[1/1] mirror: do not increase offset during initial zero_or_discard phase
diff mbox

Message ID 1486134413-19988-1-git-send-email-den@openvz.org
State New
Headers show

Commit Message

Denis V. Lunev Feb. 3, 2017, 3:06 p.m. UTC
From: Anton Nefedov <anton.nefedov@virtuozzo.com>

If explicit zeroing out before mirroring is required for the target image,
it moves the block job offset counter to EOF, then offset and len counters
count the image size twice. There is no harm but stats are confusing,
specifically the progress of the operation is always reported as 99% by
management tools.

The patch skips offset increase for the first "technical" pass over the
image. This should not cause any further harm.

Signed-off-by: Anton Nefedov <anton.nefedov@virtuozzo.com>
Signed-off-by: Denis V. Lunev <den@openvz.org>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
CC: Jeff Cody <jcody@redhat.com>
CC: Kevin Wolf <kwolf@redhat.com>
CC: Max Reitz <mreitz@redhat.com>
---
 block/mirror.c | 8 +++++++-
 1 file changed, 7 insertions(+), 1 deletion(-)

Comments

Denis V. Lunev Feb. 3, 2017, 3:08 p.m. UTC | #1
On 02/03/2017 06:06 PM, Denis V. Lunev wrote:
> From: Anton Nefedov <anton.nefedov@virtuozzo.com>
>
> If explicit zeroing out before mirroring is required for the target image,
> it moves the block job offset counter to EOF, then offset and len counters
> count the image size twice. There is no harm but stats are confusing,
> specifically the progress of the operation is always reported as 99% by
> management tools.
>
> The patch skips offset increase for the first "technical" pass over the
> image. This should not cause any further harm.
>
> Signed-off-by: Anton Nefedov <anton.nefedov@virtuozzo.com>
> Signed-off-by: Denis V. Lunev <den@openvz.org>
> Reviewed-by: Eric Blake <eblake@redhat.com>
> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
> CC: Jeff Cody <jcody@redhat.com>
> CC: Kevin Wolf <kwolf@redhat.com>
> CC: Max Reitz <mreitz@redhat.com>
actually this is V3 patch. Sorry for broken subject.

Den
Denis Lunev Feb. 13, 2017, 7:10 a.m. UTC | #2
On 02/03/2017 06:08 PM, Denis V. Lunev wrote:
> On 02/03/2017 06:06 PM, Denis V. Lunev wrote:
>> From: Anton Nefedov <anton.nefedov@virtuozzo.com>
>>
>> If explicit zeroing out before mirroring is required for the target image,
>> it moves the block job offset counter to EOF, then offset and len counters
>> count the image size twice. There is no harm but stats are confusing,
>> specifically the progress of the operation is always reported as 99% by
>> management tools.
>>
>> The patch skips offset increase for the first "technical" pass over the
>> image. This should not cause any further harm.
>>
>> Signed-off-by: Anton Nefedov <anton.nefedov@virtuozzo.com>
>> Signed-off-by: Denis V. Lunev <den@openvz.org>
>> Reviewed-by: Eric Blake <eblake@redhat.com>
>> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
>> CC: Jeff Cody <jcody@redhat.com>
>> CC: Kevin Wolf <kwolf@redhat.com>
>> CC: Max Reitz <mreitz@redhat.com>
> actually this is V3 patch. Sorry for broken subject.
>
> Den
ping
Max Reitz Feb. 13, 2017, 5:16 p.m. UTC | #3
On 13.02.2017 08:10, Denis V. Lunev wrote:
> On 02/03/2017 06:08 PM, Denis V. Lunev wrote:
>> On 02/03/2017 06:06 PM, Denis V. Lunev wrote:
>>> From: Anton Nefedov <anton.nefedov@virtuozzo.com>
>>>
>>> If explicit zeroing out before mirroring is required for the target image,
>>> it moves the block job offset counter to EOF, then offset and len counters
>>> count the image size twice. There is no harm but stats are confusing,
>>> specifically the progress of the operation is always reported as 99% by
>>> management tools.
>>>
>>> The patch skips offset increase for the first "technical" pass over the
>>> image. This should not cause any further harm.
>>>
>>> Signed-off-by: Anton Nefedov <anton.nefedov@virtuozzo.com>
>>> Signed-off-by: Denis V. Lunev <den@openvz.org>
>>> Reviewed-by: Eric Blake <eblake@redhat.com>
>>> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
>>> CC: Jeff Cody <jcody@redhat.com>
>>> CC: Kevin Wolf <kwolf@redhat.com>
>>> CC: Max Reitz <mreitz@redhat.com>
>> actually this is V3 patch. Sorry for broken subject.
>>
>> Den
> ping

Didn't Jeff merge v2?

http://lists.nongnu.org/archive/html/qemu-devel/2017-02/msg01319.html

Max
Jeff Cody Feb. 13, 2017, 7:13 p.m. UTC | #4
On Mon, Feb 13, 2017 at 06:16:36PM +0100, Max Reitz wrote:
> On 13.02.2017 08:10, Denis V. Lunev wrote:
> > On 02/03/2017 06:08 PM, Denis V. Lunev wrote:
> >> On 02/03/2017 06:06 PM, Denis V. Lunev wrote:
> >>> From: Anton Nefedov <anton.nefedov@virtuozzo.com>
> >>>
> >>> If explicit zeroing out before mirroring is required for the target image,
> >>> it moves the block job offset counter to EOF, then offset and len counters
> >>> count the image size twice. There is no harm but stats are confusing,
> >>> specifically the progress of the operation is always reported as 99% by
> >>> management tools.
> >>>
> >>> The patch skips offset increase for the first "technical" pass over the
> >>> image. This should not cause any further harm.
> >>>
> >>> Signed-off-by: Anton Nefedov <anton.nefedov@virtuozzo.com>
> >>> Signed-off-by: Denis V. Lunev <den@openvz.org>
> >>> Reviewed-by: Eric Blake <eblake@redhat.com>
> >>> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
> >>> CC: Jeff Cody <jcody@redhat.com>
> >>> CC: Kevin Wolf <kwolf@redhat.com>
> >>> CC: Max Reitz <mreitz@redhat.com>
> >> actually this is V3 patch. Sorry for broken subject.
> >>
> >> Den
> > ping
> 
> Didn't Jeff merge v2?
> 
> http://lists.nongnu.org/archive/html/qemu-devel/2017-02/msg01319.html
> 
> Max
> 


Yes, I did.
Denis Lunev Feb. 13, 2017, 7:45 p.m. UTC | #5
On 02/13/2017 10:13 PM, Jeff Cody wrote:
> On Mon, Feb 13, 2017 at 06:16:36PM +0100, Max Reitz wrote:
>> On 13.02.2017 08:10, Denis V. Lunev wrote:
>>> On 02/03/2017 06:08 PM, Denis V. Lunev wrote:
>>>> On 02/03/2017 06:06 PM, Denis V. Lunev wrote:
>>>>> From: Anton Nefedov <anton.nefedov@virtuozzo.com>
>>>>>
>>>>> If explicit zeroing out before mirroring is required for the target image,
>>>>> it moves the block job offset counter to EOF, then offset and len counters
>>>>> count the image size twice. There is no harm but stats are confusing,
>>>>> specifically the progress of the operation is always reported as 99% by
>>>>> management tools.
>>>>>
>>>>> The patch skips offset increase for the first "technical" pass over the
>>>>> image. This should not cause any further harm.
>>>>>
>>>>> Signed-off-by: Anton Nefedov <anton.nefedov@virtuozzo.com>
>>>>> Signed-off-by: Denis V. Lunev <den@openvz.org>
>>>>> Reviewed-by: Eric Blake <eblake@redhat.com>
>>>>> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
>>>>> CC: Jeff Cody <jcody@redhat.com>
>>>>> CC: Kevin Wolf <kwolf@redhat.com>
>>>>> CC: Max Reitz <mreitz@redhat.com>
>>>> actually this is V3 patch. Sorry for broken subject.
>>>>
>>>> Den
>>> ping
>> Didn't Jeff merge v2?
>>
>> http://lists.nongnu.org/archive/html/qemu-devel/2017-02/msg01319.html
>>
>> Max
>>
>
> Yes, I did.
thank you very much for clarification

Den

Patch
diff mbox

diff --git a/block/mirror.c b/block/mirror.c
index 301ba92..688bc91 100644
--- a/block/mirror.c
+++ b/block/mirror.c
@@ -69,6 +69,7 @@  typedef struct MirrorBlockJob {
     bool waiting_for_io;
     int target_cluster_sectors;
     int max_iov;
+    bool init_pass;
 } MirrorBlockJob;
 
 typedef struct MirrorOp {
@@ -117,7 +118,9 @@  static void mirror_iteration_done(MirrorOp *op, int ret)
         if (s->cow_bitmap) {
             bitmap_set(s->cow_bitmap, chunk_num, nb_chunks);
         }
-        s->common.offset += (uint64_t)op->nb_sectors * BDRV_SECTOR_SIZE;
+        if (!s->init_pass) {
+            s->common.offset += (uint64_t)op->nb_sectors * BDRV_SECTOR_SIZE;
+        }
     }
 
     qemu_iovec_destroy(&op->qiov);
@@ -566,6 +569,7 @@  static int coroutine_fn mirror_dirty_init(MirrorBlockJob *s)
             return 0;
         }
 
+        s->init_pass = true;
         for (sector_num = 0; sector_num < end; ) {
             int nb_sectors = MIN(end - sector_num,
                 QEMU_ALIGN_DOWN(INT_MAX, s->granularity) >> BDRV_SECTOR_BITS);
@@ -573,6 +577,7 @@  static int coroutine_fn mirror_dirty_init(MirrorBlockJob *s)
             mirror_throttle(s);
 
             if (block_job_is_cancelled(&s->common)) {
+                s->init_pass = false;
                 return 0;
             }
 
@@ -587,6 +592,7 @@  static int coroutine_fn mirror_dirty_init(MirrorBlockJob *s)
         }
 
         mirror_wait_for_all_io(s);
+        s->init_pass = false;
     }
 
     /* First part, loop on the sectors and initialize the dirty bitmap.  */