[Qemu-devel] [PATCH 5/9] mirror: improve performance of mirroring of empty disk

["Denis V. Lunev" <den@openvz.org>]

We should not take into account zero blocks for delay calculations.
They are not read and thus IO throttling is not required. In the
other case VM migration with 16 Tb QCOW2 disk with 4 Gb of data takes
days.

Signed-off-by: Denis V. Lunev <den@openvz.org>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
CC: Stefan Hajnoczi <stefanha@redhat.com>
CC: Fam Zheng <famz@redhat.com>
CC: Kevin Wolf <kwolf@redhat.com>
CC: Max Reitz <mreitz@redhat.com>
CC: Jeff Cody <jcody@redhat.com>
CC: Eric Blake <eblake@redhat.com>
---
 block/mirror.c | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

diff --git a/block/mirror.c b/block/mirror.c
index c2f8773..d8be80a 100644
--- a/block/mirror.c
+++ b/block/mirror.c
@@ -363,7 +363,7 @@ static uint64_t coroutine_fn mirror_iteration(MirrorBlockJob *s)
     bitmap_set(s->in_flight_bitmap, sector_num / sectors_per_chunk, nb_chunks);
     while (nb_chunks > 0 && sector_num < end) {
         int ret;
-        int io_sectors;
+        int io_sectors, io_sectors_acct;
         BlockDriverState *file;
         enum MirrorMethod {
             MIRROR_METHOD_COPY,
@@ -399,12 +399,15 @@ static uint64_t coroutine_fn mirror_iteration(MirrorBlockJob *s)
         switch (mirror_method) {
         case MIRROR_METHOD_COPY:
             io_sectors = mirror_do_read(s, sector_num, io_sectors);
+            io_sectors_acct = io_sectors;
             break;
         case MIRROR_METHOD_ZERO:
             mirror_do_zero_or_discard(s, sector_num, io_sectors, false);
+            io_sectors_acct = 0;
             break;
         case MIRROR_METHOD_DISCARD:
             mirror_do_zero_or_discard(s, sector_num, io_sectors, true);
+            io_sectors_acct = 0;
             break;
         default:
             abort();
@@ -412,7 +415,7 @@ static uint64_t coroutine_fn mirror_iteration(MirrorBlockJob *s)
         assert(io_sectors);
         sector_num += io_sectors;
         nb_chunks -= DIV_ROUND_UP(io_sectors, sectors_per_chunk);
-        delay_ns += ratelimit_calculate_delay(&s->limit, io_sectors);
+        delay_ns += ratelimit_calculate_delay(&s->limit, io_sectors_acct);
     }
     return delay_ns;
 }
-- 
2.5.0