Eric Blake f1426881a8 nbd/server: Request TCP_NODELAY ... Nagle's algorithm adds latency in order to reduce network packet overhead on small packets. But when we are already using corking to merge smaller packets into transactional requests, the extra delay from TCP defaults just gets in the way (see recent commit bd2cd4a4). For reference, qemu as an NBD client already requests TCP_NODELAY (see nbd_connect() in nbd/client-connection.c); as does libnbd as a client [1], and nbdkit as a server [2]. Furthermore, the NBD spec recommends the use of TCP_NODELAY [3]. [1] https://gitlab.com/nbdkit/libnbd/-/blob/a48a1142/generator/states-connect.c#L39 [2] https://gitlab.com/nbdkit/nbdkit/-/blob/45b72f5b/server/sockets.c#L430 [3] https://github.com/NetworkBlockDevice/nbd/blob/master/doc/proto.md#protocol-phases CC: Florian Westphal <fw@strlen.de> Signed-off-by: Eric Blake <eblake@redhat.com> Message-Id: <20230404004047.142086-1-eblake@redhat.com> Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>		2023-04-04 08:13:15 -05:00
..
client-connection.c	coroutine: Split qemu/coroutine-core.h off qemu/coroutine.h	2023-01-20 07:21:46 +01:00
client.c	block/nbd: drop connection_co	2021-09-29 13:46:33 -05:00
common.c
meson.build	nbd: move connection code from block/nbd to nbd/client-connection	2021-06-18 10:59:53 -05:00
nbd-internal.h	coroutine: Clean up superfluous inclusion of qemu/coroutine.h	2023-01-19 10:18:28 +01:00
server.c	nbd/server: Request TCP_NODELAY	2023-04-04 08:13:15 -05:00
trace-events	nbd: trace long NBD operations	2022-06-29 10:57:02 +03:00
trace.h