Re:Building columnar compression in a row-oriented database
by fabian2k on Thursday October 31, 2019 @04:27PM
(#21413365)
I really like Postgres as an allround database, and putting monitoring data into Postgres is an attractive idea to me. One thing I'm wondering is how well you can mix a regular Postgres database with timeseries data using TimescaleDB? If I have a regular application using Postgres, can I just add timeseries in there with TimescaleDB in the same database? Or is that either not possible, or simply a bad idea?
[ Parent
| on HN ]
Re:Building columnar compression in a row-oriented database
by RobAtticus on Thursday October 31, 2019 @04:38PM
(#21413478)
You definitely can mix the two, and we actually find it a very useful thing so you can have your metadata available to query alongside the time series data. TimescaleDB will only make a table a hypertable if you tell it too, otherwise regular tables should work as you'd expect. And since we appear as a table in the contexts that it matters (JOINs, querying, inserting, etc), you shouldn't have any issues mixing the two.
Re:Building columnar compression in a row-oriented database
by GordonS on Thursday October 31, 2019 @07:18PM
(#21414778)
I do this, and it works just fine - no caveats, it's actually one of the selling points of TimescaleDB.
Re:Building columnar compression in a row-oriented database
by jeltz on Friday November 01, 2019 @09:41AM
(#21418636)
Yes you can, and I think this is one of the main selling points of TimescaleDB.