Often the root cause is in dev code without proper logging and documentation = dev problem. The only way for Ops to troubleshoot might be to sniff the network a

Often the root cause is in dev code without proper logging and documentation = dev problem. The only way for Ops to troubleshoot might be to sniff the network and make educated guesses, if it's recurring. All that extra work costs much more and halts more urgent production work. You broke it, you fix it.