We've had a few reports from the field of users getting back "Operation timed out" errors (see SERVER-22392 and SERVER-22273 for some examples of this). This message is only returned by NetworkInterfaceASIO when a remote command scheduled through it times out. When that happens though, it's just about impossible to figure out what the actual root cause is. We need a better way to diagnose issues of this type. At the very least it would be nice to know what the name of the command being run was, to whom it was sent, and what the timeout that was exceeded actually was.