Any reason not to combine those options?

Personal preference, but that’s certainly something being considered.

There are more services being impacted by the shutdown than just Lemmy so there are multiple factors influencing the final decision on what moves where and how it all shakes out.

I haven’t had time to dig into it deeper yet and didn’t want to leave things in a mess for long yesterday. Everything appeared to update successfully but when the UI was updated and restarted it had several problems that made it unusable. Anything that uses the API had no trouble reaching the site after the upgrade which leads me to believe there’s an issue specific to the webui and how our instance is currently configured/hosted. I don’t want to speculate too much without more information. The setup at our host isn’t typical for most Lemmy instances and occasionally we run into issues that don’t impact instances run from a single server or from another host.

And we’re back up on 0.19.11. That’s enough instability for one day. We’ll come back to this another day once we have the network issue sorted.

Sorry again for the downtime for those of you using the webui, and for the repeated restarts for our app users. As always thank you for your patience and understanding.

Unfortunately it looks like we’re going to need to roll back to 0.19.11 for now. We’re seeing some network errors when trying to access the site from the webui on the 0.19.12 build that are going to take more time to investigate. Apologies, these things happen from time to time.

It took a little longer than expected to come back online, but we’re back now and things look normal. Things might be a touch slow for the next couple hours as things sync up again. Long term performance should be much more stable with fewer dips in performance when new communities are added or a new instance discovers ours.

Thank you again for your patience. We’re doing our best to stay ahead of the performance curve so our members can have the smoothest experience we can offer.

You’re welcome! There’s a lot of learning as we go and some things will always be out of our control.

Keeping everything running smoothly has been more hands on than expected with a managed hosting service. We have a good working relationship with our host but they are a relatively small operation so fixes and upgrades aren’t as seamless as they probably could be. Communicating downtime windows could also use some work as we’ve recently seen. Still, they do a great job responding to our concerns and helping with things that require more heavy lifting than we’re comfortable with.

I hope people are finding it an enjoyable place to spend their time online despite the hiccups here and there.

This has been rescheduled yet again. Sorry for the on again off again updates.

A further update after more discussion with our host-

The major factor at play here is the increase in scrapers/bots/AI traffic to servers hosting federated content. There are large waves of bot requests that are pushing the postgres service to capacity across multiple customer sites within our host’s infrastructure. Our host blocks most scrapers automatically, and is actively working to block new bots to minimize the impact of the increased traffic as much as possible. Despite those efforts we’re still experiencing some problems here and there that we hope to have solved after the upgrades this weekend.

Again, thank you for understanding.

There have been a few short (less than 5 minute) service interruptions over the last couple days related to the coming upgrade at our host. We often aren’t given any warning if the downtime is expected to be very brief. Apologies if the experience is less than stellar. That isn’t what we want to provide.

Hopefully things will be running smoother than ever after this weekends maintenance.

The adjustments have been completed and everything looks good. The next expected downtime will be for our storage upgrades on June 14th.

Adjustments are done and everything looks to be back up as expected.

UPDATE: This has been pushed back one week by our host as there are a wider range of upgrades, mostly unrelated to our instance, being done at the same time to minimize downtime.

The new config wasn’t working as expected so we rolled back to our old email setup for now. Sorry for the minor disruption. Registrations are enabled again. We’ll come back to this project soon once we’ve got some of the underlying issues sorted.

The upgrades are complete and everything is back up and running. Please report anything out of the ordinary in terms of site performance.

Thank you.

Unpinning this since it’s been a couple days with no reported issues. If you run into anything just let us know.

And we’re back. Please report any issues you encounter.

We’re beginning the update now. The instance will go down in just a moment and be back in a few minutes. See you soon.