Can influxdb downsamples batch imported past-data with a continuous query automatically?
I have started to handle data from 10,000 sensor and 100,000 network devices for Machine Learning based anomaly-detection.
I want to process data like following;
- Create CQ, which aggregate minute interval data to hourly interval or daily interval.
- Import daily or weekly data from csv to influxdb with golang client.
- Downsample data by some trigger.
I want to downsample from data like this,
select count from data where time >= ‘2016-12-31T15:00:00Z’ and time <= ‘2016-12-31T16:59:59Z’
name: data
time count
2016-12-31T15:00:00Z 100
2016-12-31T15:01:00Z 100
2016-12-31T15:02:00Z 100
2016-12-31T15:03:00Z 100
:
:
:
2016-12-31T15:57:00Z 100
2016-12-31T15:58:00Z 100
2016-12-31T15:59:00Z 100
2016-12-31T16:00:00Z 200
2016-12-31T16:01:00Z 200
2016-12-31T16:02:00Z 200
2016-12-31T16:03:00Z 200
:
:
:
2016-12-31T16:57:00Z 200
2016-12-31T16:58:00Z 200
2016-12-31T16:59:00Z 200
to data like following;
select count from data where time >= ‘2016-12-31T15:00:00Z’ and time <= ‘2016-12-31T16:59:59Z’
name: data
time count
2016-12-31T15:00:00Z 6000
2016-12-31T16:00:00Z 12000
And, a continuous query that I have created is following:
name: cq_test
name query
sum_count_10m_cq CREATE CONTINUOUS QUERY sum_count_10m_cq3 ON cq_test RESAMPLE FOR 10m BEGIN SELECT sum(count) INTO cq_test.autogen.sum_count_10m_cq3 FROM cq_test.autogen.data GROUP BY time(10m) END
Is it possible for bulk-import to trigger/execute continuous queries or I have to aggregate it manually?
//Regards
Taisuke Sasaya