Pretty minor but dev/test deps have changed a lot recently. I'd never heard of YugabyteDB before this PR landed!
Hah! Been a while since I heard that name. Story time... M'colleagues at Helpshift (@h0bbit, @kapil, @contact365) were evaluating Yugabyte, along with a bunch of other options, circa 2016/17 because, by then, our workloads had started breaking mongo in ways they couldn't fix (in mongo itself) or help us with (find workarounds). As early as 2015, mongo had proven itself to be an obviously bad idea (including losing our data for us), but pulling collections out of that thing was a heavy, heavy lift, for a whole bunch of reasons. Over the years, some collections were pulled out into ScyllaDB https://www.scylladb.com/2019/04/03/scylla-the-helpshift-story/ which @mourjo led, I think. And it looks like they finally ditched mongo circa 2020/2021, replacing it with Yugabyte: https://www.yugabyte.com/success-stories/helpshift-adopts-yugabytedb-distributed-sql-database/ No idea who did that work. But good riddance, I'm sure. They must have slept peacefully the day they finally trashed the old db.
At work, we leaned into MongoDB pretty heavily for a while, and while we never lost any data, we realized that it wasn't going to be cost-effective because we still needed MySQL for certain types of data, even if MongoDB was a good fit for everything else. So we migrated everything back into MySQL over about a year (including some collections that had half a billion documents), and then turned MongoDB off. Good riddance, indeed. I maintained CongoMongo during the time we were using MongoDB, then stepped down as maintainer once we stopped using it. I used to really enjoy all the one-day MongoDB conferences tho' and I have lots of free T shirts from that era 🙂