You want the FOR clause, which will cause the CQs to recalculate prior
intervals and pick up any newly written points in those intervals.
https://docs.influxdata.com/influxdb/v1.0/query_language/continuous_queries/#advanced-syntax.
E.g.

CREATE CONTINUOUS QUERY foo ON db
RESAMPLE EVERY 20s *FOR 2m *
BEGIN
  SELECT median(aa) AS aa INTO foomeas FROM barmeas GROUP BY time(20s),*
END

Note that the GROUP BY time offset just moves the beginnings of the
buckets, delaying the query execution time is just a side effect. I've
removed it from the query.

On Wed, Oct 12, 2016 at 10:43 AM, Michael C <[email protected]> wrote:

> I was wondering if someone out there could offer some advice on setting up
> a CQ for my system.  It would appear CQs might not be powerful enough to
> accomplish what i'm after, but im still new-ish to influx so maybe im wrong?
>
>
> I have a system where multiple nodes out in the world transmit packets of
> data once per minute, that contain 60 samples each, 1 for each second.  The
> data is received and stored in influxdb.  Note that this packet of 60
> samples from a given node could be received at any time over the minute.
> Thus, from influx's perspective there could be up to 1 minute of latency
> before receiving the next 60 seconds of data with respect to the start of
> each new minute.
>
> With this set up, I would like to have a CQ perform a 20second median on
> the second by second data.
>
> However it would seem that the latency of my system poses a problem to CQ
> usage. It would seem that CQ's  need to execute within the vicinity of
> now() and  for my case would occur on a 20second clock boundary.  Given my
> setup, there would be many cases where the data simply hasnt arrived yet
> when the query ran due to my 1 minute latency.  I tried adding an offset in
> the GROUP BY time(), but this doesnt seem to work (as expected after
> reading the docs, https://docs.influxdata.com/
> influxdb/v1.0/query_language/continuous_queries/)
>
>
> Below is an example of one of the many queries I have tried:
>
> CREATE CONTINUOUS QUERY foo ON db RESAMPLE EVERY 20s BEGIN SELECT
> median(aa) AS aa INTO foomeas FROM barmeas GROUP BY time(20s, -60s),*  END
>
>
> Is there any way to make CQs work with my system given the latency?
>
> I suspect it would work if I were computing 5 minute medians as I could
> then offset by -1m, but that is not what i'm after.  Or maybe im confused.
>
> Any advice would be much appreciated!
>
> Mike
>
> --
> Remember to include the version number!
> ---
> You received this message because you are subscribed to the Google Groups
> "InfluxData" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to [email protected].
> To post to this group, send email to [email protected].
> Visit this group at https://groups.google.com/group/influxdb.
> To view this discussion on the web visit https://groups.google.com/d/
> msgid/influxdb/7574ae65-fdcf-4417-b299-d72471d79ff8%40googlegroups.com
> <https://groups.google.com/d/msgid/influxdb/7574ae65-fdcf-4417-b299-d72471d79ff8%40googlegroups.com?utm_medium=email&utm_source=footer>
> .
> For more options, visit https://groups.google.com/d/optout.
>



-- 
Sean Beckett
Director of Support and Professional Services
InfluxDB

-- 
Remember to include the version number!
--- 
You received this message because you are subscribed to the Google Groups 
"InfluxData" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To post to this group, send email to [email protected].
Visit this group at https://groups.google.com/group/influxdb.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/influxdb/CALGqCvNP2uSHa1phrOpQKxojTWYcoTs7cq8d4RD9-a9dKRjV0A%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.

Reply via email to