- 09 Nov, 2012 30 commits
-
-
Philipp Reisner authored
The previous way of doing the state change was also okay since the state change on the susp flag gets propagated from the mdev to the tconn. Fortunately all this goes away in drbd-9.0 Signed-off-by:
Philipp Reisner <philipp.reisner@linbit.com> Signed-off-by:
Lars Ellenberg <lars.ellenberg@linbit.com>
-
Philipp Reisner authored
Signed-off-by:
-
Lars Ellenberg authored
If the md_sync_timer triggers a second time, while the work queued during the first time is still pending, this could result in list_add() of an already added item, and corrupt the work item list. This likely only triggered because of the erroneous batch-dequeueing of work items fixed with drbd: dequeue single work items in wait_for_work() Still, skip queueing if md_sync_work is already queued. Signed-off-by:
-
Lars Ellenberg authored
As long as we still use drbd_queue_work_front(), we must only dequeue the single first item during normal operation. The comment in drbd_worker() even says so, but bc8a5a1 drbd: remove struct drbd_tl_epoch objects (barrier works) introduced the batch dequeueing again via list_splice_init() in wait_for_work(). Change back to list_move() of the first item, if any. Signed-off-by:
-
Lars Ellenberg authored
Documentation of mutex_unlock says we must not use it in interrupt context. So do not call it while holding the spin_lock_irq, but give up the spinlock temporarily. Signed-off-by:
-
Philipp Reisner authored
If the preconditions for a state change change after the wait_event() we might hit the BUG() statement in conn_set_state(). With holding the spin_lock while evaluating the condition AND until the actual state change we ensure the the preconditions can not change anymore. Signed-off-by:
-
Lars Ellenberg authored
drbd_adm_disk_opts() does wait_event(mdev->al_wait, lc_try_lock(mdev->act_log)); drbd_al_shrink(mdev); If the device is very busy, this can take a very long time to succeed. Fix this by temporarily suspending IO, then quickly change the settings, and resume. Signed-off-by:
-
Lars Ellenberg authored
We must only send P_BARRIER for epochs we actually sent P_DATA in. If we (re-)establish a connection, we reinitialized the send.current_epoch_nr, but forgot to reset send.current_epoch_writes. This could result in a spurious P_BARRIER with stale epoch information, and a disconnect/reconnect cycle once the then "unexpected" P_BARRIER_ACK is received: BAD! BarrierAck #28823 received, expected #28829! Introduce re_init_if_first_write() and maybe_send_barrier() helpers, and call them appropriately for read/write/set-out-of-sync requests. Signed-off-by:
-
Lars Ellenberg authored
drbd_disconnected() is supposed to clear the resync lru cache, by calling drbd_rs_cancel_all(). We must do so before we call drbd_flush_workqueue(), as at least the callback w_restart_disk_io() may wait for resync progres, and would otherwise deadlock. drbd_finish_peer_reqs() may again populate that cache, which will then potentially be stale after the next resync handshake and bitmap exchange, we have to do it again after that. A stale resync lru cache causes no harm but ugly messages like this: BAD! sector=196608s enr=6 rs_left=-256 rs_failed=0 count=256 cstate=SyncTarget Signed-off-by:
-
Philipp Reisner authored
Signed-off-by:
-
Philipp Reisner authored
Disconnecting is a cluster wide state change. In case the peer node agrees to the state transition, it sends back the fact on the meta-data connection and closes both sockets. In case the node node that initiated the state transfer sees the closing action on the data-socket, before the P_STATE_CHG_REPLY packet, it was going into one of the network failure states. At least with the fencing option set to something else thatn "dont-care", the unclean shutdown of the connection causes a short IO freeze or a fence operation. Signed-off-by:
-
Philipp Reisner authored
There is at least the worker context, the receiver context, the context of receiving netlink packts. Signed-off-by:
-
Philipp Reisner authored
We need to write the whole bitmap after we moved the meta data due to an online resize operation. With the support for one peta byte devices bitmap IO was optimized to only write out touched pages. This optimization must be turned off when writing the bitmap after an online resize. This issue was introduced with drbd-8.3.10. The impact of this bug is that after an online resize, the next resync could become larger than expected. Signed-off-by:
-
Philipp Reisner authored
In various places (E.g. CONNECTION_LOST_WHILE_PENDING) the RQ_COMPLETION_SUSP mask is passed in the clear set to mod_rq_state(). The issue was that it tried to clear the RQ_COMPLETION_SUSP bit out of the state mask first, and eventuelly set it afterwards, in the drbd_req_put_completion_ref() function. Fixed that by moving the reference getting out of drbd_req_put_completion_ref() into the mod_rq_state(), before the place where the extra reference might be put. Signed-off-by:
-
Andreas Gruenbacher authored
Signed-off-by:
-
Lars Ellenberg authored
If for some reason (typically "split-brained" cluster manager) drbd replica data has diverged, we can chose a victim, and reconnect using "--discard-my-data", causing the victim to become sync-target, fetching all changed blocks from the peer. If we are Primary, we are potentially in use, and we refuse to "roll back" changes to the data below the page cache and other users. Rename the error symbol for this to ERR_DISCARD_IMPOSSIBLE. Signed-off-by:
-
Lars Ellenberg authored
We don't discard anything here, really. We resolve conflicting, concurrent writes to overlapping data blocks. Signed-off-by:
-
Lars Ellenberg authored
To avoid confusion with REQ_DISCARD aka TRIM, rename our "discard concurrent write acks" from P_DISCARD_WRITE to P_SUPERSEDED. At the same time, rename the drbd request event DISCARD_WRITE to CONFLICT_RESOLVED. It already triggers both successful completion or restart of the request, depending on our RQ_POSTPONED flag. Signed-off-by:
-
Lars Ellenberg authored
Signed-off-by:
-
Lars Ellenberg authored
Don't drop a request from the transfer log just because it was NEG_ACKED. We need it around to be able to verify P_BARRIER_ACKs against the transver log. Signed-off-by:
-
Lars Ellenberg authored
Almost all code paths calling start_new_tl_epoch() guarded it with if (... current_tle_writes > 0 ... ). Just move that inside start_new_tl_epoch(). Signed-off-by:
-
Philipp Reisner authored
Requests of an acked epoch are stored on the barrier_acked_requests list. In case the private bio of such a request completes while IO on the drbd device is suspended [req_mod(completed_ok)] then the request stays there. When thawing IO because the fence_peer handler returned, then we use tl_clear() to apply the connection_lost_while_pending event to all requests on the transfer-log and the barrier_acked_requests list. Up to now the connection_lost_while_pending event was not applied on requests on the barrier_acked_requests list. Fixed that. I.e. now the connection_lost_while_pending and resend events are applied to requests on the barrier_acked_requests list. For that it is necessary that the resend event finishes (local only) READS correctly. Signed-off-by:
-
Lars Ellenberg authored
The DISCARD_CONCURRENT flag should be set on one node and cleared on the other node. As the code was before it was theoretical possible that a node accepts the meta socket, but has to close it later on, and keeps the DISCARD_CONCURRENT flag. Correct this by moving the clear_bit(DISCARD_CONCURRENT) where the packet gets sent. Signed-off-by:
-
Lars Ellenberg authored
DRBD has a concept of request epochs or reorder-domains, which are separated on the wire by P_BARRIER packets. Older DRBD is not able to handle zero-sized requests at all, so we need to map empty flushes to these drbd barriers. These are the equivalent of empty flushes, and by default trigger flushes on the receiving side anyways (unless not supported or explicitly disabled), so there is no need to handle this differently in newer drbd either. Signed-off-by:
-
Philipp Reisner authored
Since the listening socket is open all the time, it was possible to get into stable "initial packet S crossed" loops. * when both sides realize in the drbd_socket_okay() call at the end of the loop that the other side closed the main socket you had the chance to get into a stable loop with repeated "packet S crossed" messages. * when both sides do not realize with the drbd_socket_okay() call at the end of the loop that the other side closed the main socket you had the chance to get into a stable loop with alternating "packet S crossed" "packet M crossed" messages. In order to break out these stable loops randomize the behaviour if such a crossing of P_INITIAL_DATA or P_INITIAL_META packets is detected. Signed-off-by:
-
Philipp Reisner authored
Since now our listening socket is open all the time we will get connection tries of the peer always in. No need to try it three times. This is valid when connecting to older peers as well, it simply increases the probability that the new version DRBD will accept a connection instead that it will establish one. Signed-off-by:
-
Philipp Reisner authored
Since the drbd_socket_okay() function itself tests if the the socket is NULL, the explicit test "if (sock.socket && &msock.socket)" was redundent. Apart from that the address opperator ('&') before msock.socket rendered the test pointless. Signed-off-by: -
Philipp Marek authored
Signed-off-by:
-
Lars Ellenberg authored
In 8.4, we may have bios spanning two activity log extents. Fixup drbd_al_begin_io() and drbd_al_complete_io() to deal with zero sized bios. Signed-off-by:
-
Lars Ellenberg authored
We now can schedule only a specific range of sectors for online verify, or interrupt a running verify without interrupting the connection. Had to bump the protocol version differently, we are now 101. Added verify_can_do_stop_sector() { protocol >= 97 && protocol != 100; } Also, the return value convention for worker callbacks has changed, we returned "true/false" for "keep the connection up" in 8.3, we return 0 for success and <= for failure in 8.4. Affected: receive_state() Signed-off-by:
-
- 08 Nov, 2012 10 commits
-
-
Lars Ellenberg authored
If you do back to back wait-sync/invalidate on a Primary in a tight loop, during application IO load, you could trigger a race: kernel: block drbd6: FIXME going to queue 'set_n_write from StartingSync' but 'write from resync_finished' still pending? Fix this by changing the order of the drbd_queue_work() and the wake_up() in dec_ap_pending(), and adding the additional drbd_flush_workqueue() before requesting the full sync. Signed-off-by: -
Lars Ellenberg authored
In case we want to hard-reset from the local-io-error handler, we need to call it before notifying the peer or aborting local IO. Otherwise the peer will advance its data generation UUIDs even if secondary. This way, local io error looks like a "regular" node crash, which reduces the number of different failure cases. This may be useful in a bigger picture where crashed or otherwise "misbehaving" nodes are automatically re-deployed. Signed-off-by:
-
Lars Ellenberg authored
Fix asserts like block drbd0: in got_BlockAck:4634: rs_pending_cnt = -35 < 0 ! We reset the resync lru cache and related information (rs_pending_cnt), once we successfully finished a resync or online verify, or if the replication connection is lost. We also need to reset it if a resync or online verify is aborted because a lower level disk failed. In that case the replication link is still established, and we may still have packets queued in the network buffers which want to touch rs_pending_cnt. We do not have any synchronization mechanism to know for sure when all such pending resync related packets have been drained. To avoid this counter to go negative (and violate the ASSERT that it will always be >= 0), just do not reset it when we lose a disk. It is good enough to make sure it is re-initialized before the next resync can start: reset it when we re-attach a disk. Signed-off-by:
-
Lars Ellenberg authored
We cache the congestion status in mdev->congestion_reason whenever drbd_congested() was called. Reset this cached info before reporting it when reading /proc/drbd. Signed-off-by:
-
Lars Ellenberg authored
If the drbd worker thread is synchronously waiting for some userland callback, we don't want some casual pageout to block on us. Have drbd_congested() report congestion in that case. Signed-off-by:
-
Lars Ellenberg authored
Aborting local requests (not waiting for completion from the lower level disk) is dangerous: if the master bio has been completed to upper layers, data pages may be re-used for other things already. If local IO is still pending and later completes, this may cause crashes or corrupt unrelated data. Only abort local IO if explicitly requested. Intended use case is a lower level device that turned into a tarpit, not completing io requests, not even doing error completion. Signed-off-by:
-
Lars Ellenberg authored
The two unused "global flags" in 8.3 are "per volume" flags in 8.4. Still, they are unused, so lose them. Signed-off-by:
-
Lars Ellenberg authored
We must not look at mdev->actlog, unless we have a get_ldev() reference. It also does not make much sense to try to disconnect or pull-ahead of the peer, if we don't have good local data. Only even consider congestion policies, if our local disk is D_UP_TO_DATE. Signed-off-by:
-
Lars Ellenberg authored
drbd_adm_down() does adm_detach(), which can fail with various error codes, or be interrupted by a signal. The interrupted by signal case was not properly handled, leading to block drbd0: ASSERT( mdev->state.disk == D_DISKLESS && mdev->state.conn == C_STANDALONE ) in drbd/drbd_worker.c and further to destroying objects while still in use, and resulting crashes. Detect the interruption, and take the error path out. Signed-off-by:
-
Lars Ellenberg authored
Sometimes, a lower level block device turns into a tar-pit, not completing requests at all, not even doing error completion. We can force-detach from such a tar-pit block device, either by disk-timeout, or by drbdadm detach --force. Queueing for retry only from the request destruction path (kref hit 0) makes it impossible to retry affected read requests from the peer, until the local IO completion happened, as the locally submitted bio holds a reference on the drbd request object. If we can only complete READs when the local completion finally happens, we would not need to force-detach in the first place. Instead, queue for retry where we otherwise had done the error completion. Signed-off-by:
-
