Yup. What I don't get is why they don't do a dry run, of the upgrade process, on a cloned test server. Doesn't even have to host all the same content as the main voat server: simply just enough to be useful. It's, like, what every other website in the world does. I don't even remember reddit going down this often, when they'd roll out updates, back in the day.
Reddit was a shitshow of downtime. I ‘member. I think what you’re suggesting is what was done but there were only a couple dozen users testing it
Ummm, you mean like the preview voat site that has been used for testing for a couple years now?
Not really. Because that's the site he works live on. I mean set up a dummy voat, exact clone of live voat. Then run the upgrade script on it.
He develops on the preview site and then updates to the current version on that site, repeats until he's ready to go real. There gets to be a point where as much testing as can be done with a limited userbase and then needs the real Voat to dig any deeper. I dare say he and the developers have it handled better than either of us could.
It's the load, not the upgrades themselves that cause the downtime most likely ... then combined w/ unexpected weirdness you just can't duplicate w/o real-world load as well.
This is probably the case. People trying to access the site while the upgrade is happening, throwing exceptions all over the place.
(post is archived)