Has anyone use YugabyteDB or CockroachDB in production?

Question

We are currently looking at deploying a CockroachDB or YugabyteDB cluster for some applications, and I was wondering if anyone had used it in anger?Not a lot of detailed comparisons between the 2 (that aren't written by the companies in question)

DLA · Accepted Answer

We used CRDB (and still do) in production and a scale (billions of records+) and have found it to be very good. While it does not have every advanced feature of PostgreSQL, you get very easy clustering and operational simplicity which drive reliability. The command like tools that come with it are also fantastic. Impressive database that&rsquo;s been a joy to work with. Sorry, no direct experience with YB.

travisd · Answer

Didn't make it to prod.I tried migrating to YB at one point. I had issues migrating the indices (got lots of timeout issues). Once I got past that, I found out that a particular query I needed to do didn't have predicate pushdowns implemented and so always resulted in a full table scan. Ended up giving up.Cockroach (at the time at least, ~1.25 years ago) didn't have support for some JSONB things that I needed.

gigatexal · Answer

We used CRDB 2.1.4 and then on the YYYY versions when they changed to a date based version scheme in production at a fintech. This must have been a year or so ago. I've since left. I think they might still be using it. Back then it had issues with SQLAlchemy and we had to do much in teh way of changing our application to be more cloudy in the sense of being able to retry transactions and such.I don't think we trialed YugaByte at the time because an ORM story was a dealbreaker and CRDB at least advertised support.