3 ms·
Congrats on the release! I'm in the process of migrating from influxDB to a combination of Kafka and TimeScaleDB. Has anyone beaten me to the lunch with testi
by LogicX 10y ago
Congrats on the release!
I'm in the process of migrating from influxDB to a combination of Kafka and TimeScaleDB.
Has anyone beaten me to the lunch with testing if it works with [PipelineDB](https://www.pipelinedb.com/ https://www.pipelinedb.com/)? I can see that being a powerful combination, perhaps also tied to their forthcoming [Stride](https://www.stride.io/ https://www.stride.io/) analytics API.
- RobAtticus 10y agoAwesome! Just so you're aware, upgrading between versions of TimescaleDB at this point is not fully supported. Meaning, we are still breaking some non-trivial things between release and don't have upgrade scripts. I don't want to discourage your usage, but I'm letting you know about the potential pitfalls. We hope to start supporting smoother upgrade transitions in the near future.
- zabomber 10y agoOut of interest, why are you moving away from InfluxDB? We're looking at moving to it....
- LogicX 10y agoWe can't pull stats we need. We tried: so many continuous query rollups... Eats into ability to write data... at only 1000/sec (artificially tripled due to splitting into multiple measurements to get around cardinality problem) Dropping writes even on 24 core 500gb ram, SSD Dedicated. Can't scale when also used to query against. Basically: their claimed performance looks great; but real works doesn't match up. They've pushed fixing cardinality two quarters. No reply on my github issue with broken top() Compression is good... is about all I can say lol