I like to run a max temp test run the rack maxed out and see how hot the hottest is, this will determine how critical a failsafe t-stat is. (anything over 100º I personally would say it is critical.) Then I run controlled temp tests to check the variation top to bottom typically a degree or two if the room is properly regulated to the ambient temp (usually 80º) but in unregulated rooms this can be MUCH more. I personally run a rack in a unregulated room it is 1000 times more difficult to get all your temps spot on. If you are trying this (I do NOT suggest this BTW) it can take quite some time to get it right personally it took about 3 weeks and I had the extra thermostats it required (mine took 4 probes to control the rack fully in a cool room of 65-80º)

No two set ups are the same so the time it takes varies greatly. Take your time and get it right before the snakes are housed in the new rack.