We don't use or need the pool name or value source in the zvol tasks.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Alexander Motin <mav@FreeBSD.org>
Signed-off-by: Ryan Moeller <ryan@iXsystems.com>
Closes#12361
Most of dsl_dir_diduse_space() and dsl_dir_transfer_space() CPU time
is a dd_lock overhead and time spent in dmu_buf_will_dirty(). Calling
them one after another is a waste of time and even more contention.
Doing that twice for each rewritten block within dbuf_write_done()
via dsl_dataset_block_kill() and dsl_dataset_block_born() created one
of the biggest CPU overheads in case of small blocks rewrite.
dsl_dir_diduse_transfer_space() combines functionality of these two
functions for cases where it is needed, but without double overhead,
practically for the cost of dsl_dir_diduse_space() or even cheaper.
While there, optimize dsl_dir_phys() calls in dsl_dir_diduse_space()
and dsl_dir_transfer_space(). It seems Clang detects some aliasing
there, repeating dd->dd_dbuf->db_data dereference multiple times,
increasing dd_lock scope and contention.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Matthew Ahrens <mahrens@delphix.com>
Author: Ryan Moeller <ryan@iXsystems.com>
Signed-off-by: Alexander Motin <mav@FreeBSD.org>
Sponsored-By: iXsystems, Inc.
Closes#12300
* Tinker with slop space accounting with dedup
Do not include the deduplicated space usage in the slop space
reservation, it leads to surprising outcomes.
* Update spa_dedup_dspace sometimes
Sometimes, we get into spa_get_slop_space() with
spa_dedup_dspace=~0ULL, AKA "unset", while spa_dspace is correctly set.
So call the code to update it before we use it if we hit that case.
Reviewed-by: Matthew Ahrens <mahrens@delphix.com>
Reviewed-by: Mark Maybee <mark.maybee@delphix.com>
Signed-off-by: Rich Ercolani <rincebrain@gmail.com>
Closes#12271
arc_evict_hdr() returns number of evicted bytes in scope of specific
state. For ghost states it does not mean the amount of really freed
memory, but the logical buffer size. It is correct for the eviction
process, but not for waking up threads waiting for ARC size reduction,
as added in "Revise ARC shrinker algorithm" commit, causing premature
wakeups while ARC is still overflowed, allowing even bigger overflow,
plus processing overhead when next allocation will also get blocked,
probably also for too short time.
To fix that make arc_evict_hdr() also return the amount of really
freed memory, which for the ghost states is only the header, and use
it to update arc_evict_count instead. Originally I was thinking to
not return it at all, since arc_get_data_impl() does not account for
the headers, but decided that some slow allocation progress is better
than long waits, reaching on my tests up to 100ms.
To reduce negative latency effects of long time periods when reclaim
thread can free little real memory, start reclamation process earlier,
before we actually reached the overflow threshold, when we have to
throttle new allocations. We can also do it without taking global
arc_evict_lock, reducing the contention.
Reviewed-by: George Wilson <gwilson@delphix.com>
Reviewed-by: Allan Jude <allan@klarasystems.com>
Reviewed-by: Ryan Moeller <ryan@iXsystems.com>
Signed-off-by: Alexander Motin <mav@FreeBSD.org>
Sponsored-By: iXsystems, Inc.
Closes#12279
Callers of zfs_file_get and zfs_file_put can corrupt the reference
counts for the file structure resulting in a panic or a soft lockup.
When zfs send/recv runs, it will add a reference count to the
open file, and begin to send or recv the stream. If the file descriptor
is closed, then when dmu_recv_stream() or dmu_send() return we will
call zfs_file_put to remove the reference we placed on the file
structure. Unfortunately, because zfs_file_put() uses the file
descriptor to lookup the file structure, it may end up finding that
the file descriptor table no longer contains the file struct, thus
leaking the file structure. Or it might end up finding a file
descriptor for a different file and blindly updating its reference
counts. Other failure modes probably exists.
This change reworks the zfs_file_[get|put] interface to not rely
on the file descriptor but instead pass the zfs_file_t pointer around.
Reviewed-by: Matthew Ahrens <mahrens@delphix.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Mark Maybee <mark.maybee@delphix.com>
Reviewed-by: Ryan Moeller <ryan@iXsystems.com>
Co-authored-by: Allan Jude <allan@klarasystems.com>
Signed-off-by: George Wilson <gwilson@delphix.com>
External-issue: DLPX-76119
Closes#12299
This dramatically reduces the lock contention on systems with slower
(non-TSC) timecounters. With TSC the difference is minimal, but since
this lock is pretty congested, any improvement counts. Plus I don't
see any reason to do it under the lock other than the latency of the
lock itself, which this change actually reduces.
Reviewed-by: Matthew Ahrens <mahrens@delphix.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Alexander Motin <mav@FreeBSD.org>
Sponsored-By: iXsystems, Inc.
Closes#12281
With default dbuf cache size of 1/32 of ARC, it makes no sense to have
hash table of the same size (or even bigger on Linux). Reduce it to
1/8 of ARC's one, still leaving some slack, assuming higher I/O rate
via dbuf cache than via ARC.
Remove padding from ARC hash locks array. The idea behind padding
is to avoid false sharing between locks. It would have sense if
there would be a limited number of very busy locks. But since we
have no limit on the number, using the same memory for more locks we
can achieve even lower lock contention with the same false sharing,
or we can use less memory for the same contention level.
Reduce number of hash locks from 8192 to 2048. The number is still
big enough to not cause contention, but reduced memory size improves
cache hit rate for mutex_tryenter() in ARC eviction thread, saving
about 1% of the thread time.
Reviewed-by: Matthew Ahrens <mahrens@delphix.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Alexander Motin <mav@FreeBSD.org>
Sponsored-By: iXsystems, Inc.
Closes#12289
Fix a leak of abd_t that manifested mostly when using
raidzN with at least as many columns as N (e.g. a
four-disk raidz2 but not a three-disk raidz2).
Sufficiently heavy raidz use would eventually run a system
out of memory.
Additionally:
* Switch abd_cache arena to FIRSTFIT, which empirically
improves perofrmance.
* Make abd_chunk_cache more performant and debuggable.
* Allocate the abd_zero_buf from abd_chunk_cache rather
than the heap.
* Don't try to reap non-existent qcaches in abd_cache arena.
* KM_PUSHPAGE->KM_SLEEP when allocating chunks from their
own arena
Reviewed-by: Matthew Ahrens <mahrens@delphix.com>
Reviewed-by: Alexander Motin <mav@FreeBSD.org>
Signed-off-by: Jorgen Lundman <lundman@lundman.net>
Co-authored-by: Sean Doran <smd@use.net>
Closes#12295
dmu_zfetch_stream_fini() is missing calls to destroy the refcounts,
leaking them and the mutex inside.
Reviewed-by: Matthew Ahrens <mahrens@delphix.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Alexander Motin <mav@FreeBSD.org>
Signed-off-by: Jorgen Lundman <lundman@lundman.net>
Closes#12294
Use dp_dirty_pertxg[] for txg_kick(), instead of dp_dirty_total in
original code. Extra parameter "txg" is added for txg_kick(), thus it
knows which txg to kick. Also txg_kick() call is moved from
dsl_pool_need_dirty_delay() to dsl_pool_dirty_space() so that we can
know the txg number assigned for txg_kick().
Some unnecessary code regarding dp_dirty_total in txg_sync_thread() is
also cleaned up.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Matthew Ahrens <mahrens@delphix.com>
Reviewed-by: Alexander Motin <mav@FreeBSD.org>
Signed-off-by: jxdking <lostking2008@hotmail.com>
Closes#12274
The only reason for spa_config_*() to use refcount instead of simple
non-atomic (thanks to scl_lock) variable for scl_count is tracking,
hard disabled for the last 8 years. Switch to simple int scl_count
reduces the lock hold time by avoiding atomic, plus makes structure
fit into single cache line, reducing the locks contention.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Matthew Ahrens <mahrens@delphix.com>
Reviewed-by: Mark Maybee <mark.maybee@delphix.com>
Signed-off-by: Alexander Motin <mav@FreeBSD.org>
Sponsored-By: iXsystems, Inc.
Closes#12287
LLVM's Polly (ISL to be precise) is unhappy with the loop from
ddt_stat_add():
CC [M] fs/zfs/zfs/ddt.o
../lib/External/isl/isl_schedule_node.c:2470: cannot insert node
between set or sequence node and its filter children
(building with the custom patch which adds Polly support to Kbuild)
The mentioned loop is rather suboptimal. All that we need is to just
treat ddt_stat_t as an array of u64 and perform 1:1 addition or
substraction. This can be done in simpler for-loop with the
determined index and bounds. Compiler will expand d_end - d into
a number of ddt_stat_t fields at compile time.
This prevents Polly from failing on this file.
Reviewed-by: Matthew Ahrens <mahrens@delphix.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Alexander Lobakin <alobakin@pm.me>
Closes#12253
The number of sublists in a multilist is relatively small. We dont need
64 bits to calculate an index. 32 bits is sufficient and makes the
code more efficient.
Reviewed-by: Matthew Ahrens <mahrens@delphix.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Mark Maybee <mark.maybee@delphix.com>
Signed-off-by: Alexander Motin <mav@FreeBSD.org>
Sponsored-By: iXsystems, Inc.
Closes#12288
While abd_verify() does nothing when built without debug, compiler
can't optimize it out by itself due to calls to external list_*()
and abd_verify_scatter(). This commit makes it explicit.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Adam Moss <c@yotes.com>
Reviewed-by: George Melikov <mail@gmelikov.ru>
Signed-off-by: Alexander Motin <mav@FreeBSD.org>
Sponsored-By: iXsystems, Inc.
Closes#12280
Unlike most other properties the 'compatibility' property is stored
in the pool config object and not the DMU_OT_POOL_PROPS object.
This had the advantage that the compatibility information is available
without needing to fully import the pool (it can be read with zdb).
However, this means we need to make sure to update both the copy of
the config in the MOS and the cache file. This wasn't being done.
This commit adds a call to spa_async_request() to ensure the copy of
the config in the cache file gets updated as well as the one stored
in the pool. This same change is made for the 'comment' property
which suffers from the same inconsistency.
Reviewed-by: Sean Eric Fagan <sef@ixsystems.com>
Reviewed-by: Colm Buckley <colm@tuatha.org>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes#12261Closes#12276
According to current zfs man page zfs_metaslab_mem_limit should be
25 instead of 75.
Reviewed-by: Matthew Ahrens <mahrens@delphix.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Paul Dagnelie <pcd@delphix.com>
Reviewed-by: Mark Maybee <mark.maybee@delphix.com>
Signed-off-by: jumbi77@users.noreply.github.comCloses#12273
ZFS loves using %llu for uint64_t, but that requires a cast to not
be noisy - which is even done in many, though not all, places.
Also a couple places used %u for uint64_t, which were promoted
to %llu.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Rich Ercolani <rincebrain@gmail.com>
Closes#12233
This reverts commit 13fac09868.
Per the discussion in #11531, the reverted commit---which intended only
to be a cleanup commit---introduced a subtle, unintended change in
behavior.
Care was taken to partially revert and then reapply 10b3c7f5e4
which would otherwise have caused a conflict. These changes were
squashed in to this commit.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Suggested-by: @chrisrd
Suggested-by: robn@despairlabs.com
Signed-off-by: Antonio Russo <aerusso@aerusso.net>
Closes#11531Closes#12227
In all places except two spa_get_random() is used for small values,
and the consumers do not require well seeded high quality values.
Switch those two exceptions directly to random_get_pseudo_bytes()
and optimize spa_get_random(), renaming it to random_in_range(),
since it is not related to SPA or ZFS in general.
On FreeBSD directly map random_in_range() to new prng32_bounded() KPI
added in FreeBSD 13. On Linux and in user-space just reduce the type
used to uint32_t to avoid more expensive 64bit division.
Reviewed-by: Ryan Moeller <ryan@iXsystems.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Alexander Motin <mav@FreeBSD.org>
Sponsored-By: iXsystems, Inc.
Closes#12183
wmsum was designed exactly for cases like these with many updates
and rare reads. It allows to completely avoid atomic operations on
congested global variables.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Mark Maybee <mark.maybee@delphix.com>
Signed-off-by: Alexander Motin <mav@FreeBSD.org>
Sponsored-By: iXsystems, Inc.
Closes#12172
In case we have I/O and try to remove an L2ARC device a deadlock might
occur. arc_read()->zio_read()->zfs_blkptr_verify() waits for SCL_VDEV
to be dropped while holding the hash_lock. However, spa_l2cache_load()
holds SCL_ALL and waits for the hash_lock in l2arc_evict().
Fix this by moving zfs_blkptr_verify() to the top top arc_read() before
the hash_lock is taken. Verify the block pointer and return a checksum
error if damaged rather than halting the system, by using
BLK_VERIFY_LOG instead of BLK_VERIFY_HALT.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Mark Maybee <mark.maybee@delphix.com>
Signed-off-by: George Amanakis <gamanakis@gmail.com>
Closes#12054
vdev_draid_min_asize() returns the minimum size of a child vdev. This
is used when determining if a disk is big enough to replace a child.
It's also used by zdb to determine how big of a child to make to test
replacement.
vdev_draid_min_asize() says that the child’s asize has to be at least
1/Nth of the entire draid’s asize, which is the same logic as raidz.
However, this contradicts the code in vdev_draid_open(), which
calculates the draid’s asize based on a reduced child size:
An additional 32MB of scratch space is reserved at the end of each
child for use by the dRAID expansion feature
So the problem is that you can replace a draid disk with one that’s
vdev_draid_min_asize(), but it actually needs to be larger to accommodate
the additional 32MB. The replacement is allowed and everything works at
first (since the reserved space is at the end, and we don’t try to use
it yet), but when you try to close and reopen the pool,
vdev_draid_open() calculates a smaller asize for the draid, because of
the smaller leaf, which is not allowed.
I think the confusion is that vdev_draid_min_asize() is correctly
returning the amount of required *allocatable* space in a leaf, but the
actual *size* of the leaf needs to be at least 32MB more than that.
ztest_vdev_attach_detach() assumes that it can attach that size of
device, and it actually can (the kernel/libzpool accepts it), but it
then later causes zdb to not be able to open the pool.
This commit changes vdev_draid_min_asize() to return the required size
of the leaf, not the size that draid will make available to the metaslab
allocator.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Mark Maybee <mark.maybee@delphix.com>
Signed-off-by: Matthew Ahrens <mahrens@delphix.com>
Closes#11459Closes#12221
This commit partially reverts changes to multilists in PR 7968
(multi-threaded spa-sync()) and adds some cache line alignments to
separate read-only multilists and heavily modified refcount's to different
cache lines.
Reviewed-by: Matthew Ahrens <mahrens@delphix.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Alexander Motin <mav@FreeBSD.org>
Sponsored-by: iXsystems, Inc.
Closes#12158
This mostly reverts "3537 want pool io kstats" commit of 8 years ago.
From one side this code using pool-wide locks became pretty bad for
performance, creating significant lock contention in I/O pipeline.
From another, there are more efficient ways now to obtain detailed
statistics, while this statistics is illumos-specific and much less
usable on Linux and FreeBSD, reported only via procfs/sysctls.
This commit does not remove KSTAT_TYPE_IO implementation, that may
be removed later together with already unused KSTAT_TYPE_INTR and
KSTAT_TYPE_TIMER.
Reviewed-by: Matthew Ahrens <mahrens@delphix.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Alexander Motin <mav@FreeBSD.org>
Sponsored-By: iXsystems, Inc.
Closes#12212
`getfsstat(2)` is used to retrieve the list of mounted file systems,
which libzfs uses when fetching properties like mountpoint, atime,
setuid, etc. The `mode` parameter may be `MNT_NOWAIT`, which uses
information in the VFS's cache, or `MNT_WAIT`, which effectively does a
`statfs` on every single mounted file system in order to fetch the most
up-to-date information. As far as I can tell, the only fields that
libzfs cares about are the filesystem's name, mountpoint, fstypename,
and mount flags. Those things are always updated on mount and unmount,
so they will always be accurate in the VFS's mount cache except in two
circumstances:
1) When a file system is busy unmounting
2) When a ZFS file system changes the value of a mount-overridable
property like atime or setuid, but doesn't remount the file system.
Right now that only happens when the property is changed by an
unprivileged user who has delegated authority to change the property
but not to mount the dataset. But perhaps libzfs could choose to do
it for other reasons in the future.
Switching to `MNT_NOWAIT` will greatly improve speed with no downside,
as long as we explicitly update the mount cache whenever we change a
mount-overridable property.
For comparison, Illumos gets this information using the native
`getmntany` and `getmntent` functions, which also use cached
information. The illumos function that would refresh the cache,
`resetmnttab`, is never called by libzfs.
And on GNU/Linux, `getmntany` and `getmntent` don't even communicate
with the kernel directly. They simply parse the file they are given,
which is usually /etc/mtab or /proc/mounts. Perhaps the implementation
of /proc/mounts is synchronous, ala MNT_WAIT; I don't know.
Sponsored-by: Axcient
Reviewed-by: Ryan Moeller <ryan@iXsystems.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Alan Somers <asomers@gmail.com>
Closes: #12091
Update the logic to handle the dedup-case of consecutive
FREEs in the livelist code. The logic still ensures that
all the FREE entries are matched up with a respective
ALLOC by keeping a refcount for each FREE blkptr that we
encounter and ensuring that this refcount gets to zero
by the time we are done processing the livelist.
zdb -y no longer panics when encountering double frees
Reviewed-by: Matthew Ahrens <mahrens@delphix.com>
Reviewed-by: John Kennedy <john.kennedy@delphix.com>
Reviewed-by: Don Brady <don.brady@delphix.com>
Signed-off-by: Serapheim Dimitropoulos <serapheim@delphix.com>
Closes#11480Closes#12177
- Avoid atomic_add() when updating as_lower_bound/as_upper_bound.
Previous code was excessively strong on 64bit systems while not
strong enough on 32bit ones. Instead introduce and use real
atomic_load() and atomic_store() operations, just an assignments
on 64bit machines, but using proper atomics on 32bit ones to avoid
torn reads/writes.
- Reduce number of buckets on large systems. Extra buckets not as
much improve add speed, as hurt reads. Unlike wmsum for aggsum
reads are still important.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Ryan Moeller <ryan@iXsystems.com>
Signed-off-by: Alexander Motin <mav@FreeBSD.org>
Sponsored-By: iXsystems, Inc.
Closes#12145
wmsum counters are a reduced version of aggsum counters, optimized for
write-mostly scenarios. They do not provide optimized read functions,
but instead allow much cheaper add function. The primary usage is
infrequently read statistic counters, not requiring exact precision.
The Linux implementation is directly mapped into percpu_counter KPI.
The FreeBSD implementation is directly mapped into counter(9) KPI.
In user-space due to lack of better implementation mapped to aggsum.
Unfortunately neither Linux percpu_counter nor FreeBSD counter(9)
provide sufficient functionality to completelly replace aggsum, so
it still remains to be used for several hot counters.
Reviewed-by: Paul Dagnelie <pcd@delphix.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Ryan Moeller <ryan@iXsystems.com>
Signed-off-by: Alexander Motin <mav@FreeBSD.org>
Sponsored-By: iXsystems, Inc.
Closes#12114
Previously, ZFS scaled maxinflight_bytes based on total number of
disks in the pool. A 3-wide mirror was receiving a queue depth of 3
disks, which it should not, since it reads from all the disks inside.
For wide raidz the situation was slightly better, but still a 3-wide
raidz1 received a depth of 3 disks instead of 2.
The new code counts only unique data disks, i.e. 1 disk for mirrors
and non-parity disks for raidz/draid. For draid the math is still
imperfect, since vdev_get_nparity() returns number of parity disks
per group, not per vdev, but still some better than it was.
This should slightly reduce scrub influence on payload for some pool
topologies by avoiding excessive queuing.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Ryan Moeller <ryan@iXsystems.com>
Signed-off-by: Alexander Motin <mav@FreeBSD.org>
Sponsored-By: iXsystems, Inc.
Closing #12046
Propagate vdev child state to parents on invalid label
Add VDEV_AUX_BAD_LABEL to print_import_config()
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Mark Maybee <mark.maybee@delphix.com>
Co-authored-by: Srikanth N S <srikanth.nagasubbaraoseetharaman@hpe.com>
Signed-off-by: Vipin Kumar Verma <vipin.verma@hpe.com>
Closes#12088
This change addresses two distinct scenarios which are possible
when performing a sequential resilver to a dRAID pool with vdevs
that contain silent unknown damage. Which in this circumstance
took the form of the devices being intentionally overwritten with
zeros. However, it could also result from a device returning incorrect
data while a sequential resilver was in progress.
Scenario 1) A sequential resilver is performed while all of the
dRAID vdevs are ONLINE and there is silent damage present on the
vdev being resilvered. In this case, nothing will be repaired
by vdev_raidz_io_done_reconstruct_known_missing() because
rc->rc_error isn't set on any of the raid columns. To address
this vdev_draid_io_start_read() has been updated to always mark
the resilvering column as ESTALE for sequential resilver IO.
Scenario 2) Multiple columns contain silent damage for the same
block and a sequential resilver is performed. In this case it's
impossible to generate the correct data from parity unless all of
the damaged columns are being sequentially resilvered (and thus
only good data is used to generate parity). This is as expected
and there's nothing which can be done about it. However, we need
to be careful not to make to situation worse. Since we can't
verify the data is actually good without a checksum, we must
only repair the devices which are being sequentially resilvered.
Otherwise, an incorrect repair to a device which previously
contained good data could effectively lock in the damage and
make reconstruction impossible. A check for this was added to
vdev_raidz_io_done_verified() along with a new test case.
Lastly, this change updates the redundancy_draid_spare1 and
redundancy_draid_spare3 test cases to be more representative
of normal dRAID replacement operation. Specifically, what we
care about is that the scrub run after a sequential resilver
does not find additional blocks which need repair. This would
indicate the sequential resilver failed to rebuild a section of
one of the devices. Note also the tests were switched to using
the verify_pool() function which still checks for checksum errors.
Reviewed-by: Mark Maybee <mark.maybee@delphix.com>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes#12061
While use of dynamic taskqs allows to reduce number of idle threads,
hardcoded 8 taskqs of each kind is a big overkill for small systems,
complicating CPU scheduling, increasing I/O reorder, etc, while
providing no real locking benefits, just not needed there.
On another side, 12*8 worker threads per kind are able to overload
almost any system nowadays. For example, pool of several fast SSDs
with SHA256 checksum makes system barely responsive during scrub, or
with dedup enabled barely responsive during large file deletion.
To address both problems this patch introduces ZTI_SCALE macro, alike
to ZTI_BATCH, but with multiple taskqs, depending on number of CPUs,
to be used in places where lock scalability is needed, while request
ordering is not so much. The code is made to create new taskq for
~6 worker threads (less for small systems, but more for very large)
up to 80% of CPU cores (previous 75% was not good for rounding down).
Both number of threads and threads per taskq are now tunable in case
somebody really wants to use all of system power for ZFS.
While obviously some benchmarks show small peak performance reduction
(not so big really, especially on systems with SMT, where use of the
second threads does not give as much performance as the first ones),
they also show dramatic latency reduction and much more smooth user-
space operation in case of high CPU usage by ZFS.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Alexander Motin <mav@FreeBSD.org>
Sponsored-By: iXsystems, Inc.
Closes#11966
Use dsl_dataset_has_resume_receive_state()
not dsl_dataset_is_zapified() to check if
stream is resumable.
Reviewed-by: Matthew Ahrens <mahrens@delphix.com>
Reviewed-by: Alek Pinchuk <apinchuk@axcient.com>
Reviewed-by: Ryan Moeller <ryan@ixsystems.com>
Signed-off-by: Paul Zuchowski <pzuchowski@datto.com>
Closes#12034
Commit d1d4769 takes into account the encryption key version to
decide if the local_mac could be zeroed out. However, this could lead
to failure mounting encrypted datasets created with intermediate
versions of ZFS encryption available in master between major releases.
In order to prevent this situation revert d1d4769 pending a more
comprehensive fix which addresses the mount failure case.
Reviewed-by: George Amanakis <gamanakis@gmail.com>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Issue #11294
Issue #12025
Issue #12300Closes#12033
zfs_zevent_console committed multiple printk()s per line without
properly continuing them ‒ a single event could easily be fragmented
across over thirty lines, making it useless for direct application
zfs_zevent_cols exists purely to wrap the output from zfs_zevent_console
The niche this was supposed to fill can be better served by something
akin to the all-syslog ZEDLET
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Ahelenia Ziemiańska <nabijaczleweli@nabijaczleweli.xyz>
Closes#7082Closes#11996
When dRAID performs a normal read operation only the data columns
in the raid map are read from disk. This is enough information to
calculate the checksum, verify it, and return the needed data to the
application. It's only in the event of a checksum failure that the
additional parity and any empty columns must be read since they are
required for parity reconstruction.
Reading these additional columns is handled by vdev_raidz_read_all()
which calls vdev_draid_map_alloc_empty() to expand the raid_map_t
and submit IOs for the missing columns. This all works correctly,
but it fails to account for any "short" columns. These are data
columns which are padded with a empty skip sector at the end.
Since that empty sector is not needed for a normal read it's not
read when columns is first read from disk. However, like the parity
and empty columns the skip sector is needed to perform reconstruction.
The fix is to mark any "short" columns as never being read by clearing
the rc_tried flag when expanding the raid_map_t. This will cause
the entire column to re-read from disk in the event of a checksum
failure allowing the self-healing functionality to repair the block.
Note that this only effects the self-healing feature because when
scrubbing a pool the parity, data, and empty columns are all read
initially to verify their contents. Furthermore, only blocks which
contain "short" columns would be effected, and only when the memory
backing the skip sector wasn't already zeroed out.
This change extends the existing redundancy_raidz.ksh test case to
verify self-healing (as well as resilver and scrub). Then applies
the same test case to dRAID with a slightly modified version of
the test script called redundancy_draid.ksh. The unused variable
combrec was also removed from both test cases.
Reviewed-by: Matthew Ahrens <mahrens@delphix.com>
Reviewed-by: Mark Maybee <mark.maybee@delphix.com>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes#12010
Previous code tried to keep prefetch streams while moving dnode. But
it was at least not updating per-stream zs_fetchback pointers, causing
use-after-free on next access. Instead of that I see much easier and
cleaner to just drop old prefetch state and start new from scratch.
Reviewed-by: Matthew Ahrens <mahrens@delphix.com>
Reviewed-by: Igor Kozhukhov <igor@dilos.org>
Signed-off-by: Alexander Motin <mav@FreeBSD.org>
Sponsored-By: iXsystems, Inc.
Closes#11936Closes#11998
This ensures that we don't accumulate checksum errors against offline or
unavailable devices but, more importantly, means that we don't
needlessly create DTL entries for offline devices that are already
up-to-date.
Consider a 3-way mirror, with disk A always online (and so always with
an empty DTL) and B and C only occasionally online. When A & B resilver
with C offline, B's DTL will effectively be appended to C's due to these
spurious ZIOs even as the resilver empties B's DTL:
* These ZIOs land in vdev_mirror_scrub_done() and flag an error
* That flagged error causes vdev_mirror_io_done() to see
unexpected_errors, so it issues a ZIO_TYPE_WRITE repair ZIO, which
inherits ZIO_FLAG_SCAN_THREAD because zio_vdev_child_io() includes
that flag in ZIO_VDEV_CHILD_FLAGS.
* That ZIO fails, too, and eventually zio_done() gets its hands on it
and calls vdev_stat_update().
* vdev_stat_update() sees the error and this zio...
* is not speculative,
* is not due to EIO (but rather ENXIO, since the device is closed)
* has an ->io_vd != NULL (specifically, the offline leaf device)
* is a write
* is for a txg != 0 (but rather the read block's physical birth txg)
* has ZIO_FLAG_SCAN_THREAD asserted
* So: vdev_stat_update() calls vdev_dtl_dirty() on the offline vdev.
Then, when A & C resilver with B offline, that story gets replayed and
C's DTL will be appended to B's.
In fact, one does not need this permanently-broken-mirror scenario to
induce badness: breaking a mirror with no DTLs and then scrubbing will
create DTLs for all offline devices. These DTLs will persist until the
entire mirror is reassembled for the duration of the *resilver*, which,
incidentally, will not consider the devices with good data to be sources
of good data in the case of a read failure.
Reviewed-by: Mark Maybee <mark.maybee@delphix.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Nathaniel Wesley Filardo <nwfilardo@gmail.com>
Closes#11930
This deduplicates 2 sets of caches which use the same allocation size.
Memory savings fluctuate a lot, one sample result is FreeBSD running
"make buildworld" saving ~180MB RAM in reduced page count associated
with zio caches.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Mateusz Guzik <mjguzik@gmail.com>
Closes#11877
Fix NULL pointer dereference when reporting
checksum error for gang block in zio_done.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Paul Zuchowski <pzuchowski@datto.com>
Closes#11872Closes#11896
Traversal code, traverse_visitbp() does visit blocks recursively.
Indirect (Non L0) Block of size 128k could contain, 1024 block pointers
of 128 bytes. In case of full traverse OR incremental traverse, where
all blocks were modified, it could traverse large number of blocks
pointed by indirect. Traversal code does issue prefetch of blocks
traversed below indirect. This could result into large number of
async reads queued on vdev queue. So, account for prefetch issued for
blocks pointed by indirect and limit max prefetch in one go.
Module Param:
zfs_traverse_indirect_prefetch_limit: Limit of prefetch while traversing
an indirect block.
Local counters:
prefetched: Local counter to account for number prefetch done.
pidx: Index for which next prefetch to be issued.
ptidx: Index at which next prefetch to be triggered.
Keep "ptidx" somewhere in the middle of blocks prefetched, so that
blocks prefetch read gets the enough time window before their demand
read is issued.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Alexander Motin <mav@FreeBSD.org>
Signed-off-by: Jitendra Patidar <jitendra.patidar@nutanix.com>
Closes#11802Closes#11803
It's been observed in the CI that the required 25% of obsolete bytes
in the mapping can be to high a threshold for this test resulting in
condensing never being triggered and a test failure. To prevent these
failures make the existing zfs_condense_indirect_obsolete_pct tuning
available so the obsolete percentage can be reduced from 25% to 5%
during this test.
Reviewed-by: Ryan Moeller <ryan@iXsystems.com>
Reviewed-by: George Melikov <mail@gmelikov.ru>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes#11869
After 3937ab20f zfsdev_get_state_impl can become zfsdev_get_state.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Ryan Moeller <freqlabs@FreeBSD.org>
Closes#11833
