Major Description
I think UI should show current values for the following query metrics in server UI:
query.n1ql_avg_req_time
|
query.n1ql_avg_svc_time
|
query.n1ql_avg_response_size
|
query.n1ql_avg_result_count
|
Showing cumulative values for those metrics doesn't make much sense.
These are derived metrics that are calculated on the fly by ns_server the following way (the example shows req_time, other stats are calculated likewise):
(sum n1ql_req_time across cluster) / (sum n1ql_requests across cluster)
|
In order to show current value we should apply rate function to req_time and to number of requests (because rate shows change of the counter per second = number of new request):
(sum rate(n1ql_req_time) across cluster) / (sum rate(n1ql_requests) across cluster)
|
