Well, we're accessible over IPv6. We can't actually talk to other IPv6 instances because of docker-compose nonsense.
Looks like I'm gonna have to start preparing for a migration from docker to a proper VM.
π I was sooooo close.....
Well everyone, I think Whitespashe is now available over IPv6! Does anyone know of any IPv6 only Fedi Instances I can try out?
Is there a method of culling servers that have broken SSL/Excessive timeouts from my instance?
I know `tootctl accounts cull` does something similar, but surely there's a way for mastodon to go: "That instance is dead. I'm cutting all ties with it."
Today I implemented a master/master MySQL setup, and integrated it into my HAProxy install.
And nothing is on fire!
Yesterday I successfully deployed backup mail servers in my homelab AND in the cloud.
This is the year of high-availability (or at the very least: basic redundancy) folks!
Ugh! Spent 2 days troubleshooting media issues with proxying my s3 media storage for mastodon.
I'm pretty sure I've reverted back to my original config (the one I thought I had issues with), and now it's all okay again!
This weekend has been sent to try me!!!
And now Whitespashe is running on Mastodon 3.2.0!
We're now as up-to-date and efficient as we can reasonably be! Time for a well-earned coffee!
After watching the slow loading of assets via backblaze, I've now implemented an nginx cache and haproxy wizardry to try and speed stuff up somewhat!
@angristan contributed to my success via his blog post: https://stanislas.blog/2018/05/moving-mastodon-media-files-to-wasabi-object-storage/
After following that, I then stole a load of code from the mastodon docs and https://blog.sentry.io/2017/03/01/dodging-s3-downtime-with-nginx-and-haproxy
Now I've got a fairly resilient media hosting stack!
My hard drives will surely thank me! Backups too!
I've updated Whitespashe to use backblaze for its media storage! Hopefully I'll have a much less sad hard drive now!
Ugh. Had so many issues with networking yesterday. I lost internet for a fair part of the day, my VM hosts would not stay clustered or in sync, either!
I had to completely redesign the network connectivity at home and finally for sorted at about 1am today!
Things seem stable now, but yesterday was sent to try me. π
Whitespashe is going offline for the 3.1.4 update!
Whitespashe is going offline whilst I bring us up to the most recent version. It's gonna get wild.
Whitespashe.uk is now running on Mastodon Master branch, as the v3.1.2 release can't be built due to a dependency being yanked.
We're bleeding edge guys!
Ugh. I've had to disable full-text searching, as my VM is unable to handle the high memory requirements. Looks like a hardware upgrade is in order soon. π₯
Some memory issues on whitespashe.uk
Rebuilding my elasticsearch db, as it hasn't been running for a couple of months.
Friendly Neighbourhood Sysadmin.