← Back to team overview

yellow team mailing list archive

Re: parallel test results

 

On 03/21/12 20:28, Gary Poster wrote:
1 core:  4:17
  2 cores: 2:23
3 cores:
4 cores: 1:21
5 cores:
6 cores:
7 cores:
8 cores: 0:51

Actually right now it looks kind of like

[time] = 4 hours/[number of cores] + about 20 minutes

...

Out of curiosity, I'm going to run 1 core without the /dev/random hack turned on over night. I'm wondering if that might alone be slowing us down in the current production buildbot/ec2 environment.

Gary


References