using the report flag.
Questions? DM the mods!
founded 2 years ago
MODERATORS
Not sure if this is the same problem, but it has been noticed that passing messages between federated instances has been hampered by the way the "worker threads" are currently been done. The big instances are essentially getting overloaded with outbound federation work. Solutions for that are in the works. issue/3230
I had a similar problem that ended up having to do with my server's language settings. I didn't realize you can select more than just "Undetermined" by holding down shift or ctrl. So everything tagged as English didn't appear until I went back and fixed it. You can read the whole odyssey here: https://lemmy.ml/post/1205767
Ha! Love it when people post troubleshooting in real time. I have both undetermined and english selected, so I don't think that's it.
Yeah, people around here are super helpful! Sorry we don't have the same problem, hopefully someone else can give some insight!
Hopefully it'll mostly resolve given time. Just like when there was a mass exodus to mastodon, the major thrediverse servers are getting overloaded. This creates delays and it seems even lost messages. From experience I've had messages from fedia.io take anywhere from 20-120 minutes to show up on my mastodon. As admins beef up their servers, but also as excitement tones down and there is less enthusiastic usage, hopefully the problem will resolve for the most part.
I'm not sure what causes messages to be lost forever vs. just delayed. Maybe Kbin or Lemmy aren't quite as robust in that? When mastodon was melting down, messages could take 8+ hours to get through, but they seemed to get through. I think your idea about a resync idea is a great one.
Yeah I started on a dinky little server and now it's massive.... mostly seems to be a memleak or something going on though that's causing OOM and a massive CPU spike. Anyone know anything? 4GB of RAM getting eaten for breakfast... just had to up it to 8
That's not happening on my own instance. Are you sure it's a leak per se? Could just be actual usage from volume.
I'm curious. How many users in that instance? I assume that an instance for a few people don't eat that amount of ram.
About 80 total, but I'd say only a handful are active every day
I have the same issue. I use multiple accounts across different servers, and get a different subset of comments across each one. Not sure the solution right now.
I'm glad I'm not the only one. I am commenting on this post from my own instance and it says it has 0 replies which I know is wrong. I did only federate with Selfhosted just a few hours ago. Who know how long it will take this post to get back to lemmy.world.
I see your comment.
Upvoting since I plan to run my own Lemmy instance too. Would be interested in force-sync
Thanks for pointing this out, I am just now realizing I am also missing comments. For example this post, I have 7 comments, but the actual post has 10.
Yep, it's because of the volume of comments/posts going through the big instances right now, and how many people are subscribed outside of those instances. It's too much for the server to keep up. I noticed over the last day or two my feed has mostly been Beehaw.org stuff because they don't seem to be having as hard of a time. It sure is looking like we need more separation like startrek.website, and maybe it was a bad idea to allow communities to just be made by anyone on lemmy.world and sh.itjust.works. Would've made more sense for all the users to be on these instances, but the communities hosted elsewhere.
Yeah, I definitely agree that open communities was/is a mistake. If a server is that large, it should be admin only. They can always run polls to gauge interest before making new communities.
Communities are the biggest problem. One solution would be to make community names more like a federation signature that only allows one community to have that exact name. Example: if lemmy.world has a community called !AITA, and you try to make a community on beehaw with the same !AITA name, it would not allow it and say "sorry that community name has already been registered by another instance".
Of course, the major issue with this idea is that one could defederate with lemmy.world, in this example, and make the community name anyway, then refederate. Who would keep the name? That's why I don't think this solution is ideal, but it's fun to share ideas and something to think about.
Try upping your federation workers
I’m seeing this issue primarily with communities hosted on lemmy.ml and lemmy.world. Admittedly, that’s a lot of the active ones at the moment, but for some of the smaller ones I follow, they seem to show up.
Its likely not a fault on your end, but overloaded servers on the other side failing to push updates or pushing them more that 10 seconds delayed after which your instance will discard them.
The latter part could be improved in the codebase I guess, but overloaded severs are just a reality right now, especially lemmy.ml
I believe the solution is to wait until 0.18 is released and everyone updates to it.
i am preparing to launch my own (likely single-user) instance and i'm not sure if i should wait until 0.18 proper or just throw the rc and see if that's enough.
Why wait? Can't you just deploy now and upgrade when .18 hits?
I have the same issue. Comments on some posts dont seem to update at all for me