Hi.
I have the following retention policies on my Telegraf database.
name duration shardGroupDuration replicaN default
---- -------- ------------------ -------- -------
autogen 168h0m0s 12h0m0s 1 true
30days5min 720h0m0s 24h0m0s 1 false
and use this continuous query to downsample the data:
name: telegraf
name query
---- -----
cq_5m_avg CREATE CONTINUOUS QUERY cq_5m_avg ON telegraf BEGIN SELECT mean(*) INTO telegraf."30days5min".:MEASUREMENT FROM telegraf.autogen./.*/ GROUP BY time(5m), * END
This is working perfectly except for two, of the in total 50, measurements are missing from the 30days5min RP
I haven’t found any errors in the logs, I’ve tried dropping and recreating the CQ, but to no avail.
Any hints or tips would be highly appreciated.
Thanks,
Morten
UPDATE:
To try and narrow it down a little. Running this query manually works:
SELECT mean(*) INTO telegraf.oneyear.cloudwatch_aws_rds FROM telegraf.autogen.cloudwatch_aws_rds GROUP BY time(5m), *
Running the exact same query in a CQ does not:
CREATE CONTINUOUS QUERY rds_5m_avg ON telegraf BEGIN SELECT mean(*) INTO telegraf.oneyear.cloudwatch_aws_rds FROM telegraf.autogen.cloudwatch_aws_rds GROUP BY time(5m), * END