Il 18/07/2013 08:39, Fam Zheng ha scritto: > On Wed, 07/17 13:04, Ian Main wrote: >> This patch adds the original source drive as a backing drive to our target >> image so that the target image will appear complete during backup. This >> is especially useful for SYNC_MODE_NONE as it allows export via NBD to >> have a complete point-in-time snapshot available for export. >> >> Signed-off-by: Ian Main <im...@redhat.com> >> --- >> block/backup.c | 13 +++++++++++++ >> 1 file changed, 13 insertions(+) >> >> diff --git a/block/backup.c b/block/backup.c >> index 68abd23..e086e76 100644 >> --- a/block/backup.c >> +++ b/block/backup.c >> @@ -323,6 +323,10 @@ static void coroutine_fn backup_run(void *opaque) >> >> hbitmap_free(job->bitmap); >> >> + /* Set the target backing drive back to NULL before calling delete or >> + * it will also delete the underlying drive. */ >> + target->backing_hd = NULL; >> + >> bdrv_iostatus_disable(target); >> bdrv_delete(target); >> >> @@ -362,6 +366,15 @@ void backup_start(BlockDriverState *bs, >> BlockDriverState *target, >> return; >> } >> >> + /* Manually set the backing hd to be the backup source drive so >> + * that all reads done while we are backing up will be passed >> + * on to the original source drive. This allows reading from the >> + * image while the backup is in progress, or in the case of >> + * SYNC_MODE_NONE allows a complete image to be present for export. >> + * Note that we do this for all modes including SYNC_MODE_TOP as >> + * even then it allows on-the-fly reading. */ >> + target->backing_hd = bs; >> + > > Also set target->backing_file and target->backing_format here? Paolo?
I don't think so, it is temporary while the job runs so that the NBD server can already return the actual data. Paolo