Instead of artificially enlarging the command decoding arrays to
P_MAX_CMD entries, check if an index is within the valid range using the
ARRAY_SIZE() macro.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
The previous algorithm for dealing with overlapping concurrent writes
was generating unnecessary warnings for scenarios which could be
legitimate, and did not always handle partially overlapping requests
correctly. Improve it algorithm as follows:
* While local or remote write requests are in progress, conflicting new
local write requests will be delayed (commit 82172f7).
* When a conflict between a local and remote write request is detected,
the node with the discard flag decides how to resolve the conflict: It
will ask its peer to discard conflicting requests which are fully
contained in the local request and retry requests which overlap only
partially. This involves a protocol change.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
When the node with the discard flag resolves write conflicts in
dual-primary mode, it may determine that its peer has sent ack packets
on the metadata socket which did not arrive, yet. Wait for the next ack
with ping-timeout instead of a hard-coded 30 seconds.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Commit 9b1e63e changed the concurrent write detection algorithm to only insert
peer requests into write_requests tree after determining that there is no
conflict. With this change, new conflicting local requests could be added
while the algorithm runs, but this case was not handled correctly. Instead of
making the algorithm deal with this case, switch back to adding peer requests
to the write_requests tree immediately: this improves fairness.
When a peer request is discarded, remove that request from the write_requests
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
In compatibility mode with old DRBDs, use that as the state_mutex
as well.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
The lock they constructed is only taken when the state_mutex
was already taken. It is superficial.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
No longer work callbacks must operate on a mdev. From now on they
can also operate on a tconn.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
The drbd_md_sync(mdev) happens in the after state change anyways...
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Make it more clear that update_peer_seq() is supposed to wake up the
seq_wait queue whenever the sequence number changes.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Instead of keeping a separate tree for local and remote write requests
for finding requests and for conflict detection, use the same tree for
both purposes. Introduce a flag to allow distinguishing the two
possible types of entries in this tree.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
This flag is set when a processes puts itself to sleep to wait for a
conflicting request to complete.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
These things are only used there.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Now drbd communication with protocol 100 actually works.
Replaced the remaining p_header80 with p_header where we
no longer know which header it is.
In the places where p_header80 is still in use, it is on
purpose, because we know that it is an old style header
there.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
The new header layout will only be used if the peer supports
it of course.
For the first packet and the handshake packet the old (h80)
layout is used for compatibility reasons.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
recv_bm_rle_bits() should not make any assumptions abou the layout
of the packet header
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Remove the file name and line number from the syslog messages generated:
we have no duplicate function names, and no function contains the same
assertion more than once.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Factor out duplicate code in got_NegAck().
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Get rid of the ar_id_to_req() and ack_id_to_req() wrappers.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Unify the ar_id_to_req() and ack_id_to_req() functions: make both fail
if the consistency check fails. Move the request lookup code now
duplicated in both functions into its own function.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Move _ar_id_to_req() to drbd_receiver.c and mark it non-inline. Remove
the leading underscores from _ar_id_to_req() and _ack_id_to_req(). Mark
ar_hash_slot() inline.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
This is the only place where this function is used. Make it static.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
The ID_VACANT definition has become entirely irrelevant by now.
The is_syncer_block_id() macro does not improve the code, so eliminated
it.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Converting the constants happens at compile time.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
DRBD_MAGIC has nothing to do with block ids and the funny values
computed were not actually used, anyway.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
If we have an asymetrically congested network, we may send P_PING,
but due to congestion, the corresponding P_PING_ACK would time out,
and we would drop a (congested, but otherwise) healthy connection
("PingAck did not arrive in time.")
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Found these with the help of ispell -l.
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
The old (optimistic) implementation could shrink the bio size
on an primary device.
Shrinking the bio size on a primary device is bad. Since there
we might get BIOs with the old (bigger) size shortly after
we published the new size.
The new implementation is more conservative, and eventually
increases the max_bio_size on a primary device (which is valid).
It does so, when it knows the local limit AND the remote limit.
We cache the last seen max_bio_size of the peer in the meta
data, and rely on that, to make the operation of single
nodes more efficient.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
It seems that the real cause of all the issues where that
we did not noticed in drbd_try_connect() when the other
guy closes one socket if the round trip time gets higher
than 100ms. There were that 100ms hard coded!
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
If there is no replication traffic within the idle timeout
(ping-int seconds), DRBD will send a P_PING,
and adjust the timeout to ping-timeout.
If there is no P_PING_ACK received within this ping-timeout,
DRBD finally drops the connection, and tries to re-establish it.
To decide which timeout was active, we compared the current timeout
with the ping-timeout, and dropped the connection, if that was the case.
By default, ping-int is 10 seconds, ping-timeout is 500 ms.
Unfortunately, if you configure ping-timeout to be the same as ping-int,
expiry of the idle-timeout had been mistaken for a missing ping ack,
and caused an immediate reconnection attempt.
Fix:
Allow both timeouts to be equal, use a local variable
to store which timeout is active.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Just deal with it more gracefully, if we fail to add even a single page
to an empty bio. We used to BUG_ON() there, but it has been observed in
some Xen deployment, so we need to handle that case more robustly now.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
If we fail to send the information that we lost our disk,
we have no connection, and no disk: no access to data anymore.
That is either expected (deconfiguration), or there will be so much
noise in the logs that "Sending state failed" is not useful at all.
Drop it.
If the reason for a shorter than expected receive was a signal,
which we sent because we already decided to disconnect,
these additional log messages are confusing and useless.
This patch follows this pattern:
- dev_warn(DEV, "short read expecting header on sock: r=%d\n", r);
+ if (!signal_pending(current))
+ dev_warn(DEV, "short read expecting header on sock: r=%d\n", r);
Also make them all dev_warn for consistency.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Now that we do no longer in-place endian-swap the bitmap, we allow
selected bitmap operations (testing bits, sometimes even settting bits)
during some bulk operations.
This caused us to hit a lot of FIXME asserts similar to
FIXME asender in drbd_bm_count_bits,
bitmap locked for 'write from resync_finished' by worker
Which now is nonsense: looking at the bitmap is perfectly legal
as long as it is not being resized.
This cosmetic patch defines some flags to describe expectations in finer
detail, so the asserts in e.g. bm_change_bits_to() can be skipped if
appropriate.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
All decisions about sync, sync direction, and wether or not to
allow a connect or attach are based on our set of UUIDs to tag a
data generation.
Log changes to the UUIDs whenever they occur,
logging "new current UUID P:Q:R:S" is more useful
than "Creating new current UUID".
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
The "lazy writeout" of cleared bitmap pages happens during resync, and
should happen again once the resync finishes cleanly, or is aborted.
If resync finished cleanly, or was aborted because of peer disk
failure, we trigger the writeout from worker context in the after
state change work.
If resync was aborted because of connection failure, we should not
immediately trigger bitmap writeout, but rather postpone the
writeout to after the connection cleanup happened. We now do it
in the receiver context from drbd_disconnect().
If resync was aborted because of local disk failure, well, there
is nothing to write to anymore.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Protocol A has no P_WRITE_ACKs, but has P_NEG_ACKs.
The master bio might already be completed, therefore the
request is no longer in the collision hash.
=> Do not try to validate block_id as request
In Protocol B we might already have got a P_RECV_ACK
but then get a P_NEG_ACK after wards.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
The point is that drbd_disconnect() can be called with a cstate of
WFConnection.
That happens if the user issues "drbdsetup disconnect" while the
drbd_connect() function executes. Then drbdd_init() will call
drbdd(), which in turn will return without receiving any
packets. Then drbdd_init() will end up calling drbd_disconnect()
with a cstate of WFConnection.
Bottom line: This assertion is wrong as it is, and we do not
see value in fixing it. => Removing it.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
The test if rs_pending_cnt == 0 was too weak. Using Test for
unacked_cnt == 0 instead. Moved that into the worker.
Since unacked_cnt gets already increased when an P_RS_DATA_REQ
comes in.
Also using a timer to make Ahead -> SyncSource -> Ahead cycles
slower...
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
See also commit from 2009-08-15
"drbd_uuid_compare(): Do not full sync in case a P_SYNC_UUID packet gets lost."
We saw cases where the History UUIDs where not as expected. So the
detection of the special case did not trigger. With the sync UUID
no longer being a random number, but deducible from the previous
bitmap UUID, the detection of this special case becomes more
reliable.
The SyncUUID now is the previous bitmap UUID + 0x1000000000000.
Rule 5a:
Cs = H1p & H1p + Offset = Bp
Connection was lost before SyncUUID Packet came through.
Corrent (peer) UUIDs:
Bp = H1p
H1p = H2p
H2p = 0
Become Sync target.
Rule 7a:
Cp = H1s & H1s + Offset = Bs
Connection was lost before SyncUUID Packet came through.
Correct (own) UUIDs:
Bs = H1s
H1s = H2s
H2s = 0
Become Sync source.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
We may not get from SyncSource to Ahead if we have sent some
P_RS_DATA_REPLY packets to the peer and are waiting for
P_WRITE_ACK.
Again, this is not relevant for proper tuned systems, but makes
sure that the not-tuned system does not get diverging bitmaps.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
When the sync source node replies to a P_RS_DATA_REQUEST packet
when it is already in ahead mode. I.e. those two packets
crossed each other on the wire, that may lead to diverging
bitmaps.
This never happens in a well-tuned-system. In a well-tuned-
system the resync controller has reduced the resync speed
to zero long before we got into ahead-mode.
But we have to be prepared for the not-well-tuned-system
of course as well.
Because -> diverging bitmaps = non terminating resync.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
We expect to only receive the recently introduced "set out of sync"
packets in specific states. If we receive them in different states, that
may confuse the resync process to the point where it won't terminate, or
think it made negative progress.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
The FAULT_ACTIVE macro just wraps the drbd_insert_fault macro for no
apparent reason.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
* C_STARTING_SYNC_S, C_STARTING_SYNC_T In these states the bitmap gets
written to disk. Locking out of app-IO is done by using the
drbd_queue_bitmap_io() and drbd_bitmap_io() functions these days.
It is no longer necessary to lock out app-IO based on the connection
state.
App-IO that may come in after the BITMAP_IO flag got cleared before the
state transition to C_SYNC_(SOURCE|TARGET) does not get mirrored, sets
a bit in the local bitmap, that is already set, therefore changes nothing.
* C_WF_BITMAP_S In this state we send updates (P_OUT_OF_SYNC packets).
With that we make sure they have the same number of bits when going
into the C_SYNC_(SOURCE|TARGET) connection state.
* C_UNCONNECTED: The receiver starts, no need to lock out IO.
* C_DISCONNECTING: in drbd_disconnect() we had a wait_event()
to wait until ap_bio_cnt reaches 0. Removed that.
* C_TIMEOUT, C_BROKEN_PIPE, C_NETWORK_FAILURE
C_PROTOCOL_ERROR, C_TEAR_DOWN: Same as C_DISCONNECTING
* C_WF_REPORT_PARAMS: IO still possible since that is still
like C_WF_CONNECTION.
And we do not need to send barriers in C_WF_BITMAP_S connection state.
Allow concurrent accesses to the bitmap when receiving the bitmap.
Everything gets ORed anyways.
A drbd_free_tl_hash() is in after_state_chg_work(). At that point
all the work items of the last connections must have been processed.
Introduced a call to drbd_free_tl_hash() into drbd_free_mdev()
for paranoia reasons.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
We only issue resync requests if there is no significant application IO
going on. = Application IO has higher priority than resnyc IO.
If application IO can not be started because the resync process locked
an resync_lru entry, start the IO operations necessary to release the
lock ASAP.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
In this connection mode, the ahead node no longer replicates
application IO. The behind's disk becomes out dated.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
To ease tracking of bios in some hash tables, we want it to
not cross certain boundaries (128k, used to be 32k).
We limit the maximum bio size using queue parameters.
Historically some defines and variables we use there have been named
max_segment_size, which was misguided. Rename them to max_bio_size,
and use [blk_]queue_max_hw_sectors where appropriate.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
With data-integrity digest enabled, double-check on the sending side
for modifications by upper layers of buffers under write back,
so we can tell it appart from corruption on the "wire".
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Show progressbar and ETA always, with proc_details >= 1 also show the
current sector position for both resync and online-verify on both nodes.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
For partial (resumed) online verify, initialize the resync step marks
once we know what the online verify start sector is.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
For a partial (resumed) online-verify, initialize rs_total not to total
bits, but to number of bits to check in this run, to match the meaning
rs_total has for actual resync.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
With the plugging now being explicitly controlled by the
submitter, callers need not pass down unplugging hints
to the block layer. If they want to unplug, it's because they
manually plugged on their own - in which case, they should just
unplug at will.
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
Code has been converted over to the new explicit on-stack plugging,
and delay users have been converted to use the new API for that.
So lets kill off the old plugging along with aops->sync_page().
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
This should fix a performance degradation we observed recently.
If we don't expect any subheader, we should not call into the tcp stack,
as that may add considerable latency if there is no data available at
this point.
For a synthetic synchronous write load with single outstanding writes,
this additional latency when processing the "unplug remote" packet
added up to a performance degradation factor >= 10.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
The big kernel lock has been removed from all these files at some point,
leaving only the #include.
Remove this too as a cleanup.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* 'for-2.6.37/barrier' of git://git.kernel.dk/linux-2.6-block: (46 commits)
xen-blkfront: disable barrier/flush write support
Added blk-lib.c and blk-barrier.c was renamed to blk-flush.c
block: remove BLKDEV_IFL_WAIT
aic7xxx_old: removed unused 'req' variable
block: remove the BH_Eopnotsupp flag
block: remove the BLKDEV_IFL_BARRIER flag
block: remove the WRITE_BARRIER flag
swap: do not send discards as barriers
fat: do not send discards as barriers
ext4: do not send discards as barriers
jbd2: replace barriers with explicit flush / FUA usage
jbd2: Modify ASYNC_COMMIT code to not rely on queue draining on barrier
jbd: replace barriers with explicit flush / FUA usage
nilfs2: replace barriers with explicit flush / FUA usage
reiserfs: replace barriers with explicit flush / FUA usage
gfs2: replace barriers with explicit flush / FUA usage
btrfs: replace barriers with explicit flush / FUA usage
xfs: replace barriers with explicit flush / FUA usage
block: pass gfp_mask and flags to sb_issue_discard
dm: convey that all flushes are processed as empty
...
* 'for-2.6.37/drivers' of git://git.kernel.dk/linux-2.6-block: (95 commits)
cciss: fix PCI IDs for new Smart Array controllers
drbd: add race-breaker to drbd_go_diskless
drbd: use dynamic_dev_dbg to optionally log uuid changes
dynamic_debug.h: Fix dynamic_dev_dbg() macro if CONFIG_DYNAMIC_DEBUG not set
drbd: cleanup: change "<= 0" to "== 0"
drbd: relax the grace period of the md_sync timer again
drbd: add some more explicit drbd_md_sync
drbd: drop wrong debug asserts, fix recently introduced race
drbd: cleanup useless leftover warn/error printk's
drbd: add explicit drbd_md_sync to drbd_resync_finished
drbd: Do not log an ASSERT for P_OV_REQUEST packets while C_CONNECTED
drbd: fix for possible deadlock on IO error during resync
drbd: fix unlikely access after free and list corruption
drbd: fix for spurious fullsync (uuids rotated too fast)
drbd: allow for explicit resync-finished notifications
drbd: preparation commit, using full state in receive_state()
drbd: drbd_send_ack_dp must not rely on header information
drbd: Fix regression in recv_bm_rle_bits (compressed bitmap)
drbd: Fixed a stupid copy and paste error
drbd: Allow larger values for c-fill-target.
...
Fix up trivial conflict in drivers/block/ataflop.c due to BKL removal
If we get an IO-error during an activity log transaction,
if we failed to write the bitmap of the evicted extent,
we must not write the transaction itself.
If we failed to write the transaction,
we must not even submit the corresponding bio,
as its extent is not yet marked in the activity log.
Otherwise, if this was a disconneted Primary (degraded cluster), which
now lost its disk as well, and we later re-attach the same backend
storage, we possibly "forget" to resync some parts of the disk that
potentially have been changed.
On the receiving side, when receiving from a peer with unhealthy disk,
checking for pdsk == D_DISKLESS is not enough, we need to set out of
sync and do AL transactions for everything pdsk < D_INCONSISTENT on the
receiving side.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
If we have contention in drbd_al_begin_iod (heavy randon IO),
an administrative request to detach the disk may deadlock
for similar reasons as the recently fixed deadlock if detaching
because of IO-error.
The approach taken here is to either go through the intermediate
cleanup state D_FAILED, or first lock out application io,
don't just go directly to D_DISKLESS.
We need an additional state bit (WAS_IO_ERROR) to distinguish
the -> D_FAILED because of IO-error from other failures.
Sanitize D_ATTACHING -> D_FAILED to D_ATTACHING -> D_DISKLESS.
If only attaching, ldev may be missing still, but would be referenced
from within the after_state_ch for -> D_FAILED, potentially
dereferencing a NULL pointer.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
It sometimes may take a while for the after state change work to be
scheduled, which does drbd_md_sync. At convenient places, we should do
explicit drbd_md_sync to have the new state information on disk as soon
as possible.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
This might happen if on the VERIFY_S node the disk gets dropped.
Although this is an cluster wide state transition, the VERIFY_T node,
updates it connection state first. Then the ack packet for the
cluster wide state transition travels back, and the VERIFY_S node
stops to produce the P_OV_REQUEST packets.
There is absolutely nothing wrong with that.
Further, do not log "Can not satisfy peer's..." on the VERIFY_S
node in this case, but pretend that they had equal checksum.
[Bugz 327]
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Various cleanup paths have been incomplete, for the very unlikely case
that we cannot allocate enough bios from process context when submitting
on behalf of the peer or resync process.
Never observed.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Preparation patch so more drbd_send_state() usage on the peer
will not confuse drbd in receive_state().
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
no functional change, just using full state instead of just the .conn
part of it for comparisons.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
drbd commit 17c854fea474a5eb3cfa12e4fb019e46debbc4ec
drbd: receiving of big packets, for payloads between 64kByte and 4GByte
introduced a new on-the-wire packet header format. We must no longer
assume either format, but use the result of whatever drbd_recv_header
has decoded.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
We used to be16_to_cpu the length field in our received packet header.
drbd commit 17c854fea474a5eb3cfa12e4fb019e46debbc4ec
drbd: receiving of big packets, for payloads between 64kByte and 4GByte
changed this, but forgot to adjust a few places where we relied on
h->length being in native byte order.
This broke the receiving side of the RLE compressed bitmap exchange.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Two missing corner cases to the "maximum packet size" handshake.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
There are three ways to get IO suspended:
* Loss of any access to data
* Fence-peer-handler running
* User requested to suspend IO
Track those in different bits, so that one condition clearing its
state bit does not interfere with the other two conditions.
Only when the user resumes IO he overrules all three bits.
The fact is hidden from the user, he sees only a single suspend
bit.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
If we cannot satisfy a request (because our disk just broke),
we still need to drain the payload. Or we'll get a protocol error
when interpreting the payload as DRBD packet header.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
BUG trace would look like:
lc_find
drbd_rs_complete_io
got_OVResult
drbd_asender
Could be triggered by explicit, or IO-error policy based,
detach during online-verify.
We may only dereference mdev->resync, if we first get_ldev(), as the
disk may break any time, causing mdev->resync to disappear once all
ldev references have been returned.
Already in flight online-verify requests or replies may still come in,
which we then need to ignore.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
The actual race happened int the drbd_start_resync() function. Where
drbd_resync_finished() -> __drbd_set_state() set STOP_SYNC_TIMER and
armed the timer.
If the timer fired before execution reaches the mod_timer statement
at the end of drbd_start_resync() the latter would cause an
unexpected call to w_make_resync_request().
Removed the STOP_SYNC_TIMER bit, and base it on the connection state.
The STOP_SYNC_TIMER bit probably originates probably the time before
the state engine.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Now we have multiple BIOs per ee, packets with a 32 bit length field,
it gets time to use these goodies.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
We now track the data rate of locally submitted resync related requests,
and can thus detect non-resync activity on the lower level device.
If the current sync rate is above c-min-rate, and the lower level device
appears to be busy, we throttle the resyncer.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
also canonicalize the return values of read_for_csum
and drbd_rs_begin_io to return -ESOMETHING, or 0 for success.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
We may not free tl_hash when IO is suspended, since we can not wait
until ap_bio_cnt reaches zero.
We can do this after susp reched 0, since then tl_clear was called
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
After disconnect (most likely mdev->net_cnt == 0) and we are
still in an unstable state (!drbd_state_is_stable()). When we
get an IO request in drbd_get_max_buffers() (called from
__inc_ap_bio_cond(), called from inc_ap_bio()) we wake up
misc_wait. Misc_wait is also used in inc_ap_bio() to sleep
until the outcome of __inc_ap_bio_cond() changes. => Busy loop!
Solution: Have a dedicated wait queue for get_net_conf() and
put_net_conf().
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Make sure the state engine can deny two primaries to connect
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Since we can not thaw the transfer log, the next logical step is
to allow reconnects while the fence-peer handler runs.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
If IO was frozen for a temporal network outage, resend the
content of the transfer-log into the newly established connection.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
With that the drbd_fail_pending_reads() function becomes obsolete.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
All the blkdev_issue_* helpers can only sanely be used for synchronous
caller. To issue cache flushes or barriers asynchronously the caller needs
to set up a bio by itself with a completion callback to move the asynchronous
state machine ahead. So drop the BLKDEV_IFL_WAIT flag that is always
specified when calling blkdev_issue_* and also remove the now unused flags
argument to blkdev_issue_flush and blkdev_issue_zeroout. For
blkdev_issue_discard we need to keep it for the secure discard flag, which
gains a more descriptive name and loses the bitops vs flag confusion.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
We have several users of min_not_zero, each of them using their own
definition. Move the define to kernel.h.
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Jens Axboe <axboe@carl.home.kernel.dk>
* 'for-2.6.36' of git://git.kernel.dk/linux-2.6-block: (149 commits)
block: make sure that REQ_* types are seen even with CONFIG_BLOCK=n
xen-blkfront: fix missing out label
blkdev: fix blkdev_issue_zeroout return value
block: update request stacking methods to support discards
block: fix missing export of blk_types.h
writeback: fix bad _bh spinlock nesting
drbd: revert "delay probes", feature is being re-implemented differently
drbd: Initialize all members of sync_conf to their defaults [Bugz 315]
drbd: Disable delay probes for the upcomming release
writeback: cleanup bdi_register
writeback: add new tracepoints
writeback: remove unnecessary init_timer call
writeback: optimize periodic bdi thread wakeups
writeback: prevent unnecessary bdi threads wakeups
writeback: move bdi threads exiting logic to the forker thread
writeback: restructure bdi forker loop a little
writeback: move last_active to bdi
writeback: do not remove bdi from bdi_list
writeback: simplify bdi code a little
writeback: do not lose wake-ups in bdi threads
...
Fixed up pretty trivial conflicts in drivers/block/virtio_blk.c and
drivers/scsi/scsi_error.c as per Jens.
It was a now abandoned attempt to throttle resync bandwidth
based on the delay it causes on the bulk data socket.
It has no userbase yet, and has been disabled by
9173465ccb51c09cc3102a10af93e9f469a0af6f already.
This removes the now unused code.
The basic feature, namely using up "idle" bandwith
of network and disk IO subsystem, with minimal impact
to application IO, is being reimplemented differently.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
Remove the current bio flags and reuse the request flags for the bio, too.
This allows to more easily trace the type of I/O from the filesystem
down to the block driver. There were two flags in the bio that were
missing in the requests: BIO_RW_UNPLUG and BIO_RW_AHEAD. Also I've
renamed two request flags that had a superflous RW in them.
Note that the flags are in bio.h despite having the REQ_ name - as
blkdev.h includes bio.h that is the only way to go for now.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
drbd/drbd_receiver.c: linux/mm.h is included more than once.
Signed-off-by: Andrea Gelmini <andrea.gelmini@gelma.net>
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
quoting tcp(7):
On individual connections, the socket buffer size must be set prior to the
listen(2) or connect(2) calls in order to have it take effect.
This adds a wrapper to do so, and uses it appropriately.
Improves performance in certain situations.
Note that because we cannot easily determine which socket will be
"meta" and wich "data" (bulk) socket, we adjust both sockets.
Previously, DRBD only adjusted the bufsizes of the "data" socket.
Thanks again to Eduard.Guzovsky@stratus.com.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
The late-UUID writing is delayed until the next release.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
The choice was to either delay creation of the new UUID until
IO got thawed or to delay it until the first IO request.
Both are correct, the later is more friendly to users of
dual-primary setups, that actually only write on one side.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Now that the peer may handle multi-bio EEs,
we can ignore the peer's limit,
and concentrate on the limits of the local IO stack.
This is safe accross drbd protocol versions,
as our queue_max_sectors() will be adjusted accordingly.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
This should allow for better performance if the lower level IO stack
of the peers differs in limits exposed either via the queue,
or via some merge_bvec_fn.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Delay_probes are new packets in the DRBD protocol, which allow
DRBD to know the current delay packets have on the data socket.
(relative to the meta data socket)
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Some wish to be notified of all instances of split brain, not just those that
go unresolved. The initial-split-brain handler is called to notify someone
upon detection of all split brain conditions even if auto-recovery policies
are configured.
Signed-off-by: Adam Gandelman <adam.gandelman@linbit.com>
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Even if it should never happen if the peer does behave, we need to
double check, and not even attempt access beyond end of device.
It usually would be caught by lower layers, resulting in "IO error",
but may also end up in the internal meta data area.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Don't forget to drain the digest in case we cannot satisfy a
checksum based resync or online-verify request.
It would additionally cause a protocoll error,
dropping the connection.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
block_id may be ID_SYNCER,
as well as checksum based resync request magic, or online verify magic.
Let's just drop that ASSERT.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
The patch just convert all blkdev_issue_xxx function to common
set of flags. Wait/allocation semantics preserved.
Signed-off-by: Dmitry Monakhov <dmonakhov@openvz.org>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
Fixes a race and potential kernel panic if e.g. the worker was just
about to send a few P_RS_IS_IN_SYNC via the meta socket for checksum
based resync, while the receiver destroys the sockets in
drbd_disconnect.
To make sure no-one is using the meta socket,
it is not enough to stop the asender...
Grab the meta socket mutex before destroying it.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Situation:
we have diverging data sets, i.e. we had a split brain somewhen,
but currently are connected, one node diskless.
Then we try to attach that disk, figure it is consistent,
but has a diverging data set, we refuse to attach.
This led to strange state changes:
22:18:35 bb drbd1: peer( Unknown -> Primary ) conn( WFReportParams -> Connected) pdsk( DUnknown -> UpToDate )
22:19:30 bb drbd1: disk( Diskless -> Attaching )
22:19:30 bb drbd1: disk( Attaching -> Negotiating )
22:19:30 bb drbd1: drbd_sync_handshake:
22:19:30 bb drbd1: self 97BF25798B9D5222:F33D1F62ADE698DD:4269796F9D027C83:AC45D8B5C3C1BF93 bits:19449 flags:0
22:19:30 bb drbd1: peer 280DFB6E125465D3:F33D1F62ADE698DC:4269796F9D027C82:AC45D8B5C3C1BF93 bits:2575806 flags:0
22:19:30 bb drbd1: uuid_compare()=100 by rule 90
22:19:30 bb drbd1: Split-Brain detected, dropping connection!
22:19:30 bb drbd1: disk( Negotiating -> Diskless )
while the other side says:
22:19:30 aa drbd1: Split-Brain detected, dropping connection!
22:19:30 aa drbd1: Disk attach process on the peer node was aborted.
22:19:30 aa drbd1: conn( Connected -> TOO_LARGE ) pdsk( Diskless -> Consistent )
This should be fixed now.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Test the just-allocated value for NULL rather than some other value.
The semantic patch that makes this change is as follows:
(http://coccinelle.lip6.fr/)
// <smpl>
@@
expression x,y;
statement S;
@@
x = \(kmalloc\|kcalloc\|kzalloc\)(...);
(
if ((x) == NULL) S
|
if (
- y
+ x
== NULL)
S
)
// </smpl>
Signed-off-by: Julia Lawall <julia@diku.dk>
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Immediately after changing the write ordering method, the epoch can already
be finished at this point.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Since 8.3.3 we fail to do the resync when a partial resynch is not
possible, but a full synch is necessary.
This regression was introduced with 7101539930c0a89146959e7a39c09ad9c3516434
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Otherwise the 'state fixup' in the receiver will change to Unconnected,
but the receiver will terminate itself, and any attempt at 'down'ing
that drbd later will block forever.
see also Bugz. #259
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
When there are many blocks on the fly (ua), and the AL gets into "starving"
mode (random IO, scattered all over the device), and the connections gets
interrupted, the receiver thread deadlocks in the drbd_disconnect() code path.
Affected are only nodes in Primary role.
The bug triggers most likely on system that mirror over "long distances"
Regression introduced shortly before 8.3.3
with git commit 31e0f1250f174ac1ee317f360943a0159e19edc8
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
It is force-included on the gcc command line since at least 2.6.15.
Explicit include lines seem to break compilation now in certain configurations.
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Kamalesh Babulal <kamalesh@linux.vnet.ibm.com>
Acked-by: Sam Ravnborg <sam@ravnborg.org>
Andreas Gruenbacher