Hacker News Reader

We ran this experiment Vs influx some time ago for ingestion here https://medium.com/questdb/speeding-up-influx-line-protocol-...

In terms of queries, see our previous post on HN to get a sense https://news.ycombinator.com/item?id=22803504

View on HN

Ingestion rate is pretty good in Influx and we are happy with that for now. If QuestDB is better I definitely won't complain. The issues with Influx is data integrity and also the difficulty in manipulating time series data if that becomes an issue (e.g turning all "0" string values into a 0i). How does QuestDB fare in this regard?

We take data integrity very seriously. This is one of the reasons QuestDB actually exists. We haven't had reports of data corruption or loss yet. We can commit durably if needed too.

To manipulate data we support "add" and "delete" column on the fly. We can also add column replace and type change if needed. This is pretty easy to do.

PostgreSQL wire is in beta. It works with JDBC driver and we will add metadata support quite soon.

Reply on news.ycombinator.com