Has anyone used a subquery along with difference or derivative such as
SELECT difference(sum(“value”)) FROM (SELECT MAX(“value”) as value FROM “duration” WHERE $timeFilter GROUP BY bucket,api) GROUP by time(1m), bucket
The data is basically a set of buckets for times.
time bucket api value
===== ====== ======= =====
xxxx 1 /api/xxx 10
xxxx 1 /api/xxx 3
nnnn 1 /api/yyy 0
nnnn 2 /api/yyy 20
Each measurement is a counter of the number of API calls that met the bucket criteria (response time). Granted the counter can reset to 0 and I will likely use non_negative_difference but not before I can get difference, derivative or anything to work
The query w/o the difference works fine.
SELECT sum(“value”) FROM (SELECT MAX(“value”) as value FROM “duration” WHERE $timeFilter GROUP BY bucket,api) GROUP by time(1m), bucket
returns data fine and correct however:
SELECT difference(sum(“value”)) FROM (SELECT MAX(“value”) as value FROM “duration” WHERE $timeFilter GROUP BY bucket,api) GROUP by time(1m), bucket
returns an empty result set each time. Similar queries without the subquery work fine however I need to get the last value for each bucket x API and then sum them together to get the correct values to use.
Had anyone used difference with subqueries and grouping ?
Thanks,