[Alerts] [HealthAvailability] Streams perMinuteRequests and perMinuteResponses gets events with a wrong timeout window

[rmsamitha]

Description:
In APIMAnalytics-RequestSummarizer-RequestSummarizer-realtime1.siddhiql, Streams perMinuteRequests and perMinuteResponses receives events with a uniqueExternalTimeBatch window which has lesser timeout value than the window.

This occurs when events are inserted into perMinuteRequestsTemp stream from perSecondRequestsWithUniqueId stream. Time Window is defined as 60 seconds but timeout value is 10 seconds. So events are flushed into perMinuteRequests stream multiple times within 60 seconds. So time window of 60 seconds gets neglected.
[This occurs similarly when events are inserted into perMinuteResponsesTemp stream from perSecondResponsesWithUniqueId stream ]

This bug causes the health availability alert's request/response count percentiles related triggering cause. Pattern count can gets incremented more than once in each minute and so alerts get triggered by lesser number of occurrences, than expected.

Affected Product Version:
2.0.0, 2.1.0, 2.2.0