Could be what communities you are subscribed to. I run a small instance with about 3ish users, and here are my stats after about 3 months as well:
9.5G ./pictrs
12G ./postgres
8.0K ./lemmy-ui
What version of lemmy are you using? A recent update also introduced some space savings in the database (I think).
I think this is part of the recommended (external) nginx configuration for lemmy:
limit_req_zone $binary_remote_addr zone={{domain}}_ratelimit:10m rate=1r/s;
Which can be found here
There is a matrix chat: https://matrix.to/#/#lemmy-admin-support-topics:discuss.online
POSTs are how federation works (ActivityPub is a Push-based protocol). When you “subscribe” to a community on say lemmy.ml, you are telling it to periodically send you updates about that community. This comes in the form of POSTS.
As to the frequency of the POSTs, I can imagine something like lemmy.ml having a lot of activity that it needs to inform your instance of (new votes, new comments, new posts, etc)… but I’m not sure if one request per second is reasonable or not.
For Lemmy, I primarily use Photon and have contributed a little bit to the project.
Familiarity (my client distro is Pop and is based on Ubuntu), and I like the LTS life cycle (predictable).
I do uninstall snaps, though, and mostly just use Docker for things. I could use Debian, but again, for me it was about familiarity and support (a lot more Ubuntu specific documentation).
Currently self-hosting my own mastodon server and honestly the setup wasn’t too bad (using docker)… much more straight-forward than I feared.
My main concerns, which Julia mentions, is that if you have a small instance, you are very much an island as the way federation work is not what you expect. For instance, as Julia notes, if you view a new person’s profile on your own instance, it will look empty (as if they haven’t posted anything). Lemmy also has this issue if you view a community you have not subscribed to yet for the first time.
Likewise, my “#explore” tab is basically always empty and discovering new tags or people is difficult if you are just looking on your own instance (I basically have to go to Fossotodon or another instance to find new things and then import them into my own instance). I’ve recently learned that you have to have a third party application basically seed your instance with posts… again, similar to the bot tricks use for seeding Lemmy with communities.
Overall, I think discovery is a big pain point for the fediverse and ActivityPub. It’s great that we can have our own instances and control our own small communities, but it seems that we are lacking the ability to really connect across instances and form experiences that really bridge across multiple communities.
Twitter and Reddit may have only lost a few million users to Mastodon and Lemmy so far, but these are nation-sized numbers, comparable to what Scandinavia is to the United States of America. The incumbents have allowed the fediverse to reach critical mass. It’s only gonna get bigger, but it already works well enough that I’ve no need for any other social network. It’s nicer here.
This resonates with me. Although they are still lacking for the long tail of small niche communities, Lemmy and Mastodon now have enough people and content that I rarely find myself going to Reddit or Twitter. The fediverse is not perfect and there is a lot of room for growth, but it is now large enough to be viable and hopefully sustainable.
Do you have a searxng
folder in the same folder as your docker-compose.yml
? If so, perhaps it is not mounting inside the container properly.
I think this is missing an article link: https://www.phoronix.com/review/downfall
Downfall, or as Intel prefers to call it is GDS: Gather Data Sampling. GDS/Downfall affects the gather instruction with AVX2 and AVX-512 enabled processors. At least the latest-generation Intel CPUs are not affected but Tigerlake / Ice Lake back to Sandy Bridge is confirmed to be impacted. There is microcode mitigation available but it will be costly for AVX2/AVX-512 workloads with GATHER instructions in hot code-paths and thus widespread software exposure particularly for HPC and other compute-intensive workloads that have relied on AVX2/AVX-512 for better performance.
Rough day for CPU makers…
Update: Of course there is a dedicated page for it: https://downfall.page/
No, but basically jmp.chat takes over your phone number… it acts as your carrier for voice and SMS (similar to Google Voice). Maybe not exactly what you want.
From the FAQ:
You can use JMP to communicate with your contacts without them changing anything on their end, just like with any other telephone provider. JMP works wherever you have an Internet connection. JMP can be used alongside, or instead of, a traditional wireless carrier subscription.
The benefit of this is that you can receive voice and text on anything that can serve as a XMPP client.
You could consider using something like jmp.chat. It delivers SMS via XMPP (aka jabber), so you could self-host a XMPP server and receive SMS that way. It also has some support for MMS (group chat, media), but my experience with it was mixed (I used it for about 3-4 years).
I wonder if it is because of the various outages on both instance and the new “dead instance” detection, lemmy.ml has temporarily stopped receiving updates?
The federation code now includes a check for dead instances which is used when sending activities. This helps to reduce the amount of outgoing POST requests, and also reduce server load.
Yeah, I got a cordless stick vac a few months ago and it has been a game changer, particularly for the stairs. It is powerful but also lightweight enough that my kids can use it (and thus help around the house) effectively.
Also joined the club today :)
Ugh. I shared this with my wife and she is really saddened by the news. We have over a decade of historical data on Mint and she really values the insights it has been able to give to her in planning for a family.
Hopefully, the community can suggest some viable alternatives :|