Continuous Aggregates - User feedback

[ramonguiu]

Promscale continuous aggregates leverage a TimescaleDB feature of the same name to have the database manage downsampling and data materialization.

Read more details in the launch blog post and the documentation.

The feature is currently in beta and we are using this discussion to capture user feedback.

You are also welcome to join our Slack community and post questions in the #promscale channel.

[toymachiner62]

I've tried to follow the blog post and i'm hitting road blocks left and right. The functions don't work unless you prefix with toolkit_experimental and then if I do that I get an error that aggregates cannot use named arguments.

Here's a link to my issue in slack https://timescaledb.slack.com/archives/C4GT3N90X/p1635305238080800

[ramonguiu]

Thanks for the feedback @toymachiner62 . What version of the toolkit are you using? You can use this query: SELECT extname, extversion FROM pg_extension;
I think you may be using an old one.

I can resolve the error if I remove bounds => from the query. I'm checking internally why we are using it and I'll provide an update.

[toymachiner62]

plpgsql	1.0
promscale	0.1.2
timescaledb	2.4.2
timescaledb_toolkit	1.1

We're using managed timescaledb (as opposed to timescale cloud I think)

[toymachiner62]

This view creation does not work as documented in the blog post. It's missing the AS

CREATE OR REPLACE VIEW cpu_usage_30m 
SELECT
    bucket + '30 min' as time,
    series_id,
    extrapolated_delta(cnt_agg, method =>'prometheus') as increase,
    extrapolated_rate(cnt_agg, method => 'prometheus') as rate,
    irate_right(cnt_agg) as irate,
    num_resets(cnt_agg)::float as resets
FROM cpu_usage_30m_aggregate;

[toymachiner62]

Also after creating that view the query doesn't work as described in the blog post

[ramonguiu]

@toymachiner62 sorry for the delay and thanks so much for your feedback. We have updated the blog post and you should be able to run those queries now.

[mwolter805]

Hello, How would you recommend aggregating a Prometheus metric that only has a binary state? On and off, or actually 0 and 1.

Thanks!