I'm going to investigate today. The timing is suspiciously close to when the backup process runs overnight, so that will be the first place I investigate.

[–] Axisential 2 points 1 year ago (1 children)

Out of interest, is this instance running on your own hardware or virtualised? If the former, what's the bandwidth used? Just curious!

[–] Dave 1 points 1 year ago* (last edited 1 year ago) (1 children)

So in the early days (waaaay back in June 2023) it was run on a commercial VPS (a $2/month one), but I got an offer from @idanoo@lemmy.nz to have it hosted on the hardware being used by fediservices.nz, which hosts a bunch of federated sites. They have a server set up in an Auckland data centre, with full access to install the hardware they like. It runs proxmox and allocates us a virtualised server to run lemmy.nz.

We have cloudflare sitting in front, and cloudflare reports 15GB total data served over the last 24 hours. 27k unique visitors made 930k requests. There's some weird stuff though that makes it hard to judge exactly how many people. For example, the join-lemmy.org/instances page hotlinks to the instance logos, including ours. So if they get a million hits, that loads our logo a million times, so registers on cloudflare as a million requests. And people on other instances looking at news articles we posted will have the article image loaded from our cache which will count as a visit/request. Stuff like that makes it hard to know who is directly accessing lemmy.nz.

Previously we have had much higher bandwidth usage, but I think this was largely from the instances page being loaded a lot from lots of reddit users looking for servers.

[–] Axisential 2 points 1 year ago

Outstanding! What a great setup

[–] flambonkscious@sh.itjust.works 2 points 1 year ago (1 children)

Bummer dude! I'm grappling with a failing hard disk (the data is fine), but just the sheer amount of shitting around to prove it is or isn't my recent half baked theory is something I just don't have the emotional energy for.

As much as I'd like to, ain't no way I'm self-hosting this stack!

[–] Dave 1 points 1 year ago (2 children)

My personal stuff is all hosted on a Raspberry Pi with an SSD attached, and the SSD failed earlier in the year (as in completely dead). I was trying to troubleshoot, and switched out the hard drive for another one I had and it still didn't work. I ended up going down the wrong track trying to solve it because what are the chances of two dead hard drives? Anyway, long story short, I had two dead hard drives. Neither of them had SMART errors, though when I was searching around at the time I remember reading that Google reported about 60% of hard drive failures had SMART errors, so it was very common for them to not have any indication of failure before it happened.

At least with Lemmy, there are a bunch of tech savy users to ask for help 🙂.

[–] flambonkscious@sh.itjust.works 1 points 1 year ago

Wild but in a bad way - not what I signed up for

I'm getting old and jaded - this was not in the plan?!?

[–] Dave 2 points 1 year ago (1 children)

Last night I played Jaipur with my wife. It was our first time playing it, and she won both games and all but one round. So if anyone has any tips 😆