0b6715de3c
- kvm-migration-Handle-block-device-inactivation-failures-.patch [bz#2177957] - kvm-migration-Minor-control-flow-simplification.patch [bz#2177957] - kvm-migration-Attempt-disk-reactivation-in-more-failure-.patch [bz#2177957] - kvm-nbd-server-push-pending-frames-after-sending-reply.patch [bz#2035712] - kvm-nbd-server-Request-TCP_NODELAY.patch [bz#2035712] - Resolves: bz#2177957 (Qemu core dump if cut off nfs storage during migration) - Resolves: bz#2035712 ([qemu] Booting from Guest Image over NBD with TLS Is Slow)
73 lines
2.4 KiB
Diff
73 lines
2.4 KiB
Diff
From 170872370c6f3c916e741eb32d80431995d7a870 Mon Sep 17 00:00:00 2001
|
|
From: Florian Westphal <fw@strlen.de>
|
|
Date: Fri, 24 Mar 2023 11:47:20 +0100
|
|
Subject: [PATCH 4/5] nbd/server: push pending frames after sending reply
|
|
|
|
RH-Author: Eric Blake <eblake@redhat.com>
|
|
RH-MergeRequest: 274: nbd: improve TLS performance of NBD server
|
|
RH-Bugzilla: 2035712
|
|
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
|
|
RH-Acked-by: Kevin Wolf <kwolf@redhat.com>
|
|
RH-Acked-by: Stefano Garzarella <sgarzare@redhat.com>
|
|
RH-Commit: [1/2] ab92c06c48810aa40380de0433dcac4c6e4be9a5 (ebblake/qemu-kvm)
|
|
|
|
qemu-nbd doesn't set TCP_NODELAY on the tcp socket.
|
|
|
|
Kernel waits for more data and avoids transmission of small packets.
|
|
Without TLS this is barely noticeable, but with TLS this really shows.
|
|
|
|
Booting a VM via qemu-nbd on localhost (with tls) takes more than
|
|
2 minutes on my system. tcpdump shows frequent wait periods, where no
|
|
packets get sent for a 40ms period.
|
|
|
|
Add explicit (un)corking when processing (and responding to) requests.
|
|
"TCP_CORK, &zero" after earlier "CORK, &one" will flush pending data.
|
|
|
|
VM Boot time:
|
|
main: no tls: 23s, with tls: 2m45s
|
|
patched: no tls: 14s, with tls: 15s
|
|
|
|
VM Boot time, qemu-nbd via network (same lan):
|
|
main: no tls: 18s, with tls: 1m50s
|
|
patched: no tls: 17s, with tls: 18s
|
|
|
|
Future optimization: if we could detect if there is another pending
|
|
request we could defer the uncork operation because more data would be
|
|
appended.
|
|
|
|
Signed-off-by: Florian Westphal <fw@strlen.de>
|
|
Message-Id: <20230324104720.2498-1-fw@strlen.de>
|
|
Reviewed-by: Eric Blake <eblake@redhat.com>
|
|
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
|
|
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
|
|
(cherry picked from commit bd2cd4a441ded163b62371790876f28a9b834317)
|
|
Signed-off-by: Eric Blake <eblake@redhat.com>
|
|
---
|
|
nbd/server.c | 3 +++
|
|
1 file changed, 3 insertions(+)
|
|
|
|
diff --git a/nbd/server.c b/nbd/server.c
|
|
index 4630dd7322..a5edc7f681 100644
|
|
--- a/nbd/server.c
|
|
+++ b/nbd/server.c
|
|
@@ -2647,6 +2647,8 @@ static coroutine_fn void nbd_trip(void *opaque)
|
|
goto disconnect;
|
|
}
|
|
|
|
+ qio_channel_set_cork(client->ioc, true);
|
|
+
|
|
if (ret < 0) {
|
|
/* It wans't -EIO, so, according to nbd_co_receive_request()
|
|
* semantics, we should return the error to the client. */
|
|
@@ -2672,6 +2674,7 @@ static coroutine_fn void nbd_trip(void *opaque)
|
|
goto disconnect;
|
|
}
|
|
|
|
+ qio_channel_set_cork(client->ioc, false);
|
|
done:
|
|
nbd_request_put(req);
|
|
nbd_client_put(client);
|
|
--
|
|
2.39.1
|
|
|