On Sun, Jun 7, 2020, at 12:02, Brandon Martin wrote: > This is difficult to understate. "People" are continually amazed when I > show them that I can leave TCP sessions up for days at a time (with > properly configured endpoints) with absolutely zero keepalive traffic > being exchanged.
On the other hand, I'm constantly having to remind developers at $job that while this may be normal, it's not guaranteed, and they need to deal with TCP sessions that go down without requiring operator intervention. Reliable networks do not teach developers about fault tolerance ;) -- Harald
