it would be awesome to test how nerve reacts to slow responses and dropped packets; we can probably add this to the helloworld service by having it conditionally reply slightly after the timeout interval. another approach, which would generalize to testing all kinds of services, would be to add network latency at the kernel level.
i was looking into netem for this, using tc. there's more information on stack overflow
it would be awesome to test how nerve reacts to slow responses and dropped packets; we can probably add this to the helloworld service by having it conditionally reply slightly after the timeout interval. another approach, which would generalize to testing all kinds of services, would be to add network latency at the kernel level.
i was looking into netem for this, using tc. there's more information on stack overflow