Currently backup to nbd target is broken, as nbd doesn't have
.bdrv_get_info realization.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
---

v3: fix compilation (I feel like an idiot)
    adjust wording (Fam)
    
v2: add WARNING

===

Since commit

commit 4c9bca7e39a6e07ad02c1dcde3478363344ec60b
Author: John Snow <jsnow@redhat.com>
Date:   Thu Feb 25 15:58:30 2016 -0500

    block/backup: avoid copying less than full target clusters

backup to nbd target is broken, we have "Couldn't determine the cluster size of
the target image".

Proposed NBD protocol extension - NBD_OPT_INFO should finally solve this problem.
But until it is not realized, we need allow backup to nbd target due to backward
compatibility.

Furthermore, is it entirely ok to disallow backup if bds lacks .bdrv_get_info?
Which behavior should be default: to fail backup or to use default cluster size?


 block/backup.c | 11 ++++++++++-
 1 file changed, 10 insertions(+), 1 deletion(-)

diff --git a/block/backup.c b/block/backup.c
index ea38733..d800a24 100644
--- a/block/backup.c
+++ b/block/backup.c
@@ -638,7 +638,16 @@ BlockJob *backup_job_create(const char *job_id, BlockDriverState *bs,
      * backup cluster size is smaller than the target cluster size. Even for
      * targets with a backing file, try to avoid COW if possible. */
     ret = bdrv_get_info(target, &bdi);
-    if (ret < 0 && !target->backing) {
+    if (ret == -ENOTSUP) {
+        /* Cluster size is not defined */
+        fprintf(stderr,
+                "WARNING: Target block device doesn't provide information "
+                "about block size and it doesn't have backing file. Default "
+                "block size of %u bytes is used. If actual block size of "
+                "target exceeds this default, backup may be unusable",
+                BACKUP_CLUSTER_SIZE_DEFAULT);
+        job->cluster_size = BACKUP_CLUSTER_SIZE_DEFAULT;
+    } else if (ret < 0 && !target->backing) {
         error_setg_errno(errp, -ret,
             "Couldn't determine the cluster size of the target image, "
             "which has no backing file");
-- 
1.8.3.1

Am 15.02.2017 um 15:58 hat Vladimir Sementsov-Ogievskiy geschrieben:
> Currently backup to nbd target is broken, as nbd doesn't have
> .bdrv_get_info realization.
> 
> Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
> ---
> 
> v3: fix compilation (I feel like an idiot)
>     adjust wording (Fam)
>     
> v2: add WARNING
> 
> ===
> 
> Since commit
> 
> commit 4c9bca7e39a6e07ad02c1dcde3478363344ec60b
> Author: John Snow <jsnow@redhat.com>
> Date:   Thu Feb 25 15:58:30 2016 -0500
> 
>     block/backup: avoid copying less than full target clusters
> 
> backup to nbd target is broken, we have "Couldn't determine the cluster size of
> the target image".
> 
> Proposed NBD protocol extension - NBD_OPT_INFO should finally solve this problem.
> But until it is not realized, we need allow backup to nbd target due to backward
> compatibility.
> 
> Furthermore, is it entirely ok to disallow backup if bds lacks .bdrv_get_info?
> Which behavior should be default: to fail backup or to use default cluster size?
> 
> 
>  block/backup.c | 11 ++++++++++-
>  1 file changed, 10 insertions(+), 1 deletion(-)
> 
> diff --git a/block/backup.c b/block/backup.c
> index ea38733..d800a24 100644
> --- a/block/backup.c
> +++ b/block/backup.c
> @@ -638,7 +638,16 @@ BlockJob *backup_job_create(const char *job_id, BlockDriverState *bs,
>       * backup cluster size is smaller than the target cluster size. Even for
>       * targets with a backing file, try to avoid COW if possible. */
>      ret = bdrv_get_info(target, &bdi);
> -    if (ret < 0 && !target->backing) {
> +    if (ret == -ENOTSUP) {
> +        /* Cluster size is not defined */
> +        fprintf(stderr,

error_report() was better because with HMP, the message would appear in
the monitor (where the user is working) instead of stderr (where they
might easily miss the message).

If this means that you need to include another header file, just do
that.

> +                "WARNING: Target block device doesn't provide information "
> +                "about block size and it doesn't have backing file. Default "
> +                "block size of %u bytes is used. If actual block size of "
> +                "target exceeds this default, backup may be unusable",

This error message could use a few more articles. :-)

    "WARNING: The target block device doesn't provide information "
    "about the block size and it doesn't have a backing file. The default "
    "block size of %u bytes is used. If the actual block size of "
    "the target exceeds this default, the backup may be unusable",


> +                BACKUP_CLUSTER_SIZE_DEFAULT);
> +        job->cluster_size = BACKUP_CLUSTER_SIZE_DEFAULT;
> +    } else if (ret < 0 && !target->backing) {
>          error_setg_errno(errp, -ret,
>              "Couldn't determine the cluster size of the target image, "
>              "which has no backing file");

I'm not completely sure whether this is the right fix or whether it
would be better to address this is NBD, e.g. by adding an option where
you specify the block size when opening the NBD driver. (Later, when NBD
can communicate this, it would check if the option matches what the
server says and error out if it doesn't.)

Kevin

On Wed, Feb 15, 2017 at 05:58:13PM +0300, Vladimir Sementsov-Ogievskiy wrote:
> Currently backup to nbd target is broken, as nbd doesn't have
> .bdrv_get_info realization.
> 
> Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
> ---
> 
> v3: fix compilation (I feel like an idiot)
>     adjust wording (Fam)
>     
> v2: add WARNING
> 
> ===
> 
> Since commit
> 
> commit 4c9bca7e39a6e07ad02c1dcde3478363344ec60b
> Author: John Snow <jsnow@redhat.com>
> Date:   Thu Feb 25 15:58:30 2016 -0500
> 
>     block/backup: avoid copying less than full target clusters
> 
> backup to nbd target is broken, we have "Couldn't determine the cluster size of
> the target image".
> 
> Proposed NBD protocol extension - NBD_OPT_INFO should finally solve this problem.
> But until it is not realized, we need allow backup to nbd target due to backward
> compatibility.
> 
> Furthermore, is it entirely ok to disallow backup if bds lacks .bdrv_get_info?
> Which behavior should be default: to fail backup or to use default cluster size?
> 
> 
>  block/backup.c | 11 ++++++++++-
>  1 file changed, 10 insertions(+), 1 deletion(-)
> 
> diff --git a/block/backup.c b/block/backup.c
> index ea38733..d800a24 100644
> --- a/block/backup.c
> +++ b/block/backup.c
> @@ -638,7 +638,16 @@ BlockJob *backup_job_create(const char *job_id, BlockDriverState *bs,
>       * backup cluster size is smaller than the target cluster size. Even for
>       * targets with a backing file, try to avoid COW if possible. */
>      ret = bdrv_get_info(target, &bdi);
> -    if (ret < 0 && !target->backing) {
> +    if (ret == -ENOTSUP) {
> +        /* Cluster size is not defined */
> +        fprintf(stderr,
> +                "WARNING: Target block device doesn't provide information "
> +                "about block size and it doesn't have backing file. Default "
> +                "block size of %u bytes is used. If actual block size of "
> +                "target exceeds this default, backup may be unusable",
> +                BACKUP_CLUSTER_SIZE_DEFAULT);

You should use error_report, like in your v2 patch.  Just make sure to
include qemu/error-report.h.

> +        job->cluster_size = BACKUP_CLUSTER_SIZE_DEFAULT;
> +    } else if (ret < 0 && !target->backing) {
>          error_setg_errno(errp, -ret,
>              "Couldn't determine the cluster size of the target image, "
>              "which has no backing file");
> -- 
> 1.8.3.1
>