I have the following to display average latency.
It can be accelerated (vs. using the transaction command).
Now I would like to update this to "average latency per request type". Bu the requestType can be extracted only from the first event of each of the "transactions" below.
...stats min(_time) AS startOfTransaction max(_time) AS endOfTransaction by context_flow | eval transactionDuration=endOfTransaction-startOfTransaction | bucket startOfTransaction span=1h | eval startOfTransaction = strftime( startOfTransaction,"%m/%d/%Y %I:%M:%S") | stats avg(transactionDuration) as avgDuration by startOfTransaction | chart sum(avgDuration) as "Average Latency in Seconds" by startOfTransaction
Include first(requestType) as requestType in your stats calls to carry that field through the entire pipeline.
Include first(requestType) as requestType in your stats calls to carry that field through the entire pipeline.