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
Compiling with gcc 11.1.0 produces three new warnings.
Change the code slightly to avoid them.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Matthew Ahrens <mahrens@delphix.com>
Signed-off-by: Attila Fülöp <attila@fueloep.org>
Closes#12130Closes#12188Closes#12237
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
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
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
- 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
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
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
As found by
git grep -E '(open|setmntent|pipe2?)\(' |
grep -vE '((zfs|zpool)_|fd|dl|lzc_re|pidfile_|g_)open\('
FreeBSD's pidfile_open() says nothing about the flags of the files it
opens, but we can't do anything about it anyway; the implementation does
open all files with O_CLOEXEC
Consider this output with zpool.d/media appended with
"pid=$$; (ls -l /proc/$pid/fd > /dev/tty)":
$ /sbin/zpool iostat -vc media
lrwx------ 0 -> /dev/pts/0
l-wx------ 1 -> 'pipe:[3278500]'
l-wx------ 2 -> /dev/null
lrwx------ 3 -> /dev/zfs
lr-x------ 4 -> /proc/31895/mounts
lrwx------ 5 -> /dev/zfs
lr-x------ 10 -> /usr/lib/zfs-linux/zpool.d/media
vs
$ ./zpool iostat -vc vendor,upath,iostat,media
lrwx------ 0 -> /dev/pts/0
l-wx------ 1 -> 'pipe:[3279887]'
l-wx------ 2 -> /dev/null
lr-x------ 10 -> /usr/lib/zfs-linux/zpool.d/media
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Ahelenia Ziemiańska <nabijaczleweli@nabijaczleweli.xyz>
Closes#11866
When a rebuild completes it will automatically schedule a follow up
scrub to verify all of the block checksums. Before setting up the
scrub execute the counterpart dsl_scan_setup_check() function to
confirm the scrub can be started. Prior to this change we'd only
check vdev_rebuild_active() which isn't as comprehensive, and using
the check function keeps all of this logic in one place.
Reviewed-by: Mark Maybee <mark.maybee@delphix.com>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes#11849
Just as delay zevents can flood the zevent pipe when a vdev becomes
unresponsive, so do the deadman zevents.
Ratelimit deadman zevents according to the same tunable as for delay
zevents.
Enable deadman tests on FreeBSD and add a test for deadman event
ratelimiting.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Don Brady <don.brady@delphix.com>
Signed-off-by: Ryan Moeller <ryan@iXsystems.com>
Closes#11786
Correct an assortment of typos throughout the code base.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Matthew Ahrens <mahrens@delphix.com>
Reviewed-by: Ryan Moeller <ryan@iXsystems.com>
Signed-off-by: Andrea Gelmini <andrea.gelmini@gelma.net>
Closes#11774
For gang blocks, `DVA_GET_ASIZE()` is the total space allocated for the
gang DVA including its children BP's. The space allocated at each DVA's
vdev/offset is `vdev_psize_to_asize(vd, SPA_GANGBLOCKSIZE)`.
This commit makes this relationship more clear by using a helper
function, `vdev_gang_header_asize()`, for the space allocated at the
gang block's vdev/offset.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Matthew Ahrens <mahrens@delphix.com>
Closes#11744
Reviewed-by: Matthew Ahrens <mahrens@delphix.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Andrea Gelmini <andrea.gelmini@gelma.net>
Closes#11775
To make better predictions on parallel workloads dmu_zfetch() should
be called as early as possible to reduce possible request reordering.
In particular, it should be called before dmu_buf_hold_array_by_dnode()
calls dbuf_hold(), which may sleep waiting for indirect blocks, waking
up multiple threads same time on completion, that can significantly
reorder the requests, making the stream look like random. But we
should not issue prefetch requests before the on-demand ones, since
they may get to the disks first despite the I/O scheduler, increasing
on-demand request latency.
This patch splits dmu_zfetch() into two functions: dmu_zfetch_prepare()
and dmu_zfetch_run(). The first can be executed as early as needed.
It only updates statistics and makes predictions without issuing any
I/Os. The I/O issuance is handled by dmu_zfetch_run(), which can be
called later when all on-demand I/Os are already issued. It even
tracks the activity of other concurrent threads, issuing the prefetch
only when _all_ on-demand requests are issued.
For many years it was a big problem for storage servers, handling
deeper request queues from their clients, having to either serialize
consequential reads to make ZFS prefetcher usable, or execute the
incoming requests as-is and get almost no prefetch from ZFS, relying
only on deep enough prefetch by the clients. Benefits of those ways
varied, but neither was perfect. With this patch deeper queue
sequential read benchmarks with CrystalDiskMark from Windows via
iSCSI to FreeBSD target show me much better throughput with almost
100% prefetcher hit rate, comparing to almost zero before.
While there, I also removed per-stream zs_lock as useless, completely
covered by parent zf_lock. Also I reused zs_blocks refcount to track
zf_stream linkage of the stream, since I believe previous zs_fetch ==
NULL check in dmu_zfetch_stream_done() was racy.
Delete prefetch streams when they reach ends of files. It saves up
to 1KB of RAM per file, plus reduces searches through the stream list.
Block data prefetch (speculation and indirect block prefetch is still
done since they are cheaper) if all dbufs of the stream are already
in DMU cache. First cache miss immediately fires all the prefetch
that would be done for the stream by that time. It saves some CPU
time if same files within DMU cache capacity are read over and over.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Adam Moss <c@yotes.com>
Reviewed-by: Matthew Ahrens <mahrens@delphix.com>
Signed-off-by: Alexander Motin <mav@FreeBSD.org>
Sponsored-By: iXsystems, Inc.
Closes#11652
If TX_WRITE is create on a file, and the file is later deleted and a new
directory is created on the same object id, it is possible that when
zil_commit happens, zfs_get_data will be called on the new directory.
This may result in panic as it tries to do range lock.
This patch fixes this issue by record the generation number during
zfs_log_write, so zfs_get_data can check if the object is valid.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Chunwei Chen <david.chen@nutanix.com>
Closes#10593Closes#11682
The RAIDZ and DRAID code is responsible for reporting checksum errors on
their child vdevs. Checksum errors represent events where a disk
returned data or parity that should have been correct, but was not. In
other words, these are instances of silent data corruption. The
checksum errors show up in the vdev stats (and thus `zpool status`'s
CKSUM column), and in the event log (`zpool events`).
Note, this is in contrast with the more common "noisy" errors where a
disk goes offline, in which case ZFS knows that the disk is bad and
doesn't try to read it, or the device returns an error on the requested
read or write operation.
RAIDZ/DRAID generate checksum errors via three code paths:
1. When RAIDZ/DRAID reconstructs a damaged block, checksum errors are
reported on any children whose data was not used during the
reconstruction. This is handled in `raidz_reconstruct()`. This is the
most common type of RAIDZ/DRAID checksum error.
2. When RAIDZ/DRAID is not able to reconstruct a damaged block, that
means that the data has been lost. The zio fails and an error is
returned to the consumer (e.g. the read(2) system call). This would
happen if, for example, three different disks in a RAIDZ2 group are
silently damaged. Since the damage is silent, it isn't possible to know
which three disks are damaged, so a checksum error is reported against
every child that returned data or parity for this read. (For DRAID,
typically only one "group" of children is involved in each io.) This
case is handled in `vdev_raidz_cksum_finish()`. This is the next most
common type of RAIDZ/DRAID checksum error.
3. If RAIDZ/DRAID is not able to reconstruct a damaged block (like in
case 2), but there happens to be additional copies of this block due to
"ditto blocks" (i.e. multiple DVA's in this blkptr_t), and one of those
copies is good, then RAIDZ/DRAID compares each sector of the data or
parity that it retrieved with the good data from the other DVA, and if
they differ then it reports a checksum error on this child. This
differs from case 2 in that the checksum error is reported on only the
subset of children that actually have bad data or parity. This case
happens very rarely, since normally only metadata has ditto blocks. If
the silent damage is extensive, there will be many instances of case 2,
and the pool will likely be unrecoverable.
The code for handling case 3 is considerably more complicated than the
other cases, for two reasons:
1. It needs to run after the main raidz read logic has completed. The
data RAIDZ read needs to be preserved until after the alternate DVA has
been read, which necessitates refcounts and callbacks managed by the
non-raidz-specific zio layer.
2. It's nontrivial to map the sections of data read by RAIDZ to the
correct data. For example, the correct data does not include the parity
information, so the parity must be recalculated based on the correct
data, and then compared to the parity that was read from the RAIDZ
children.
Due to the complexity of case 3, the rareness of hitting it, and the
minimal benefit it provides above case 2, this commit removes the code
for case 3. These types of errors will now be handled the same as case
2, i.e. the checksum error will be reported against all children that
returned data or parity.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Matthew Ahrens <mahrens@delphix.com>
Closes#11735
The `rr_code` field in `raidz_row_t` is unused.
This commit removes the field, as well as the code that's used to set
it.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Matthew Ahrens <mahrens@delphix.com>
Closes#11736
Resolve some oddities in zfsdev_close() which could result in a
panic and were not present in the equivalent function for Linux.
- Remove unused definition ZFS_MIN_MINOR
- FreeBSD: Simplify zfsdev state destruction
- Assert zs_minor is valid in zfsdev_close
- Make locking around zfsdev state match Linux
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Alexander Motin <mav@FreeBSD.org>
Signed-off-by: Ryan Moeller <ryan@iXsystems.com>
Closes#11720
For some reason cppcheck 1.90 is generating an invalidSyntax warning
when the BF64_SET macro is used in the zstream source. The same
warning is not reported by cppcheck 2.3, nor is their any evident
problem with the expanded macro. This appears to be an issue with
this version of cppcheck. This commit annotates the source to suppress
the warning.
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes#11700
This change modifies the behavior of how we determine how much slop
space to use in the pool, such that now it has an upper limit. The
default upper limit is 128G, but is configurable via a tunable.
Reviewed-by: Matthew Ahrens <mahrens@delphix.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Prakash Surya <prakash.surya@delphix.com>
Closes#11023
Making uio_impl.h the common header interface between Linux and FreeBSD
so both OS's can share a common header file. This also helps reduce code
duplication for zfs_uio_t for each OS.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Brian Atkinson <batkinson@lanl.gov>
Closes#11622
Also fixes leak of the dlopen handle in the error case.
Reviewed-by: Matthew Ahrens <mahrens@delphix.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Pavel Zakharov <pavel.zakharov@delphix.com>
Signed-off-by: Christian Schwarz <me@cschwarz.com>
Closes#11602
Fix regression seen in issue #11545 where checksum errors
where not being counted or showing up in a zpool event.
Reviewed-by: Matthew Ahrens <mahrens@delphix.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Don Brady <don.brady@delphix.com>
Closes#11609
Property to allow sets of features to be specified; for compatibility
with specific versions / releases / external systems. Influences
the behavior of 'zpool upgrade' and 'zpool create'. Initial man
page changes and test cases included.
Brief synopsis:
zpool create -o compatibility=off|legacy|file[,file...] pool vdev...
compatibility = off : disable compatibility mode (enable all features)
compatibility = legacy : request that no features be enabled
compatibility = file[,file...] : read features from specified files.
Only features present in *all* files will be enabled on the
resulting pool. Filenames may be absolute, or relative to
/etc/zfs/compatibility.d or /usr/share/zfs/compatibility.d (/etc
checked first).
Only affects zpool create, zpool upgrade and zpool status.
ABI changes in libzfs:
* New function "zpool_load_compat" to load and parse compat sets.
* Add "zpool_compat_status_t" typedef for compatibility parse status.
* Add ZPOOL_PROP_COMPATIBILITY to the pool properties enum
* Add ZPOOL_STATUS_COMPATIBILITY_ERR to the pool status enum
An initial set of base compatibility sets are included in
cmd/zpool/compatibility.d, and the Makefile for cmd/zpool is
modified to install these in $pkgdatadir/compatibility.d and to
create symbolic links to a reasonable set of aliases.
Reviewed-by: ericloewe
Reviewed-by: Matthew Ahrens <mahrens@delphix.com>
Reviewed-by: Richard Laager <rlaager@wiktel.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Colm Buckley <colm@tuatha.org>
Closes#11468
FreeBSD's zfsd fails to build after e2af2acce3 due to strict type
checking errors from the implicit conversion between bool and boolean_t
in the inline predicate definitions in abd.h.
Use conditionals to return the correct value type from these functions.
Reviewed-by: Matthew Ahrens <mahrens@delphix.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Igor Kozhukhov <igor@dilos.org>
Signed-off-by: Ryan Moeller <freqlabs@FreeBSD.org>
Closes#11592
zfs_znode_update_vfs is a more platform-agnostic name than
zfs_inode_update. Besides that, the function's prototype is moved to
include/sys/zfs_znode.h as the function is also used in common code.
Reviewed-by: Ryan Moeller <ryan@ixsystems.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Ka Ho Ng <khng300@gmail.com>
Sponsored by: The FreeBSD Foundation
Closes#11580
ABD's currently track their parent/child relationship. This applies to
`abd_get_offset()` and `abd_borrow_buf()`. However, nothing depends on
knowing this relationship, it's only used for consistency checks to
verify that we are not destroying an ABD that's still in use. When we
are creating/destroying ABD's frequently, the performance impact of
maintaining these data structures (in particular the atomic
increment/decrement operations) can be measurable.
This commit removes this verification code on production builds, but
keeps it when ZFS_DEBUG is set.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Brian Atkinson <batkinson@lanl.gov>
Signed-off-by: Matthew Ahrens <mahrens@delphix.com>
Closes#11535
The runtime of vdev_validate is dominated by the disk accesses in
vdev_label_read_config. Speed it up by validating all vdevs in
parallel using a taskq.
Sponsored by: Axcient
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Alan Somers <asomers@gmail.com>
Closes#11470
metaslab_init is the slowest part of importing a mature pool, and it
must be repeated hundreds of times for each top-level vdev. But its
speed is dominated by a few serialized disk accesses. That can lead to
import times of > 1 hour for pools with many top-level vdevs on spinny
disks.
Speed up the import by using a taskqueue to parallelize vdev_load across
all top-level vdevs.
This also requires adding mutex protection to
metaslab_class_t.mc_historgram. The mc_histogram fields were
unprotected when that code was first written in "Illumos 4976-4984 -
metaslab improvements" (OpenZFS
f3a7f6610f). The lock wasn't added until
3dfb57a35e, though it's unclear exactly
which fields it's supposed to protect. In any case, it wasn't until
vdev_load was parallelized that any code attempted concurrent access to
those fields.
Sponsored by: Axcient
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Alan Somers <asomers@gmail.com>
Closes#11470
Mixing ZIL and normal allocations has several problems:
1. The ZIL allocations are allocated, written to disk, and then a few
seconds later freed. This leaves behind holes (free segments) where the
ZIL blocks used to be, which increases fragmentation, which negatively
impacts performance.
2. When under moderate load, ZIL allocations are of 128KB. If the pool
is fairly fragmented, there may not be many free chunks of that size.
This causes ZFS to load more metaslabs to locate free segments of 128KB
or more. The loading happens synchronously (from zil_commit()), and can
take around a second even if the metaslab's spacemap is cached in the
ARC. All concurrent synchronous operations on this filesystem must wait
while the metaslab is loading. This can cause a significant performance
impact.
3. If the pool is very fragmented, there may be zero free chunks of
128KB or more. In this case, the ZIL falls back to txg_wait_synced(),
which has an enormous performance impact.
These problems can be eliminated by using a dedicated log device
("slog"), even one with the same performance characteristics as the
normal devices.
This change sets aside one metaslab from each top-level vdev that is
preferentially used for ZIL allocations (vdev_log_mg,
spa_embedded_log_class). From an allocation perspective, this is
similar to having a dedicated log device, and it eliminates the
above-mentioned performance problems.
Log (ZIL) blocks can be allocated from the following locations. Each
one is tried in order until the allocation succeeds:
1. dedicated log vdevs, aka "slog" (spa_log_class)
2. embedded slog metaslabs (spa_embedded_log_class)
3. other metaslabs in normal vdevs (spa_normal_class)
The space required for the embedded slog metaslabs is usually between
0.5% and 1.0% of the pool, and comes out of the existing 3.2% of "slop"
space that is not available for user data.
On an all-ssd system with 4TB storage, 87% fragmentation, 60% capacity,
and recordsize=8k, testing shows a ~50% performance increase on random
8k sync writes. On even more fragmented systems (which hit problem #3
above and call txg_wait_synced()), the performance improvement can be
arbitrarily large (>100x).
Reviewed-by: Serapheim Dimitropoulos <serapheim@delphix.com>
Reviewed-by: George Wilson <gwilson@delphix.com>
Reviewed-by: Don Brady <don.brady@delphix.com>
Reviewed-by: Mark Maybee <mark.maybee@delphix.com>
Signed-off-by: Matthew Ahrens <mahrens@delphix.com>
Closes#11389
In FreeBSD the struct uio was just a typedef to uio_t. In order to
extend this struct, outside of the definition for the struct uio, the
struct uio has been embedded inside of a uio_t struct.
Also renamed all the uio_* interfaces to be zfs_uio_* to make it clear
this is a ZFS interface.
Reviewed-by: Ryan Moeller <ryan@iXsystems.com>
Reviewed-by: Jorgen Lundman <lundman@lundman.net>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Brian Atkinson <batkinson@lanl.gov>
Closes#11438
The `abd_get_offset_*()` routines create an abd_t that references
another abd_t, and doesn't allocate any pages/buffers of its own. In
some workloads, these routines may be called frequently, to create many
abd_t's representing small pieces of a single large abd_t. In
particular, the upcoming RAIDZ Expansion project makes heavy use of
these routines.
This commit adds the ability for the caller to allocate and provide the
abd_t struct to a variant of `abd_get_offset_*()`. This eliminates the
cost of allocating the abd_t and performing the accounting associated
with it (`abdstat_struct_size`). The RAIDZ/DRAID code uses this for
the `rc_abd`, which references the zio's abd. The upcoming RAIDZ
Expansion project will leverage this infrastructure to increase
performance of reads post-expansion by around 50%.
Additionally, some of the interfaces around creating and destroying
abd_t's are cleaned up. Most significantly, the distinction between
`abd_put()` and `abd_free()` is eliminated; all types of abd_t's are
now disposed of with `abd_free()`.
Reviewed-by: Brian Atkinson <batkinson@lanl.gov>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Matthew Ahrens <mahrens@delphix.com>
Issue #8853Closes#11439
Each zfs ioctl that changes on-disk state (e.g. set property, create
snapshot, destroy filesystem) is recorded in the zpool history, and is
printed by `zpool history -i`.
For performance diagnostic purposes, it would be useful to know how long
each of these ioctls took to run. This commit adds that functionality,
with a new `ZPOOL_HIST_ELAPSED_NS` member of the history nvlist.
Additionally, the time recorded in this history log is currently the
time that the history record is written to disk. But in many cases (CLI
args logging and ioctl logging), this happens asynchronously,
potentially many seconds after the operation completed. This commit
changes the timestamp to reflect when the history event was created,
rather than when it was written to disk.
Reviewed-by: Mark Maybee <mmaybee@cray.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Matthew Ahrens <mahrens@delphix.com>
Closes#11440
Build error on illumos with gcc 10 did reveal:
In function 'dmu_objset_refresh_ownership':
../../common/fs/zfs/dmu_objset.c:857:25: error: implicit conversion
from 'boolean_t' to 'ds_hold_flags_t' {aka 'enum ds_hold_flags'}
[-Werror=enum-conversion]
857 | dsl_dataset_disown(ds, decrypt, tag);
| ^~~~~~~
cc1: all warnings being treated as errors
libzfs_input_check.c: In function 'zfs_ioc_input_tests':
libzfs_input_check.c:754:28: error: implicit conversion from
'enum dmu_objset_type' to 'enum lzc_dataset_type'
[-Werror=enum-conversion]
754 | err = lzc_create(dataset, DMU_OST_ZFS, NULL, NULL, 0);
| ^~~~~~~~~~~
cc1: all warnings being treated as errors
The same issue is present in openzfs, and also the same issue about
ds_hold_flags_t, which currently defines exactly one valid value.
Reviewed-by: Igor Kozhukhov <igor@dilos.org>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Toomas Soome <tsoome@me.com>
Closes#11406
On a system with very high fragmentation, we may need to do lots of gang
allocations (e.g. most indirect block allocations (~50KB) may need to
gang). Before failing a "normal" allocation and resorting to ganging, we
try every metaslab. This has the impact of loading every metaslab (not
a huge deal since we now typically keep all metaslabs loaded), and also
iterating over every metaslab for every failing allocation. If there are
many metaslabs (more than the typical ~200, e.g. due to vdev expansion
or very large vdevs), the CPU cost of this iteration can be very
impactful. This iteration is done with the mg_lock held, creating long
hold times and high lock contention for concurrent allocations,
ultimately causing long txg sync times and poor application performance.
To address this, this commit changes the behavior of "normal" (not
try_hard, not ZIL) allocations. These will now only examine the 100
best metaslabs (as determined by their ms_weight). If none of these
have a large enough free segment, then the allocation will fail and
we'll fall back on ganging.
To accomplish this, we will now (normally) gang before doing a
`try_hard` allocation. Non-try_hard allocations will only examine the
100 best metaslabs of each vdev. In summary, we will first try normal
allocation. If that fails then we will do a gang allocation. If that
fails then we will do a "try hard" gang allocation. If that fails then
we will have a multi-layer gang block.
Reviewed-by: Paul Dagnelie <pcd@delphix.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Matthew Ahrens <mahrens@delphix.com>
Closes#11327
Metaslab rotor and aliquot are used to distribute workload between
vdevs while keeping some locality for logically adjacent blocks. Once
multiple allocators were introduced to separate allocation of different
objects it does not make much sense for different allocators to write
into different metaslabs of the same metaslab group (vdev) same time,
competing for its resources. This change makes each allocator choose
metaslab group independently, colliding with others only sporadically.
Test including simultaneous write into 4 files with recordsize of 4KB
on a striped pool of 30 disks on a system with 40 logical cores show
reduction of vdev queue lock contention from 54 to 27% due to better
load distribution. Unfortunately it won't help much ZVOLs yet since
only one dataset/ZVOL is synced at a time, and so for the most part
only one allocator is used, but it may improve later.
While there, to reduce the number of pointer dereferences change
per-allocator storage for metaslab classes and groups from several
separate malloc()'s to variable length arrays at the ends of the
original class and group structures.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Matthew Ahrens <mahrens@delphix.com>
Signed-off-by: Alexander Motin <mav@FreeBSD.org>
Closes#11288
Building the spa module for i386 caused gcc to emit
-Wint-to-pointer-cast "cast to pointer from integer of different size"
because spa.spa_did was uint64_t but pthread_join (via thread_join in
spa_deactivate) takes a pointer (32-bit on i386). Define spa_did to be
pointer-size instead. For now spa_did is in fact never non-zero and the
thread_join could instead be ifdef'd out, but changing the size of
spa_did may be more useful for the future.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Ryan Libby <rlibby@FreeBSD.org>
Closes#11336
The last change caused the read completion callback to not be called
if the IO was still in progress. This change restores allocation
of the arc buf callback, but in the callback path checks the new
acb_nobuf field to know to skip buffer allocation.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Matt Macy <mmacy@FreeBSD.org>
Closes#11324
The performance of `zfs receive` can be bottlenecked on the CPU consumed
by the `receive_writer` thread, especially when receiving streams with
small compressed block sizes. Much of the CPU is spent creating and
destroying dbuf's and arc buf's, one for each `WRITE` record in the send
stream.
This commit introduces the concept of "lightweight writes", which allows
`zfs receive` to write to the DMU by providing an ABD, and instantiating
only a new type of `dbuf_dirty_record_t`. The dbuf and arc buf for this
"dirty leaf block" are not instantiated.
Because there is no dbuf with the dirty data, this mechanism doesn't
support reading from "lightweight-dirty" blocks (they would see the
on-disk state rather than the dirty data). Since the dedup-receive code
has been removed, `zfs receive` is write-only, so this works fine.
Because there are no arc bufs for the received data, the received data
is no longer cached in the ARC.
Testing a receive of a stream with average compressed block size of 4KB,
this commit improves performance by 50%, while also reducing CPU usage
by 50% of a CPU. On a per-block basis, CPU consumed by receive_writer()
and dbuf_evict() is now 1/7th (14%) of what it was.
Baseline: 450MB/s, CPU in receive_writer() 40% + dbuf_evict() 35%
New: 670MB/s, CPU in receive_writer() 17% + dbuf_evict() 0%
The code is also restructured in a few ways:
Added a `dr_dnode` field to the dbuf_dirty_record_t. This simplifies
some existing code that no longer needs `DB_DNODE_ENTER()` and related
routines. The new field is needed by the lightweight-type dirty record.
To ensure that the `dr_dnode` field remains valid until the dirty record
is freed, we have to ensure that the `dnode_move()` doesn't relocate the
dnode_t. To do this we keep a hold on the dnode until it's zio's have
completed. This is already done by the user-accounting code
(`userquota_updates_task()`), this commit extends that so that it always
keeps the dnode hold until zio completion (see `dnode_rele_task()`).
`dn_dirty_txg` was previously zeroed when the dnode was synced. This
was not necessary, since its meaning can be "when was this dnode last
dirtied". This change simplifies the new `dnode_rele_task()` code.
Removed some dead code related to `DRR_WRITE_BYREF` (dedup receive).
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Paul Dagnelie <pcd@delphix.com>
Reviewed-by: George Wilson <gwilson@delphix.com>
Signed-off-by: Matthew Ahrens <mahrens@delphix.com>
Closes#11105
ZFS currently doesn't react to hotplugging cpu or memory into the
system in any way. This patch changes that by adding logic to the ARC
that allows the system to take advantage of new memory that is added
for caching purposes. It also adds logic to the taskq infrastructure
to support dynamically expanding the number of threads allocated to a
taskq.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Co-authored-by: Matthew Ahrens <matthew.ahrens@delphix.com>
Co-authored-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Paul Dagnelie <pcd@delphix.com>
Closes#11212
Add ARC_FLAG_NO_BUF to indicate that a buffer need not be
instantiated. This fixes a ~20% performance regression on
cached reads due to zfetch changes.
Reviewed-by: Tony Nguyen <tony.nguyen@delphix.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Matthew Ahrens <mahrens@delphix.com>
Signed-off-by: Matt Macy <mmacy@FreeBSD.org>
Closes#11220Closes#11232
Investigating influence of scrub (especially sequential) on random read
latency I've noticed that on some HDDs single 4KB read may take up to 4
seconds! Deeper investigation shown that many HDDs heavily prioritize
sequential reads even when those are submitted with queue depth of 1.
This patch addresses the latency from two sides:
- by using _min_active queue depths for non-interactive requests while
the interactive request(s) are active and few requests after;
- by throttling it further if no interactive requests has completed
while configured amount of non-interactive did.
While there, I've also modified vdev_queue_class_to_issue() to give
more chances to schedule at least _min_active requests to the lowest
priorities. It should reduce starvation if several non-interactive
processes are running same time with some interactive and I think should
make possible setting of zfs_vdev_max_active to as low as 1.
I've benchmarked this change with 4KB random reads from ZVOL with 16KB
block size on newly written non-fragmented pool. On fragmented pool I
also saw improvements, but not so dramatic. Below are log2 histograms
of the random read latency in milliseconds for different devices:
4 2x mirror vdevs of SATA HDD WDC WD20EFRX-68EUZN0 before:
0, 0, 2, 1, 12, 21, 19, 18, 10, 15, 17, 21
after:
0, 0, 0, 24, 101, 195, 419, 250, 47, 4, 0, 0
, that means maximum latency reduction from 2s to 500ms.
4 2x mirror vdevs of SATA HDD WDC WD80EFZX-68UW8N0 before:
0, 0, 2, 31, 38, 28, 18, 12, 17, 20, 24, 10, 3
after:
0, 0, 55, 247, 455, 470, 412, 181, 36, 0, 0, 0, 0
, i.e. from 4s to 250ms.
1 SAS HDD SEAGATE ST14000NM0048 before:
0, 0, 29, 70, 107, 45, 27, 1, 0, 0, 1, 4, 19
after:
1, 29, 681, 1261, 676, 1633, 67, 1, 0, 0, 0, 0, 0
, i.e. from 4s to 125ms.
1 SAS SSD SEAGATE XS3840TE70014 before (microseconds):
0, 0, 0, 0, 0, 0, 0, 0, 70, 18343, 82548, 618
after:
0, 0, 0, 0, 0, 0, 0, 0, 283, 92351, 34844, 90
I've also measured scrub time during the test and on idle pools. On
idle fragmented pool I've measured scrub getting few percent faster
due to use of QD3 instead of QD2 before. On idle non-fragmented pool
I've measured no difference. On busy non-fragmented pool I've measured
scrub time increase about 1.5-1.7x, while IOPS increase reached 5-9x.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Matthew Ahrens <mahrens@delphix.com>
Reviewed-by: Ryan Moeller <ryan@iXsystems.com>
Signed-off-by: Alexander Motin <mav@FreeBSD.org>
Sponsored-By: iXsystems, Inc.
Closes#11166
- Don't leave fstrans set when passed a snapshot
- Don't remove minor if volmode already matches new value
- (FreeBSD) Wait for GEOM ops to complete before trying
remove (at create time GEOM will be "tasting" in parallel)
- (FreeBSD) Don't leak zvol_state_lock on open if zv == NULL
- (FreeBSD) Don't try to unlock zv->zv_state lock if zv == NULL
Reviewed-by: Ryan Moeller <ryan@iXsystems.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Matt Macy <mmacy@FreeBSD.org>
Closes#11199
For encrypted receives, where user accounting is initially disabled on
creation, both 'zfs userspace' and 'zfs groupspace' fails with
EOPNOTSUPP: this is because dmu_objset_id_quota_upgrade_cb() forgets to
set OBJSET_FLAG_USERACCOUNTING_COMPLETE on the objset flags after a
successful dmu_objset_space_upgrade().
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Co-authored-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: loli10K <ezomori.nozomu@gmail.com>
Closes#9501Closes#9596
The output of ZFS channel programs is logged on-disk in the zpool
history, and printed by `zpool history -i`. Channel programs can use
10MB of memory by default, and up to 100MB by using the `zfs program -m`
flag. Therefore their output can be up to some fraction of 100MB.
In addition to being somewhat wasteful of the limited space reserved for
the pool history (which for large pools is 1GB), in extreme cases this
can result in a failure of `ASSERT(length <= DMU_MAX_ACCESS);` in
`dmu_buf_hold_array_by_dnode()`.
This commit limits the output size that will be logged to 1MB. Larger
outputs will not be logged, instead a entry will be logged indicating
the size of the omitted output.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Matthew Ahrens <mahrens@delphix.com>
Closes#11194
Alexander Motin
Attila Fülöp
Rich Ercolani
Alan Somers
Brian Behlendorf
наб
Ryan Moeller
Andrea Gelmini
Matthew Ahrens
Alexander Motin
Chunwei Chen
Prakash Surya
Brian Atkinson
Christian Schwarz
Don Brady
Colm
Ryan Moeller
khng300
Alan Somers
Toomas Soome
Ryan Libby
Matthew Macy
Paul Dagnelie
loli10K