Fork me on GitHub
#sql
<
2020-01-31
>
kulminaator19:01:47

btw. some feedback to cockroachdb ... gave it a few hours of testing , it actually feels pretty nice. postgresql driver's just worked. and the db feels like an elasticsearch cluster (you can add and remove nodes at will and the system nicely manages the data according to your changes). not as performant on a single node as postgresql, but i wasn't expecting that anyway.

kulminaator20:01:09

probably comes with a similar set of problems too (compared to elasticsearch and other clustered storage systems - split brain, awkward data relocation etc.) 🙂

adamfeldman19:02:49

IIUC CockroachDB handles data relocation automatically — check out their detailed blog and docs. In testing I’ve found it very robust (and they also did a Jepsen analysis — https://jepsen.io/analyses/cockroachdb-beta-20160829)

kulminaator20:02:55

elastic does it automatically as well, but sometimes this backfires badly 🙂

kulminaator22:02:21

but i agree based on initial testing, it felt that it handled weird happenings ok

kulminaator20:01:59

i have had my eye on yugabyte too which has similar properties and is also an sql database. if anyone has experiences with that one, including accessing from java/clojure, i'm all ears 🙂