I am trying to build the software as fast as possible. To accomplish this, I did the following:
* Run the agent on a machine with an SSD, the actual workspace is even mapped on a ramdisk of 50 GB
* Put lots of processors in the machine (16)
During compilation, etc, I don't see anything peek. Is there a way to make sure the processors (because I think that is the bottleneck when compiling on a ram disk?) are using their full potential?
I’m not sure there is much we can do for that, most of the time is spent running other processes. Whether those processes make full use of the available resources is up to them. The MSBuild action has a max cpu option, setting it to zero will allow it to use 1 process per core, but whether it actually does that depends on the structure of your solution and the dependencies between the projects.