I have converted this below query to flux but it does not give me the correct answer.
SELECT sum(“value”) FROM ( SELECT non_negative_derivative(mean(“log_end_offset”), 1s) AS “value” FROM “kafka_consumer_group” WHERE (“topic” =~ /^$topic$/ AND “resourcegroup” = ‘clelastic’ AND “client_id” ==vm-kafka-mirrormaker) AND $timeFilter GROUP BY time($__interval), “topic”, “partition” fill(none) ) GROUP BY time($__interval), “topic” fill(none)
from(bucket: “myvc”)
|> range(start: v.timeRangeStart, stop: v.timeRangeStop)
|> filter(fn: (r) => r[“_measurement”] == “kafka_consumer_group”)
|> filter(fn: (r) => r[“_field”] == “log_end_offset”)
|> filter(fn: (r) => r[“topic”] == “to_metrics_shared” and r[“resourcegroup”] == “clelastic” and r[“client_id”] == vm-kafka-mirrormaker)
|> aggregateWindow(every: v.windowPeriod, fn: mean, createEmpty: false)
|> derivative(unit: 1s, nonNegative: true)
|> mean(column: “_value”)
|> group()
|> sum()
This flux query doesn’t provide me the exact result, can some please help me on this. ?
Hello @Murugan_P,
Welcome!
Thanks for your question.
Hmm your query loooks good to me except for it looks like you have an extra mean. You’re already taking the mean with the aggregateWindow() function. I don’t see you taking the mean of the derivative with your influxQL query. Can you please try:
When making translations I recommend trying to get the subqueries right first and translating the queries bit by bit instead of making the translation all at once.
@Murugan_P
Awesome to hear! Thank you!
PS what are you using InfluxDB for? It really makes my day to learn about what the community is doing with InfluxDB.
Thank you!
Currently the query is displaying the topics that i have selected in the influxdb dashboard. I want another query that displays all the topics in the certian time frame. I don’t have to type the exact topics in the query.
