how could you find the problem in a large distributed system
Sigiloso
logs? no. exceptions? no. console? no. back tracing? no. line-by-line execution? no. narrow the possible failed parts? no. I wanted to talk about network failure, but they stopped me.