Not going as smoothly as I would have liked
It is back up. Sorry about the downtime.
Every opportunity is gold to learn more about the system and how to make it more reliable, right? Any backend improvements deriving from this “outage”?
Not yet because I had to roll everything back. This is a big part of the separation from Strava work. Nothing visible to users (other than bug fixes and workout builder changes described elsewhere) but the site will be faster.
Faster is already a Customer Facing improvement (if measurable, even better! :P)
Surely though, the fact that you had to rollback a migration means you learned something about the infrastructure, identified an underlying/unexpected issue, which is in turn going to be useful to better next migrations…?
Well, it is noticeably faster!
Ah yes. I learned that its better to run the same version of the database on your laptop as in live because otherwise db migration scripts might not work So tomorrow morning I need to try upgrade the database.
Oh yeah. As the joke goes: “Everybody has a testing environment. Some people are lucky enough enough to have a totally separate environment to run production in.”