The day that I’ve worried about has finally arrived…
…skyhoppers.club is dead.
I didn’t make the decision lightly, and honestly, I wish I could’ve handled it better. I think there were maybe… two? three? other active users besides me, and even they seemed to hardly use their accounts, but I certainly didn’t want to just pull the plug without notifying those users. And if you’ll allow me to half-quote Han Solo: “It’s not my fault!” I mean. It certainly is somewhere down the line, but let me elaborate.
I started my Fediverse journey on mastodon.lol back in like… 2022? but that got weird, so I moved over to tech.lgbt and enjoyed my time there! But the amateur sysadmin inside me was like “Mastodon is cool… what does it take to self-host it?” I certainly didn’t want to expose my home network to the things that could show up on Mastodon (or pay for the potential added bandwidth, for that matter), and Digital Ocean had a “one-click” deployment option, so I spun up a droplet and got started! It took me maybe two or three days, but at the end of all the work, I had a mostly-functioning instance! I never did get audio files to play from my timeline, but that was the only issue that never really got ironed out. Oh, except for Digital Ocean blocking all SMTP traffic by default and not really letting you enable it no matter what. But as I mentioned above, there were only a few users and they all got in before DO changed that rule, so it ended up being fine.
And for two years, things were mostly fine! I was able to use skyhoppers.club like any other instance with little to no problem. I had a few issues with the odd update here and there, but I was always able to figure it out and get things back up and running within a day or so.
However, with the release of 4.3.9 or thereabouts, I had a TON of issues with the update. I thought I followed all of the instructions, and it even looked like a pretty standard update. But something went sideways somewhere, because I ended up running a 4.4.0-alpha build of Mastodon and was never able to get it back to stable. I was switching git branches and shit at lightspeed trying to figure out what I’d done wrong, but eventually it just kind of… came back. I don’t even remember what I did. This was the beginning of the end.
I got a few emails notifying me of new releases, and I kept putting them off because I had a sinking feeling that something was fundamentally broken on my instance and it was waiting to come out and bite me as soon as I tried to do another update. And sure enough, that’s exactly what happened! Last week I decided to bite the bullet and attempt an upgrade, and it went just as poorly as I’d anticipated. I had Ruby version issues, gems that wouldn’t install for some reason that was lost on me because I don’t know Ruby, and all of my other troubleshooting attempts failed. It was at this point that I just decided that I didn’t want to do this anymore.
I had noticed that the instance was getting a bit more expensive with each passing month, but it was always affordable so I didn’t put too much stock in it. But once the cost eclipsed $50USD per month, I finally decided to actually look at my invoice. Turns out that I had misconfigured cache settings on skyhoppers.club and was storing basically everything that was loaded onto my instance in cache, which caused my storage costs to balloon. This and the resulting settings update to fix the cache problem actually happened before the instance died, but I felt stupid enough about it that I was already kinda feeling defeated when everything broke.
Anyway. Sorry to anyone that’s finding out like this that things broke and won’t be coming back. Thanks to anyone that spent any time on skyhoppers.club and letting me feel like a real sysadmin for a few years! Anyway. Gotta get back to work now.