Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

For the first 3 years of Stream we used Cassandra. Afterwards we switched to a custom RocksDB + Raft solution. (somewhat outdated stackshare interview: https://stackshare.io/stream/stream-and-go-news-feeds-for-ov...)

The difference is massive. Cassandra was hard to manage and after many years of our team using it still had random spikes. RocksDB+Raft has been extremely solid, doesn't require any maintenance, load times are flat, zero spikes.

Cassandra was awesome, but it definitely has some issues. That's also why companies like ScyllaDB see space in that market. I wonder if AWS's cassandra implementation is better than regular cassandra.



> Cassandra was hard to manage and after many years of our team using it still had random spikes

A lot of that is getting better over time. A non-trivial percentage of contributions are coming from companies that are only looking to improve their operational story, not adding new features.

Most of the talks at NGCC (Next generation cassandra conference) focused on operational improvements - it's something we care a lot about (myself especially).


Those that don't want to roll their own RocksDB + Raft can consider using TiKV, which uses the RocksDB + Raft architecture. https://tikv.org/


Were your spikes all p999 impacts from GC?

I'm hoping rocksandra will solve a lot of that.


Something awesome is coming up for Cassandra 4.0


Can you add anything more specific? Something related to the storage engine?




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: