Submitted 7 months ago by alphacyberranger@sh.itjust.works to selfhosted@lemmy.world
I tried self hosting it,but it felt very resource intensive on my vps. It’s a really good bookmark manager, feature rich and all. But I feel like it could have been lighter.
I have it running on an old Dell quadcore workstation with 16gb of memory, alongside around 30 other Docker services including a full Servarr stack, and haven't noticed any issues with it.
Runs smooth on my old qnap 453 :]
It’s intensive because is trying to archive the links and spends a lot of team doing so.
It’s more like a replacement for pocket.
Linkwarden has pretty minimal hardware requirements - it was tested on a VPS with 4gb of memory and it ran pretty smoothly, the most intense part is when you build the app, but once it’s running it’s relatively lightweight.
From their website. I wouldn’t consider tested on 4gb vps having minimal hardware.
I use Linkding and I am very happy with it. Less feature? Maybe. But it’s a bookmark sync. What do you need?
LOL if 4gb of memory is pretty minimal I am curious to know that they think to be resource intensive.
The maximum memory I am willing to allocate to a lxc container dedicated to bookmark management is 256mb
4gb isn’t so bad for a build, but etf does it need to run smoothly?!
I use Linkding too, it’s very light and it’s great, I only wish it had the exclude tag filter and no tag filter. That would make it perfect.
A bookmark sync. Thanks. I’ll check it out.
If you just want to sync bookmarks I don’t think linkwarden is what you want. Maybe floccus? I’m going to check out linkding that someone else mentioned because using git to sync floccus is broken on mobile platforms.
I guess my question would be what are your resources?
2 Gb ram and 2 CPU cores… I know it ain’t much, but I was not expecting a minimum requirement of 4 gb
Er, phones have had 4gb for years.
2gb for a system… My 2012 laptop has 4gb.
BinaryUnit@lemmy.world 7 months ago
Before I started to host a bookmark service o made some investigation, and the final dockerized contenders in 2023 for what relates to memory were:
Shaarli: (~ 50Mb of RAM ) Shiori: ( ~30Mb of RAM but lacks quite some features) linkding ( ~200Mb of RAM)
In the end i went with Shaarli