ta to a trickling speed compared to what it should be.
--
Carlo Wood
ce we know that the stall always seems to be 207ms, I'm pretty sure
that actually it stalled 14 ms before that, and needed 14 ms to finish the
complete burst in one direction.
--
Carlo Wood
are WAY smaller.
Obviously I have no idea if this is a problem with epoll itself,
or that there is a problem with TCP/IP for this case.
But I am not a kernel hacker, so all I could to was provide this
little test case that allows you to reproduce it.
--
Carlo Wood
