# R: looking for “time” clusters in a data set

I am new to R and seeking some advice. I have a set (~20M) of data describing on which step a process did fail or succeeded:

name;star_time;step1;step2;step3
A;2011-07-22 11:02:44;passed;passed;failed
B;2011-03-06 19:33:06;passed;failed;failed
...


I want to search for clusters of processes which have mostly succeed on stepX (for a specific month, ...) and visualize it (scatter plot).

Then if possible I would like to find some patterns in data -> as process B did in 80% of the runs successfully ended in stepY after previous run of process Q that same day.

Any advice where to start where to begin? I don't have an overly fast machine, so any efficiency advices are welcome. Thanks.

-