Mike Schroepfer Reporter
	Description • 17 years ago

Discussions with Ken and others about how to reduce the variance of Tp results suggested that we up the number of runs we do per page to reduce variance.  Right now we run 5 runs per page - can we allocate 1 Talos windows machine to run Tp only (e.g. no other tests) and up the number of runs from 5 to 100 to see if that gets us more accurate results.

	Mike Schroepfer Reporter
	Comment 1 • 17 years ago

Note if this is easier with Tp2 that's fine.. just want one of the pageload numbers to be more stable.

	alice nodelman [:alice] [:anode]
	Comment 2 • 16 years ago

Long ago completed this test - found that numbers were more consistent at 20 cycles through the page set.  There was no improvement at 50 cycles.  We currently do 10 cycles through the set as a balance between consistent numbers and talos machine test cycle times. 

	(not currently active) Ted Mielczarek
	Comment 3 • 16 years ago

Mass move of Core:Testing bugs to mozilla.org:ReleaseEngineering. Filter on RelEngMassMove to ignore.