I'll use it with block backends shortly, and the name is going to fit
badly there. It's a block layer thing anyway, not just a block driver
thing.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
I'll use BlockDriverAIOCB with block backends shortly, and the name is
going to fit badly there. It's a block layer thing anyway, not just a
block driver thing.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
device_name[] can become non-empty only in bdrv_new_root() and
bdrv_move_feature_fields(). The latter is used only to undo damage
done by bdrv_swap(). The former is called only by blk_new_with_bs().
Therefore, when a BlockDriverState's device_name[] is non-empty, then
it's been created with a BlockBackend, and vice versa. Furthermore,
blk_new_with_bs() keeps the two names equal.
Therefore, device_name[] is redundant. Eliminate it.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Before, we cancel all the child requests with bdrv_aio_cancel, then free
the acb..
Now we just kick off asynchronous cancellation of child requests and
return, we know quorum_aio_cb will be called later, so in the end
quorum_aio_finalize will take care of calling the caller's cb.
Signed-off-by: Fam Zheng <famz@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
For a fifo read pattern, we only have one running aio (possible other cases that
has less number than num_children in the future), so we need to check if
.acb is NULL against bdrv_aio_cancel() to avoid segfault.
Cc: Eric Blake <eblake@redhat.com>
Cc: Benoit Canet <benoit@irqsave.net>
Cc: Kevin Wolf <kwolf@redhat.com>
Cc: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Liu Yuan <namei.unix@gmail.com>
Signed-off-by: Fam Zheng <famz@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
This patch adds single read pattern to quorum driver and quorum vote is default
pattern.
For now we do a quorum vote on all the reads, it is designed for unreliable
underlying storage such as non-redundant NFS to make sure data integrity at the
cost of the read performance.
For some use cases as following:
VM
--------------
| |
v v
A B
Both A and B has hardware raid storage to justify the data integrity on its own.
So it would help performance if we do a single read instead of on all the nodes.
Further, if we run VM on either of the storage node, we can make a local read
request for better performance.
This patch generalize the above 2 nodes case in the N nodes. That is,
vm -> write to all the N nodes, read just one of them. If single read fails, we
try to read next node in FIFO order specified by the startup command.
The 2 nodes case is very similar to DRBD[1] though lack of auto-sync
functionality in the single device/node failure for now. But compared with DRBD
we still have some advantages over it:
- Suppose we have 20 VMs running on one(assume A) of 2 nodes' DRBD backed
storage. And if A crashes, we need to restart all the VMs on node B. But for
practice case, we can't because B might not have enough resources to setup 20 VMs
at once. So if we run our 20 VMs with quorum driver, and scatter the replicated
images over the data center, we can very likely restart 20 VMs without any
resource problem.
After all, I think we can build a more powerful replicated image functionality
on quorum and block jobs(block mirror) to meet various High Availibility needs.
E.g, Enable single read pattern on 2 children,
-drive driver=quorum,children.0.file.filename=0.qcow2,\
children.1.file.filename=1.qcow2,read-pattern=fifo,vote-threshold=1
[1] http://en.wikipedia.org/wiki/Distributed_Replicated_Block_Device
[Dropped \n from an error_setg() error message
--Stefan]
Cc: Benoit Canet <benoit@irqsave.net>
Cc: Eric Blake <eblake@redhat.com>
Cc: Kevin Wolf <kwolf@redhat.com>
Cc: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Liu Yuan <namei.unix@gmail.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
On read operations when this parameter is set and some replicas are corrupted
while quorum can be reached quorum will proceed to rewrite the correct version
of the data to fix the corrupted replicas.
This will shine with SSD where the FTL will remap the same block at another
place on rewrite.
Signed-off-by: Benoit Canet <benoit@irqsave.net>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Implement .bdrv_detach/attach_aio_context() interfaces to propagate
detach/attach to BDRVQuorumState->bs[] children. The block layer takes
care of ->file and ->backing_hd but doesn't know about our ->bs[]
BlockDriverStates, which is also part of the graph.
Reviewed-by: Benoît Canet <benoit.canet@irqsave.net>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
error_is_set(&var) is the same as var != NULL, but it takes
whole-program analysis to figure that out. Unnecessarily hard for
optimizers, static checkers, and human readers. Commit 84d18f0 dumbed
it down to obvious, but a few more have crept in since, and
documentation was overlooked. Dumb these down, too.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
If it returns an error, the migrated VM will not be started, but qemu
exits with an error message.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Benoit Canet <benoit@irqsave.net>
This patch keep the recursive way of doing things but simplify it by giving
two responsabilities to all block filters implementors.
They will need to do two things:
-Set the is_filter field of their block driver to true.
-Implement the bdrv_recurse_is_first_non_filter method of their block driver like
it is done on the Quorum block driver. (block/quorum.c)
[Paolo Bonzini <pbonzini@redhat.com> pointed out that this patch changes
the semantics of blkverify, which now recurses down both bs->file and
s->test_file.
-- Stefan]
Reported-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Benoit Canet <benoit@irqsave.net>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Insert quorum QMP events documentation alphabetically.
Also change the "ret" errno value by an optional "error" being an strerror(-ret)
in the QUORUM_REPORT_BAD qmp event.
Signed-off-by: Benoit Canet <benoit@irqsave.net>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Although it may not look like it, this patch simplifies quorum_open().
qdict_array_split() is now able to return QLists with different objects
than only QDicts, therefore it will now do all the work and
quorum_open() does not have to handle reference strings by itself.
This allows mixing full option dicts and reference strings for
specifying the child block devices of quorum; furthermore, it improves
handling of malformed specifications.
Signed-off-by: Max Reitz <mreitz@redhat.com>
Reviewed-by: Benoit Canet <benoit@irqsave.net>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Example of command line:
-drive if=virtio,driver=quorum,\
children.0.file.filename=1.raw,\
children.0.node-name=1.raw,\
children.0.driver=raw,\
children.1.file.filename=2.raw,\
children.1.node-name=2.raw,\
children.1.driver=raw,\
children.2.file.filename=3.raw,\
children.2.node-name=3.raw,\
children.2.driver=raw,\
vote-threshold=2
blkverify=on with vote-threshold=2 and two files can be passed to
emulate blkverify.
Signed-off-by: Benoit Canet <benoit@irqsave.net>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
This is used to activate quorum snapshot.
Signed-off-by: Benoit Canet <benoit@irqsave.net>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Makes a vote to select error if any.
Signed-off-by: Benoit Canet <benoit@irqsave.net>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
We really want that live migration works with quorum so implement
quorum_invalidate_cache().
Signed-off-by: Benoit Canet <benoit@irqsave.net>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Check that every bs file returns the same length.
Otherwise, return -EIO to disable the quorum and
avoid length discrepancy.
Signed-off-by: Benoit Canet <benoit@irqsave.net>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
This patchset enables the core of the quorum mechanism.
The num_children reads are compared to get the majority version and if this
version exists more than threshold times the guest won't see the error at all.
If a block is corrupted or if an error occurs during an IO or if the quorum
cannot be established QMP events are used to report to the management.
Use gnutls's SHA-256 to compare versions.
--enable-quorum must be used to enable the feature.
Signed-off-by: Benoit Canet <benoit@irqsave.net>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Add code to do num_children reads in parallel and cleanup the structures
afterwards.
Signed-off-by: Benoit Canet <benoit@irqsave.net>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Writes are mirrored num_children times on num_children devices.
Signed-off-by: Benoit Canet <benoit@irqsave.net>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Create the structure holding the quorum settings and write the minimal block
driver instanciation boilerplate.
Signed-off-by: Benoit Canet <benoit@irqsave.net>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Quorum is a block filter mirroring writes to num_children children.
For reads quorum reads each children and does a vote.
If more than vote_threshold versions are identical the quorum is reached and
this winning version is returned to the guest. So quorum prevents bit corruption.
For high availability purpose minority errors are reported via QMP but the guest
does not see them.
This patch creates the driver C source file and introduces the structures that
will be used in asynchronous reads and writes.
Signed-off-by: Benoit Canet <benoit@irqsave.net>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
