I will not be able to take a look at this dashboard before Saturday (09/15) or Sunday (09/16).

@Addshore @Lea_WMDE @Franziska_Heine

The following:

When the failure type "All" is selected the top right single stat panel does not work. This probably needs a sumSeries call to add the series together then multiple are selected.

is now hopefully fixed by using

summarize(sumSeries(MediaWiki.FileImporter.specialPage.execute.fail.$FAILTYPE.sum), '1d', 'sum', false)

in place of

summarize(MediaWiki.FileImporter.specialPage.execute.fail.$FAILTYPE.sum, '1d', 'sum', false)

as suggested by @Addshore.

I have to admit that I am still experiencing confusion when I need to use sumSeries() across metrics whose names end in .sum.

@Addshore @Lea_WMDE @Franziska_Heine

As of the following:

Max import time reported lower than the median import time in single stat panels. Currently the max is reported as 6k but the median is 9.7k (see in F25749476)

At the present moment (16:41 CEST 09/16/2018) the dashboard reports all four Import Time reporting values (min, max, median, and mean) to be the same.

Now, here are the queries that I've used in the respective singlestats; as you may observe, the only difference between the queries is in the usage of the respective aggregation (min, max, median, average).

MINIMAL IMPORT TIME

summarize(MediaWiki.FileImporter.import.timing.wholeImport.sum, '1d', 'min', false)

MAXIMUM IMPORT TIME

summarize(MediaWiki.FileImporter.import.timing.wholeImport.sum, '1d', 'max', false)

MEDIAN IMPORT TIME

summarize(MediaWiki.FileImporter.import.timing.wholeImport.sum, '1d', 'median', false)

MEAN IMPORT TIME

summarize(MediaWiki.FileImporter.import.timing.wholeImport.sum, '1d', 'average', false)

Now, in the Options (metrics tab), I've switched the Stat value back from Average to Current - I would say that should to be the correct setting here. However, I am still getting all four values the same. NOTE: By changing the Stat in Options (metrics tab) one does obtain different numbers that make sense: min, max, average, for example; however (a) it is unclear how does the usage of this parameter interact with queries that are already written out so to summarize by min, max, mean, and median, and (b) this value cannot be set to median, and we need to show the median import time.

I have then changed the value of the alignToFrom function parameter from false to true: no change at all. This change was made in accordance with the following observation from the Graphite functions overview (see under summarize):

By default, buckets are calculated by rounding to the nearest interval. This works well for intervals smaller than a day. For example, 22:32 will end up in the bucket 22:00-23:00 when the interval=1hour. Passing alignToFrom=true will instead create buckets starting at the from time. In this case, the bucket for 22:32 depends on the from time. If from=6:30 then the 1hour bucket for 22:32 is 22:30-23:30.

Then I have experimented with the MAXIMUM IMPORT TIME singlestat and changed the complete query to:

consolidateBy(MediaWiki.FileImporter.import.timing.wholeImport.sum, 'max')

(inspired by a recent @Addshore's blog post on Grafana)
getting exactly the same number as when

summarize(MediaWiki.FileImporter.import.timing.wholeImport.sum, '1d', 'max', false)

was used. I have left this singlestat with consolidateBy for us to observe the difference in its behavior (if any) in respect to the remaining three singlestats that summarize Import Time.

In total despair I have also set Max.Data Points to 9999999 (as @Addshore has already observed that might be useful in some cases, although I don't see how it would help us here). No effect; same numbers.

In my opinion: something is rotten in the dashboard of Grafana; or, to cite @Addshore

"I have no doubt that Grafana will continue to trip me and others up with little quirks like this."

@Lea_WMDE @Addshore

I wonder what is the status of this task following our joint inquiry into the inner depths and miracles of Grafana?