Oxygen and water supply was the least of the worry with the failure of the Russian TsBM and TBM computers. Without them, Russian thrusters couldn't be used meaning the Shuttle couldn't undock without leaving the ISS to tumble. American gyroscopes can be used for long duration at a static attitude but thrusters are required to do any type of maneuver. Fortunately, 4 out of the 6 computers were restarted nominally after the other two were bypassed (there are 3 redundant TsBMs and 3 redundant TBMs - at least one of each is required). The failed computers are known to be finicky about their power supply. Whether the upstream power or the computer's power supply are to blame remains to be seen.
I used to prefer the design of the Russian segment Command & Data Handling system (CDH) to the American design but now I'm having second thoughts. Rather than the Russian system which uses two sets of two-fault tolerant computers, the US uses a network of dozens of computers, all one or two-fault tolerant, which have identical hardware but dedicated software. In the event of a hardware failure, a less important computer's components can be scavenged to build or repair a more important one. One advantage to the Russian system is that each type of computer runs its calculations in parallel with the other two; they then compare data, reject anything that doesn't fit. If one computer is voted out by the other two it goes into a diagnostic state. There are pros and cons to both philosophies I suppose.
Tidbit of info: the Russian computers were designed in Germany. They are NOT the same as the ones that the Russians used in the past.
The shuttle took a sweet pics of a TIE fighter yesterday: