FETT Stats have changed - explanation needed

[rfoot]

There have some unexplained changes in the stats that we're reporting out to DARPA

Launches are down by 56 Terms up by 23 and the MIT stats seem flipped.

Need to get an update to DARPA on why they changed.

	SUM	DELTA
Reporting Metrics
Portal Data*
Launches	820	-56
Terminations	818	23
Errors Launching	0	0

• Data includes activity since 07/25/2020, 10:00AM Pacfic and does not include test profile activity   |   |   |   Launches by Type |   |   |   LMCO | CHISEL_P1 | FREERTOS | 117 | 0 |   MICHIGAN | CHISEL_P1 | FREERTOS | 199 | 0 |   GFE | CHISEL_P1 | FREERTOS | 40 | 0 |   LMCO | CHISEL_P2 | DEBIAN | 95 | 1 |   GFE | CHISEL_P2 | DEBIAN | 34 | 1 |   SRI-CAMBRIDGE | BLUESPEC_P2 | FREEBSD (purecap) | 49 | 25 |   SRI-CAMBRIDGE | BLUESPEC_P2 | FREEBSD (default) | 24 | -16 |   SRI-CAMBRIDGE | BLUESPEC_P2 | FREEBSD (temporal) | 3 | 3 |   GFE | BLUESPEC_P2 | FREEBSD | 26 | 7 |   MIT | BLUESPEC_P2 | DEBIAN | 231 | 1 |     |   |   |  

[mattlebeau-galois]

@alex-fivetalent - Is the drop in global # of launches due to the inclusion of the missed "isRedTeam" flag on the total launches lambda?

Also, looking at the stats that @rfoot produced on the evening of 09.13.20 versus those that I produced on the afternoon of 09.10.20, it appears that SRI-CAMBRIDGE | BLUESPEC_P2 | FREEBSD (default) went from 40 -> 24 launches, which I can't quite explain. Can you, sir?

@rfoot - Can you elaborate on what you mean re: "the MIT stats seem flipped", please? I don't see it when I compare the 09.10 and the 09.13 stats. Perhaps you were instead referring to the differences in the SRI-Cambridge (default) stats I asked Alex about above?

[alex-fivetalent]

For the first part. You are correct @mattlebeau-galois, the # global launches is due to including isRedTeam flag. Previously, the launch was excluding just the ftsresearcher test profile from the launch counts. Going forward we will be showing the non-RedTeam values next to the RedTeam ones on the portal metrics page.

For the second part related to the two SRI numbers. Looking at both the updated query and the previous query I get the same results for the RedTeam counts on SRI/Cambridge variants in question which is 50 default 24 purecap. The numbers showed similar to that on 9/9 as well in the screenshot I got from Andrew. I think the discrepancy could be the sortkey making the two rows reverse in order on the page when adding the new variant to the db. The numbers and row label would have been correct, but if just copy/pasting from the site, without the rows it could have appeared to have flipped. That should be a quick fix to make sure the sort order doesn't change on the metrics page.

[mattlebeau-galois]

Closing per conversation w/ @rfoot && @Abivin12