- 10 Jul, 2014 40 commits
-
-
Lars Ellenberg authored
Adding requests to per-device fifo lists as soon as possible after allocating them leaves a simple list_first_entry_or_null() to find the oldest request, regardless what it is still waiting for. Signed-off-by:
Philipp Reisner <philipp.reisner@linbit.com> Signed-off-by:
Lars Ellenberg <lars.ellenberg@linbit.com>
-
Lars Ellenberg authored
Record (in jiffies) how much time a request spends in which stages. Followup commits will use and present this additional timing information so we can better locate and tackle the root causes of latency spikes, or present the backlog for asynchronous replication. Signed-off-by:
-
Lars Ellenberg authored
For diagnostic purposes, track intent, start time and latest submit time of meta data IO. Move separate members from struct drbd_device into the embeded struct drbd_md_io. s/md_io_(page|in_use)/md_io.\1/ Signed-off-by:
-
Lars Ellenberg authored
drbd_destroy_device means to give up reference counts on the connection(s) reachable via the peer_device(s). It must not do that by iterating via device->resource->connections, resource and connections may have already been disassociated by drbd_free_resource, and we'd leak connection refs. Instead, iterate via device->peer_devices->connection. Signed-off-by:
-
Lars Ellenberg authored
Now that we have additional asynchronous kref_get/kref_put via debugfs, make sure we catch access after free. Poison struct drbd_device, drbd_connection and drbd_resource before kfree() with 0xfd, 0xfc, and 0xf2, respectively. Signed-off-by:
-
Lars Ellenberg authored
To be able to find and present such zero-out fallback peer_requests in debugfs, we add those to "active_ee", once that list drained. Signed-off-by:
-
Lars Ellenberg authored
Keep the epoch entry lists (active_ee, read_ee, sync_ee, ...) consistently "oldest first". That way finding the oldest not yet successfully processed request is simply list_first_entry_or_null. Signed-off-by:
-
Lars Ellenberg authored
The only user of drbd_md_flush was bm_rw(), and it is always followed by either a drbd_md_sync(), or an al_write_transaction(), which, if so configured, both end up submiting a FLUSH|FUA request anyways. Signed-off-by:
-
Lars Ellenberg authored
We sometimes do if (list_empty(&w.list)) drbd_queue_work(&q, &w.list); Removal (list_del_init) may happen outside all locks, after all pending work entries have been moved to an on-stack local work list. For not dynamically allocated, but embeded, work structs, we must avoid to re-add until it really was removed. Move that list_empty check inside the spin_lock(&q->q_lock) within the helper function, and change to list_empty_careful(). This may have been the reason for a list_add corruption inside drbd_queue_work(). Signed-off-by: -
Lars Ellenberg authored
We try to limit the number of "in-flight" resync requests. One condition for that is the amount of requested data should not exceed half of what can be covered by our "max-buffers" setting. However we compared number of 4k pages with number of in-flight 512 Byte sectors, and this extra throttle triggered much earlier than intended. Signed-off-by:
-
Lars Ellenberg authored
Cosmetic change only. Signed-off-by:
-
Lars Ellenberg authored
If we lost a disk during the first resync after primary crash, we could have prematurely cleared the CRASHED_PRIMARY flag. Testing on C_CONNECTED is not what we meant there, but testing for both peers to become D_UP_TO_DATE. Signed-off-by:
-
Lars Ellenberg authored
If we throttle resync because the socket sendbuffer is filling up, tell TCP about it, so it may expand the sendbuffer for us. Signed-off-by:
-
Joe Perches authored
Just about all of these have been converted to __func__, so convert the last uses. Signed-off-by:
Joe Perches <joe@perches.com> Signed-off-by:
-
Monam Agarwal authored
This patch replaces rcu_assign_pointer(x, NULL) with RCU_INIT_POINTER(x, NULL) The rcu_assign_pointer() ensures that the initialization of a structure is carried out before storing a pointer to that structure. And in the case of the NULL pointer, there is no structure to initialize. So, rcu_assign_pointer(p, NULL) can be safely converted to RCU_INIT_POINTER(p, NULL) Signed-off-by:
Monam Agarwal <monamagarwal123@gmail.com> Signed-off-by:
-
Lars Ellenberg authored
If we already "pulled ahead", we can short-circuit, and avoid logging the same messages over and over again. Signed-off-by:
-
Lars Ellenberg authored
If "dirty" blocks are written to during resync, that brings them in-sync. By explicitly requesting write-acks during resync even in protocol != C, we now can actually respect this. Signed-off-by:
-
Philipp Reisner authored
In setups involving a DRBD-proxy and connections that experience a lot of buffer-bloat it might be necessary to set ping-timeout to an unusual high value. By default DRBD uses the same value to wait if a newly established TCP-connection is stable. Since the DRBD-proxy is usually located in the same data center such a long wait time may hinder DRBD's connect process. In such setups socket-check-timeout should be set to at least to the round trip time between DRBD and DRBD-proxy. I.e. in most cases to 1. Signed-off-by:
-
Philipp Reisner authored
Before the patch 'drbd: Keep the listening socket open while trying to connect to the peer' the newly created socket inherited the receive timeout from the listen socket. The listen socket had a receive timeout of connect-intervall +- 30% random jitter. The real issue is that after the mentioned patch we had no timeout at all. Now use 4 times the ping-timeout. Signed-off-by:
-
Lars Ellenberg authored
Checksum based resync trades CPU cycles for network bandwidth, in situations where we expect much of the to-be-resynced blocks to be actually identical on both sides already. In a "network hickup" scenario, it won't help: all to-be-resynced blocks will typically be different. The use case is for the resync of *potentially* different blocks after crash recovery -- the crash recovery had marked larger areas (those covered by the activity log) as need-to-be-resynced, just in case. Most of those blocks will be identical. This option makes it possible to configure checksum based resync, but only actually use it for the first resync after primary crash. Signed-off-by:
-
Lars Ellenberg authored
During handshake, we compare backend sizes, and user set limits, and agree on what device size we are going to expose. We remember that last-agreed-size in our meta data. But if we come up diskless, we have to accept what the peer presents us with. We used to accept the peers maximum potential capacity (backend size), which is wrong, and could lead to IO errors due to access beyond end of device. Instead, we need to accept the peer's current size. Unless that is communicated as 0, in which case we accept the backend size, or the user set limit, if set. Signed-off-by:
-
Lars Ellenberg authored
We intentionally do not serialize /proc/drbd access with internal state changes or statistic updates. Because of that, cat /proc/drbd may race with resync just being finished, still see the sync state, and find information about number of blocks still to go, but then find the total number of blocks within this resync has just been reset to 0 when accessing it. This now produces bogus numbers in the resync speed estimates. Fix by accessing all relevant data only once, and fixing it up if "still to go" happens to be more than "total". Signed-off-by:
-
Andreas Gruenbacher authored
Get rid of dump_stack() debug statements. There is no point whatsoever in registering and unregistering a reboot notifier that doesn't do anything. The intention was to switch to an "emergency read-only" mode, so we won't have to resync the full activity log just because we had been Primary before the reboot. Once we have that implemented, we may re-introduce the reboot notifier. Signed-off-by:
-
Lars Ellenberg authored
Signed-off-by:
-
Lars Ellenberg authored
The textual representation of resync extents in /proc/drbd presented with proc_details >= 3 was wrong, it used bitnumbers as bitmasks. It was not particularly useful either, and I doubt anyone has even tried to look at it in the last few years. Drop it. Signed-off-by:
-
Lars Ellenberg authored
The last user was al_write_transaction, if called with "delegate", and the last user to call it with "delegate = true" was the receiver thread, which has no need to delegate, but can call it himself. Finally drop the delegate parameter, drop the extra w_al_write_transaction callback, and drop drbd_queue_work_front. Do not (yet) change dequeue_work_item to dequeue_work_batch, though. Signed-off-by:
-
Lars Ellenberg authored
This replaces the md_sync_work member of struct drbd_device by a new MD_SYNC "work bit" in device->flags. This replaces the resync_start_work member of struct drbd_device by a new RS_START "work bit" in device->flags. Signed-off-by:
-
Lars Ellenberg authored
The recent fix to put_ldev() (correct ordering of access to local_cnt and state.disk; memory barrier in __drbd_set_state) guarantees that the cleanup happens exactly once. However it does not yet guarantee that the cleanup happens from worker context, the last put_ldev() may still happen from atomic context, which must not happen: blkdev_put() may sleep. Fix this by scheduling the cleanup to the worker instead, using a couple more bits in device->flags and a new helper, drbd_device_post_work(). Generalized the "resync progress" work to cover these new work bits. Signed-off-by:
-
Lars Ellenberg authored
BUG: unable to handle kernel NULL pointer dereference at 0000000000000058 IP: bd_release+0x21/0x70 Process drbd_w_t7146 Call Trace: close_bdev_exclusive drbd_free_ldev [drbd] drbd_ldev_destroy [drbd] w_after_state_ch [drbd] Race probably went like this: state.disk = D_FAILED ... first one to hit zero during D_FAILED: put_ldev() /* ----------------> 0 */ i = atomic_dec_return() if (i == 0) if (state.disk == D_FAILED) schedule_work(go_diskless) /* 1 <------ */ get_ldev_if_state() go_diskless() do_some_pre_cleanup() corresponding put_ldev(): force_state(D_DISKLESS) /* 0 <------ */ i = atomic_dec_return() if (i == 0) atomic_inc() /* ---------> 1 */ state.disk = D_DISKLESS schedule_work(after_state_ch) /* execution pre-empted by IRQ ? */ after_state_ch() put_ldev() i = atomic_dec_return() /* 0 */ if (i == 0) if (state.disk == D_DISKLESS) if (state.disk == D_DISKLESS) drbd_ldev_destroy() drbd_ldev_destroy(); Trying to fix this by checking the disk state *before* the atomic_dec_return(), which implies memory barriers, and by inserting extra memory barriers around the state assignment in __drbd_set_state(). Signed-off-by: -
Lars Ellenberg authored
For some reason we have assumed NOIDLE was implied by one of the other flags we set. It is not (anymore?). Explicitly set REQ_NOIDLE for synchronous meta data updates, or we can seriously starve random writes when using CFQ. Signed-off-by:
-
Lars Ellenberg authored
This probably does not have any real life impact, but we should first persist any potentially new UUID and other meta data flags, as well as our new role, before we allow/disallow write access. Signed-off-by:
-
Lars Ellenberg authored
This should reduce latency for such in-DRBD-protocol "pings", and may help reduce spurious disconnect/reconnect cycles due to "PingAck did not arrive in time." Signed-off-by:
-
Lars Ellenberg authored
The conf_update mutex used to be held while clearing the net_conf->discard_my_data flag inside drbd_set_role. It was moved into drbd_adm_set_role with drbd: allow parallel promote/demote actions but then replaced at that location by the newly introduced adm_mutex with drbd: Fix a potential deadlock in drbdsetup, introduce resource->adm_mutex And I simply forgot to put it back in at the original location. Signed-off-by: -
Lars Ellenberg authored
If we re-write all meta data due to resize, we have open-coded write-out of our meta data super block. Stop the md_sync_timer, it would just trigger scary but in this case spurious "timer expired" messages. Signed-off-by:
-
Lars Ellenberg authored
This fixes one recent regresion, and one long existing bug. The bug: drbd_try_clear_on_disk_bm() assumed that all "count" bits have to be accounted in the resync extent corresponding to the start sector. Since we allow application requests to cross our "extent" boundaries, this assumption is no longer true, resulting in possible misaccounting, scary messages ("BAD! sector=12345s enr=6 rs_left=-7 rs_failed=0 count=58 cstate=..."), and potentially, if the last bit to be cleared during resync would reside in previously misaccounted resync extent, the resync would never be recognized as finished, but would be "stalled" forever, even though all blocks are in sync again and all bits have been cleared... The regression was introduced by drbd: get rid of atomic update on disk bitmap works For an "empty" resync (rs_total == 0), we must not "finish" the resync on the SyncSource before the SyncTarget knows all relevant information (sync uuid). We need to wait for the full round-trip, the SyncTarget will then explicitly notify us. Also for normal, non-empty resyncs (rs_total > 0), the resync-finished condition needs to be tested before the schedule() in wait_for_work, or it is likely to be missed. Signed-off-by: -
Lars Ellenberg authored
We may implicitly call drbd_send() from inside wait_for_work(), via maybe_send_barrier(). If the "stop" signal was send just before that, drbd_send() would call flush_signals(), and we would run an unbounded schedule() afterwards. Fix: check for thread_state == RUNNING before we schedule() Signed-off-by:
-
Lars Ellenberg authored
Just trigger the occasional lazy bitmap write-out during resync from the central wait_for_work() helper. Previously, during resync, bitmap pages would be written out separately, synchronously, one at a time, at least 8 times each (every 512 bytes worth of bitmap cleared). Now we trigger "merge friendly" bulk write out of all cleared pages every two seconds during resync, and once the resync is finished. Most pages will be written out only once. Signed-off-by:
-
Lars Ellenberg authored
Previously, once you disabled flushes as a means of enforcing write-ordering, you'd need to detach/re-attach to enable them again. Allow drbdsetup disk-options to re-enable previously disabled write-ordering policy options at runtime. While at it fix RCU in drbd_bump_write_ordering() max_allowed_wo() uses rcu_dereference, therefore it must be called within rcu_read_lock()/rcu_read_unlock() Signed-off-by:
-
Lars Ellenberg authored
Reduce the number of calls to first_peer_device(). Instead, call first_peer_device() just once to assign a local variable peer_device. Signed-off-by:
-
Lars Ellenberg authored
Instead of dropping and re-aquiring the spinlock around the submit, just remember that we want to submit, and do that only once we have dropped the spinlock for good. Signed-off-by:
-
