You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
It was easy to get counts of the number of slow-requests-per-day over the prior 7 days, for example here's what that looks like for all issues+prs across all orgs on 4th June:
[
[Fri, 29 May 2020, 10],
[Sat, 30 May 2020, 1],
[Sun, 31 May 2020, 3],
[Mon, 01 Jun 2020, 27],
[Tue, 02 Jun 2020, 4],
[Wed, 03 Jun 2020, 0],
[Thu, 04 Jun 2020, 0]
]
Couple of questions arise:
Do you want the prior 7 days from today, which will include two days that will always be zero as they aren't 48 hours old yet, or do you want the 7 days after the 48 hour slow response window has passed?
Are you sure you want the median (the middle number) which in this case would be 27, whilst the mean would be 6.42
If I look at the numbers shifted back by 48 hours we get:
[
[Wed, 27 May 2020, 17],
[Thu, 28 May 2020, 10],
[Fri, 29 May 2020, 10],
[Sat, 30 May 2020, 1],
[Sun, 31 May 2020, 3],
[Mon, 01 Jun 2020, 27],
[Tue, 02 Jun 2020, 4]
]
Here the median is 1 and the mean is 7.2.
We can of course scope these calculations to score for individual orgs, repos, languages etc
Not sure whether this should be on the "last week" view or the slow requests view, but we need to know what the rolling median is for slow requests.
Eg, on any given day there are zero or more requests which have not had a response in the last 48hrs.
What's the median for slow-requests-per-day over the prior 7 days.
The text was updated successfully, but these errors were encountered: