Run block_copy iterations in parallel in aio tasks.
Changes:
- BlockCopyTask becomes aio task structure. Add zeroes field to pass
it to block_copy_do_copy
- add call state - it's a state of one call of block_copy(), shared
between parallel tasks. For now used only to keep information about
first error: is it read or not.
- convert block_copy_dirty_clusters to aio-task loop.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20200429130847.28124-6-vsementsov@virtuozzo.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>
Instead of just relying on the comment "Called only on full-dirty
region" in block_copy_task_create() let's move initial dirty area
search directly to block_copy_task_create(). Let's also use effective
bdrv_dirty_bitmap_next_dirty_area instead of looping through all
non-dirty clusters.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-Id: <20200429130847.28124-5-vsementsov@virtuozzo.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>
We are going to use aio-task-pool API, so we'll need state pointer in
BlockCopyTask anyway. Add it now and use where possible.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-Id: <20200429130847.28124-4-vsementsov@virtuozzo.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>
We are going to use aio-task-pool API, so tasks will be handled in
parallel. We need therefore separate allocated task on each iteration.
Introduce this logic now.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-Id: <20200429130847.28124-3-vsementsov@virtuozzo.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>
We are going to use aio-task-pool API and extend in-flight request
structure to be a successor of AioTask, so rename things appropriately.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-Id: <20200429130847.28124-2-vsementsov@virtuozzo.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>
Hide structure definitions and add explicit API instead, to keep an
eye on the scope of the shared fields.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Andrey Shinkevich <andrey.shinkevich@virtuozzo.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-Id: <20200311103004.7649-10-vsementsov@virtuozzo.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>
Currently, block_copy operation lock the whole requested region. But
there is no reason to lock clusters, which are already copied, it will
disturb other parallel block_copy requests for no reason.
Let's instead do the following:
Lock only sub-region, which we are going to operate on. Then, after
copying all dirty sub-regions, we should wait for intersecting
requests block-copy, if they failed, we should retry these new dirty
clusters.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Andrey Shinkevich <andrey.shinkevich@virtuozzo.com>
Message-Id: <20200311103004.7649-9-vsementsov@virtuozzo.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>
offset/bytes pair is more usual naming in block layer, let's use it.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Andrey Shinkevich <andrey.shinkevich@virtuozzo.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-Id: <20200311103004.7649-8-vsementsov@virtuozzo.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>
We have a lot of "chunk_end - start" invocations, let's switch to
bytes/cur_bytes scheme instead.
While being here, improve check on block_copy_do_copy parameters to not
overflow when calculating nbytes and use int64_t for bytes in
block_copy for consistency.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Andrey Shinkevich <andrey.shinkevich@virtuozzo.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-Id: <20200311103004.7649-7-vsementsov@virtuozzo.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>
Split find_conflicting_inflight_req to be used separately.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Andrey Shinkevich <andrey.shinkevich@virtuozzo.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-Id: <20200311103004.7649-6-vsementsov@virtuozzo.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>
Use bdrv_block_status_above to chose effective chunk size and to handle
zeroes effectively.
This substitutes checking for just being allocated or not, and drops
old code path for it. Assistance by backup job is dropped too, as
caching block-status information is more difficult than just caching
is-allocated information in our dirty bitmap, and backup job is not
good place for this caching anyway.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Andrey Shinkevich <andrey.shinkevich@virtuozzo.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-Id: <20200311103004.7649-5-vsementsov@virtuozzo.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>
In block_copy_do_copy we fallback to read+write if copy_range failed.
In this case copy_size is larger than defined for buffered IO, and
there is corresponding commit. Still, backup copies data cluster by
cluster, and most of requests are limited to one cluster anyway, so the
only source of this one bad-limited request is copy-before-write
operation.
Further patch will move backup to use block_copy directly, than for
cases where copy_range is not supported, first request will be
oversized in each backup. It's not good, let's change it now.
Fix is simple: just limit first copy_range request like buffer-based
request. If it succeed, set larger copy_range limit.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Andrey Shinkevich <andrey.shinkevich@virtuozzo.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-Id: <20200311103004.7649-4-vsementsov@virtuozzo.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>
Assume we have two regions, A and B, and region B is in-flight now,
region A is not yet touched, but it is unallocated and should be
skipped.
Correspondingly, as progress we have
total = A + B
current = 0
If we reset unallocated region A and call progress_reset_callback,
it will calculate 0 bytes dirty in the bitmap and call
job_progress_set_remaining, which will set
total = current + 0 = 0 + 0 = 0
So, B bytes are actually removed from total accounting. When job
finishes we'll have
total = 0
current = B
, which doesn't sound good.
This is because we didn't considered in-flight bytes, actually when
calculating remaining, we should have set (in_flight + dirty_bytes)
as remaining, not only dirty_bytes.
To fix it, let's refactor progress calculation, moving it to block-copy
itself instead of fixing callback. And, of course, track in_flight
bytes count.
We still have to keep one callback, to maintain backup job bytes_read
calculation, but it will go on soon, when we turn the whole backup
process into one block_copy call.
Cc: qemu-stable@nongnu.org
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Andrey Shinkevich <andrey.shinkevich@virtuozzo.com>
Message-Id: <20200311103004.7649-3-vsementsov@virtuozzo.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>
0e2402452f allowed writes larger than cluster, but that's
unsupported for compressed write. Fix it.
Fixes: 0e2402452f
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-id: 20191029150934.26416-1-vsementsov@virtuozzo.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
No reason to limit buffered copy to one cluster. Let's allow up to 1
MiB.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-id: 20191022111805.3432-7-vsementsov@virtuozzo.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
Currently total allocation for parallel requests to block-copy instance
is unlimited. Let's limit it to 128 MiB.
For now block-copy is used only in backup, so actually we limit total
allocation for backup job.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-id: 20191022111805.3432-6-vsementsov@virtuozzo.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
Merge copying code into one function block_copy_do_copy, which only
calls bdrv_ io functions and don't do any synchronization (like dirty
bitmap set/reset).
Refactor block_copy() function so that it takes full decision about
size of chunk to be copied and does all the synchronization (checking
intersecting requests, set/reset dirty bitmaps).
It will help:
- introduce parallel processing of block_copy iterations: we need to
calculate chunk size, start async chunk copying and go to the next
iteration
- simplify synchronization improvement (like memory limiting in
further commit and reducing critical section (now we lock the whole
requested range, when actually we need to lock only dirty region
which we handle at the moment))
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-id: 20191022111805.3432-4-vsementsov@virtuozzo.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
Large copy range may imply memory allocation and large io effort, so
using 2G copy range request may be bad idea. Let's limit it to 16 MiB.
It also helps the following patch to refactor copy-with-offload
fallback to copy-with-bounce-buffer.
Note, that total memory usage of backup is still not limited, it will
be fixed in further commit.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-id: 20191022111805.3432-3-vsementsov@virtuozzo.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
Move bounce_buffer allocation block_copy_with_bounce_buffer. This
commit simplifies further work on implementing copying by larger chunks
(of different size) and further asynchronous handling of block_copy
iterations (with help of block/aio_task API).
Allocation works fast, a lot faster than disk io, so it's not a problem
that we now allocate/free bounce_buffer more times. And we anyway will
have to allocate several bounce_buffers for parallel execution of loop
iterations in future.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-id: 20191022111805.3432-2-vsementsov@virtuozzo.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
Add bs field to BdrvDirtyBitmap structure. Drop BlockDriverState
parameter from bitmap APIs where possible.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: John Snow <jsnow@redhat.com>
Message-id: 20190916141911.5255-3-vsementsov@virtuozzo.com
[Rebased on top of block-copy. --js]
Signed-off-by: John Snow <jsnow@redhat.com>
Drop write notifiers and use filter node instead.
= Changes =
1. Add filter-node-name argument for backup qmp api. We have to do it
in this commit, as 257 needs to be fixed.
2. There are no more write notifiers here, so is_write_notifier
parameter is dropped from block-copy paths.
3. To sync with in-flight requests at job finish we now have drained
removing of the filter, we don't need rw-lock.
4. Block-copy is now using BdrvChildren instead of BlockBackends
5. As backup-top owns these children, we also move block-copy state
into backup-top's ownership.
= Iotest changes =
56: op-blocker doesn't shoot now, as we set it on source, but then
check on filter, when trying to start second backup.
To keep the test we instead can catch another collision: both jobs will
get 'drive0' job-id, as job-id parameter is unspecified. To prevent
interleaving with file-posix locks (as they are dependent on config)
let's use another target for second backup.
Also, it's obvious now that we'd like to drop this op-blocker at all
and add a test-case for two backups from one node (to different
destinations) actually works. But not in these series.
141: Output changed: prepatch, "Node is in use" comes from bdrv_has_blk
check inside qmp_blockdev_del. But we've dropped block-copy blk
objects, so no more blk objects on source bs (job blk is on backup-top
filter bs). New message is from op-blocker, which is the next check in
qmp_blockdev_add.
257: The test wants to emulate guest write during backup. They should
go to filter node, not to original source node, of course. Therefore we
need to specify filter node name and use it.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-id: 20191001131409.14202-6-vsementsov@virtuozzo.com
Reviewed-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>
Split block_copy_set_callbacks out of block_copy_state_new. It's needed
for further commit: block-copy will use BdrvChildren of backup-top
filter, so it will be created from backup-top filter creation function.
But callbacks will still belong to backup job and will be set in
separate.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-id: 20191001131409.14202-4-vsementsov@virtuozzo.com
Reviewed-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>
Move synchronization mechanism to block-copy, to be able to use one
block-copy instance from backup job and backup-top filter in parallel.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-id: 20191001131409.14202-2-vsementsov@virtuozzo.com
Reviewed-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>
Split block_copy to separate file, to be cleanly shared with backup-top
filter driver in further commits.
It's a clean movement, the only change is drop "static" from interface
functions.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-id: 20190920142056.12778-8-vsementsov@virtuozzo.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
Vladimir Sementsov-Ogievskiy