Fork me on GitHub
#xtdb
<
2020-04-24
>
Dieter06:04:27

Could crux be used as replacement for influxdb?

refset08:04:22

Hi, unfortunately it's probably not going to be a suitable replacement for the time being. InfluxDB uses all kinds of columnar compression & storage techniques, with various aggregation utilities. You've reminded me though that it's worth an FAQ entry for the difference between a timeseries db and a temporal db 🙂

Dieter09:04:33

Well I'm quite new to all this 😉. I'm trying to figure out where crux fits in. Is there a use case for small/hobbiest projects? Or is it more geared towards business and advanced applications?

refset10:04:39

I can sympathise! And time is very confusing... Crux is definitely good for hobby things. We have quite a few small internal apps running on it. It's pretty simple and reliable. I would recommend RocksDB + Postgres, or RocksDB + SQLite if you really don't want have to set up Postgres for "fun" 🙂

Jorin16:04:54

@U899JBRPF so you would recommend SQLite over RocksDB for a single node Crux setup?

refset21:04:29

@U8ZN5EHGU I mean SQLite for the event-log (specifically). Using RocksDB for the event-log as well as the main kv store is a good choice too, but I suspect many people would consider it more opaque

Jorin11:04:10

Yes, event-log was also what I had in mind :) less technologies involved seems like a good thing to me. Guess the operational characteristics are something that needs to be observed in the real world to tell what scenarios plays to the strengths of different setups.

👍 4