In this talk, I will present some of the results obtained by studying the large scale transcriptional regulatory network controlling E. coli metabolism containing 583 genes and 96 external metabolites as a Boolean dynamical system. We have studied how the attractors of the boolean dynamical system depend on the initial condition of the genes and on various environmental conditions. We find that the nature of the attractors endows the cell with both stability for a fixed environment and flexibility to adapt to changes in the environment. We comment on design features of the system that are responsible for these observed system level properties. In particular, I will reflect upon our results obtained by studying this real genetic network as opposed to those obtained earlier by studying random boolean networks.