Object allocation performance can be improved for complex operations
by providing an interface which returns the newly allocated dnode.
This allows the caller to immediately use the dnode without incurring
the expense of looking up the dnode by object number.
The functions dmu_object_alloc_hold(), zap_create_hold(), and
dmu_bonus_hold_by_dnode() were added for this purpose.
The zap_create_* functions have been updated to take advantage of
this new functionality. The dmu_bonus_hold_impl() function should
really have never been included in sys/dmu.h and was removed.
It's sole caller was converted to use dmu_bonus_hold_by_dnode().
The new symbols have been exported for use by Lustre.
Reviewed-by: Tony Hutter <hutter2@llnl.gov>
Reviewed by: Matt Ahrens <mahrens@delphix.com>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes#8015
The dmu_objset_remap_indirects_impl() logic depends on dnode_hold()
returning ENOENT for dnodes which will be freed and should be skipped.
This behavior can only be relied upon when taking a new hold and
while the caller has an open transaction. This ensures that the
open txg cannot advance and that a concurrent free will end up
in the same txg (which is critical). Relying on an existing hold
will not prevent dnode_free() from succeeding.
The solution is to take an additional dnode_hold() after assigning
the transaction. This ensures the remap will never dirty the dnode
if it was freed while we were waiting in dmu_tx_assign(, TXG_WAIT).
Randomly set zfs_object_remap_one_indirect_delay_ms in ztest. This
increases the likelihood of an operation racing with the remap.
Converted from ticks to milliseconds.
Reviewed by: Matt Ahrens <mahrens@delphix.com>
Reviewed by: Tom Caputi <tcaputi@datto.com>
Reviewed by: Igor Kozhukhov <igor@dilos.org>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes#8215
As a result of the changes made in 8585, it's possible for an excessive
amount of vdev flush commands to be issued under some workloads.
Specifically, when the workload consists of mostly async write activity,
interspersed with some sync write and/or fsync activity, we can end up
issuing more flush commands to the underlying storage than is actually
necessary. As a result of these flush commands, the write latency and
overall throughput of the pool can be poorly impacted (latency
increases, throughput decreases).
Currently, any time an lwb completes, the vdev(s) written to as a result
of that lwb will be issued a flush command. The intenion is so the data
written to that vdev is on stable storage, prior to communicating to any
waiting threads that their data is safe on disk.
The problem with this scheme, is that sometimes an lwb will not have any
threads waiting for it to complete. This can occur when there's async
activity that gets "converted" to sync requests, as a result of calling
the zil_async_to_sync() function via zil_commit_impl(). When this
occurs, the current code may issue many lwbs that don't have waiters
associated with them, resulting in many flush commands, potentially to
the same vdev(s).
For example, given a pool with a single vdev, and a single fsync() call
that results in 10 lwbs being written out (e.g. due to other async
writes), that will result in 10 flush commands to that single vdev (a
flush issued after each lwb write completes). Ideally, we'd only issue a
single flush command to that vdev, after all 10 lwb writes completed.
Further, and most important as it pertains to this change, since the
flush commands are often very impactful to the performance of the pool's
underlying storage, unnecessarily issuing these flush commands can
poorly impact the performance of the lwb writes themselves. Thus, we
need to avoid issuing flush commands when possible, in order to acheive
the best possible performance out of the pool's underlying storage.
This change attempts to address this problem by changing the ZIL's logic
to only issue a vdev flush command when it detects an lwb that has a
thread waiting for it to complete. When an lwb does not have threads
waiting for it, the responsibility of issuing the flush command to the
vdevs involved with that lwb's write is passed on to the "next" lwb.
It's only once a write for an lwb with waiters completes, do we issue
the vdev flush command(s). As a result, now when we issue the flush(s),
we will issue them to the vdevs involved with that specific lwb's write,
but potentially also to vdevs involved with "previous" lwb writes (i.e.
if the previous lwbs did not have waiters associated with them).
Thus, in our prior example with 10 lwbs, it's only once the last lwb
completes (which will be the lwb containing the waiter for the thread
that called fsync) will we issue the vdev flush command; all of the
other lwbs will find they have no waiters, so they'll pass the
responsibility of the flush to the "next" lwb (until reaching the last
lwb that has the waiter).
Porting Notes:
* Reconciled conflicts with the fastwrite feature.
Authored by: Prakash Surya <prakash.surya@delphix.com>
Reviewed by: Matt Ahrens <matt@delphix.com>
Reviewed by: Brad Lewis <brad.lewis@delphix.com>
Reviewed by: Patrick Mooney <patrick.mooney@joyent.com>
Reviewed by: Jerry Jelinek <jerry.jelinek@joyent.com>
Approved by: Joshua M. Clulow <josh@sysmgr.org>
Ported-by: Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
OpenZFS-issue: https://www.illumos.org/issues/9962
OpenZFS-commit: https://github.com/openzfs/openzfs/commit/545190c6Closes#8188
The ZFS range locking code in zfs_rlock.c/h depends on ZPL-specific
data structures, specifically znode_t. However, it's also used by
the ZVOL code, which uses a "dummy" znode_t to pass to the range
locking code.
We should clean this up so that the range locking code is generic
and can be used equally by ZPL and ZVOL, and also can be used by
future consumers that may need to run in userland (libzpool) as
well as the kernel.
Porting notes:
* Added missing sys/avl.h include to sys/zfs_rlock.h.
* Removed 'dbuf is within the locked range' ASSERTs from dmu_sync().
This was needed because ztest does not yet use a locked_range_t.
* Removed "Approved by:" tag requirement from OpenZFS commit
check to prevent needless warnings when integrating changes
which has not been merged to illumos.
* Reverted free_list range lock changes which were originally
needed to defer the cv_destroy() which was called immediately
after cv_broadcast(). With d2733258 this should be safe but
if not we may need to reintroduce this logic.
* Reverts: The following two commits were reverted and squashed in
to this change in order to make it easier to apply OpenZFS 9689.
- d88895a0, which removed the dummy znode from zvol_state
- e3a07cd0, which updated ztest to use range locks
* Preserved optimized rangelock comparison function. Preserved the
rangelock free list. The cv_destroy() function will block waiting
for all processes in cv_wait() to be scheduled and drop their
reference. This is done to ensure it's safe to free the condition
variable. However, blocking while holding the rl->rl_lock mutex
can result in a deadlock on Linux. A free list is introduced to
defer the cv_destroy() and kmem_free() until after the mutex is
released.
Authored by: Matthew Ahrens <mahrens@delphix.com>
Reviewed by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed by: Serapheim Dimitropoulos <serapheim.dimitro@delphix.com>
Reviewed by: George Wilson <george.wilson@delphix.com>
Reviewed by: Brad Lewis <brad.lewis@delphix.com>
Ported-by: Brian Behlendorf <behlendorf1@llnl.gov>
OpenZFS-issue: https://illumos.org/issues/9689
OpenZFS-commit: https://github.com/openzfs/openzfs/pull/680
External-issue: DLPX-58662
Closes#7980
torvalds/linux@59b57717f ("blkcg: delay blkg destruction until
after writeback has finished") added a refcount_t to the blkcg
structure. Due to the refcount_t compatibility code, zfs_refcount_t
was used by mistake.
Resolve this by removing the compatibility code and replacing the
occurrences of refcount_t with zfs_refcount_t.
Reviewed-by: Franz Pletz <fpletz@fnordicwalking.de>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Tim Schumacher <timschumi@gmx.de>
Closes#7885Closes#7932
Allocation Classes add the ability to have allocation classes in a
pool that are dedicated to serving specific block categories, such
as DDT data, metadata, and small file blocks. A pool can opt-in to
this feature by adding a 'special' or 'dedup' top-level VDEV.
Reviewed by: Pavel Zakharov <pavel.zakharov@delphix.com>
Reviewed-by: Richard Laager <rlaager@wiktel.com>
Reviewed-by: Alek Pinchuk <apinchuk@datto.com>
Reviewed-by: Håkan Johansson <f96hajo@chalmers.se>
Reviewed-by: Andreas Dilger <andreas.dilger@chamcloud.com>
Reviewed-by: DHE <git@dehacked.net>
Reviewed-by: Richard Elling <Richard.Elling@RichardElling.com>
Reviewed-by: Gregor Kopka <gregor@kopka.net>
Reviewed-by: Kash Pande <kash@tripleback.net>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Matthew Ahrens <mahrens@delphix.com>
Signed-off-by: Don Brady <don.brady@delphix.com>
Closes#5182
This patch fixes 2 issues with raw, deduplicated send streams. The
first is that datasets who had been completely received earlier in
the stream were not still marked as raw receives. This caused
problems when newly received datasets attempted to fetch raw data
from these datasets without this flag set.
The second problem was that the arc freeze checksum code was not
consistent about which locks needed to be held while performing
its asserts. The proper locking needed to run these asserts is
actually fairly nuanced, since the asserts touch the linked list
of buffers (requiring the header lock), the arc_state (requiring
the b_evict_lock), and the b_freeze_cksum (requiring the
b_freeze_lock). This seems like a large performance sacrifice and
a lot of unneeded complexity to verify that this relatively small
debug feature is working as intended, so this patch simply removes
these asserts instead.
Reviewed by: Matthew Ahrens <mahrens@delphix.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Tom Caputi <tcaputi@datto.com>
Closes#7701
It is helpful to tune zfs_per_txg_dirty_frees_percent for commit
539d33c7(OpenZFS 6569 - large file delete can starve out write ops).
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed by: Richard Elling <Richard.Elling@RichardElling.com>
Signed-off-by: Feng Sun <loyou85@gmail.com>
Closes#7718
This project's goal is to make read-heavy channel programs and zfs(1m)
administrative commands faster by caching all the metadata that they will
need in the dbuf layer. This will prevent the data from being evicted, so
that any future call to i.e. zfs get all won't have to go to disk (very
much). There are two parts:
The dbuf_metadata_cache. We identify what to put into the cache based on
the object type of each dbuf. Caching objset properties os
{version,normalization,utf8only,casesensitivity} in the objset_t. The reason
these needed to be cached is that although they are queried frequently,
they aren't stored in a dbuf type which we can easily recognize and cache in
the dbuf layer; instead, we have to explicitly store them. There's already
existing infrastructure for maintaining cached properties in the objset
setup code, so I simply used that.
Performance Testing:
- Disabled kmem_flags
- Tuned dbuf_cache_max_bytes very low (128K)
- Tuned zfs_arc_max very low (64M)
Created test pool with 400 filesystems, and 100 snapshots per filesystem.
Later on in testing, added 600 more filesystems (with no snapshots) to make
sure scaling didn't look different between snapshots and filesystems.
Results:
| Test | Time (trunk / diff) | I/Os (trunk / diff) |
+------------------------+---------------------+---------------------+
| zpool import | 0:05 / 0:06 | 12.9k / 12.9k |
| zfs get all (uncached) | 1:36 / 0:53 | 16.7k / 5.7k |
| zfs get all (cached) | 1:36 / 0:51 | 16.0k / 6.0k |
Authored by: Matthew Ahrens <mahrens@delphix.com>
Reviewed by: Prakash Surya <prakash.surya@delphix.com>
Reviewed by: George Wilson <george.wilson@delphix.com>
Reviewed by: Thomas Caputi <tcaputi@datto.com>
Reviewed by: Brian Behlendorf <behlendorf1@llnl.gov>
Approved by: Richard Lowe <richlowe@richlowe.net>
Ported-by: Alek Pinchuk <apinchuk@datto.com>
Signed-off-by: Alek Pinchuk <apinchuk@datto.com>
OpenZFS-issue: https://illumos.org/issues/9337
OpenZFS-commit: https://github.com/openzfs/openzfs/commit/7dec52fCloses#7668
This patch fixes a small bug found where receive_spill() sometimes
attempted to decrypt spill blocks when doing a raw receive. In
addition, this patch fixes another small issue in arc_buf_fill()'s
error handling where a decryption failure (which could be caused by
the first bug) would attempt to set the arc header's IO_ERROR flag
without holding the header's lock.
Reviewed-by: Matthew Thode <prometheanfire@gentoo.org>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed by: Matthew Ahrens <mahrens@delphix.com>
Signed-off-by: Tom Caputi <tcaputi@datto.com>
Closes#7564Closes#7584Closes#7592
Minimal changes required to integrate the SPL sources in to the
ZFS repository build infrastructure and packaging.
Build system and packaging:
* Renamed SPL_* autoconf m4 macros to ZFS_*.
* Removed redundant SPL_* autoconf m4 macros.
* Updated the RPM spec files to remove SPL package dependency.
* The zfs package obsoletes the spl package, and the zfs-kmod
package obsoletes the spl-kmod package.
* The zfs-kmod-devel* packages were updated to add compatibility
symlinks under /usr/src/spl-x.y.z until all dependent packages
can be updated. They will be removed in a future release.
* Updated copy-builtin script for in-kernel builds.
* Updated DKMS package to include the spl.ko.
* Updated stale AUTHORS file to include all contributors.
* Updated stale COPYRIGHT and included the SPL as an exception.
* Renamed README.markdown to README.md
* Renamed OPENSOLARIS.LICENSE to LICENSE.
* Renamed DISCLAIMER to NOTICE.
Required code changes:
* Removed redundant HAVE_SPL macro.
* Removed _BOOT from nvpairs since it doesn't apply for Linux.
* Initial header cleanup (removal of empty headers, refactoring).
* Remove SPL repository clone/build from zimport.sh.
* Use of DEFINE_RATELIMIT_STATE and DEFINE_SPINLOCK removed due
to build issues when forcing C99 compilation.
* Replaced legacy ACCESS_ONCE with READ_ONCE.
* Include needed headers for `current` and `EXPORT_SYMBOL`.
Reviewed-by: Tony Hutter <hutter2@llnl.gov>
Reviewed-by: Olaf Faaland <faaland1@llnl.gov>
Reviewed-by: Matthew Ahrens <mahrens@delphix.com>
Reviewed-by: Pavel Zakharov <pavel.zakharov@delphix.com>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
TEST_ZIMPORT_SKIP="yes"
Closes#7556
In the existing code, when doing a raw (encrypted) zfs receive,
we call arc_convert_to_raw() from open context. This creates a
race condition between arc_release()/arc_change_state() and
writing out the block from syncing context (arc_write_ready/done()).
This change makes it so that when we are doing a raw (encrypted)
zfs receive, we save the crypt parameters (salt, iv, mac) of dnode
blocks in the dbuf_dirty_record_t, and call arc_convert_to_raw()
from syncing context when writing out the block of dnodes.
Additionally, we can eliminate dr_raw and associated setters, and
instead know that dnode blocks are always raw when doing a zfs
receive (see the new field os_raw_receive).
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Tom Caputi <tcaputi@datto.com>
Signed-off-by: Matthew Ahrens <mahrens@delphix.com>
Closes#7424Closes#7429
OpenZFS 7614 - zfs device evacuation/removal
OpenZFS 9064 - remove_mirror should wait for device removal to complete
This project allows top-level vdevs to be removed from the storage pool
with "zpool remove", reducing the total amount of storage in the pool.
This operation copies all allocated regions of the device to be removed
onto other devices, recording the mapping from old to new location.
After the removal is complete, read and free operations to the removed
(now "indirect") vdev must be remapped and performed at the new location
on disk. The indirect mapping table is kept in memory whenever the pool
is loaded, so there is minimal performance overhead when doing operations
on the indirect vdev.
The size of the in-memory mapping table will be reduced when its entries
become "obsolete" because they are no longer used by any block pointers
in the pool. An entry becomes obsolete when all the blocks that use
it are freed. An entry can also become obsolete when all the snapshots
that reference it are deleted, and the block pointers that reference it
have been "remapped" in all filesystems/zvols (and clones). Whenever an
indirect block is written, all the block pointers in it will be "remapped"
to their new (concrete) locations if possible. This process can be
accelerated by using the "zfs remap" command to proactively rewrite all
indirect blocks that reference indirect (removed) vdevs.
Note that when a device is removed, we do not verify the checksum of
the data that is copied. This makes the process much faster, but if it
were used on redundant vdevs (i.e. mirror or raidz vdevs), it would be
possible to copy the wrong data, when we have the correct data on e.g.
the other side of the mirror.
At the moment, only mirrors and simple top-level vdevs can be removed
and no removal is allowed if any of the top-level vdevs are raidz.
Porting Notes:
* Avoid zero-sized kmem_alloc() in vdev_compact_children().
The device evacuation code adds a dependency that
vdev_compact_children() be able to properly empty the vdev_child
array by setting it to NULL and zeroing vdev_children. Under Linux,
kmem_alloc() and related functions return a sentinel pointer rather
than NULL for zero-sized allocations.
* Remove comment regarding "mpt" driver where zfs_remove_max_segment
is initialized to SPA_MAXBLOCKSIZE.
Change zfs_condense_indirect_commit_entry_delay_ticks to
zfs_condense_indirect_commit_entry_delay_ms for consistency with
most other tunables in which delays are specified in ms.
* ZTS changes:
Use set_tunable rather than mdb
Use zpool sync as appropriate
Use sync_pool instead of sync
Kill jobs during test_removal_with_operation to allow unmount/export
Don't add non-disk names such as "mirror" or "raidz" to $DISKS
Use $TEST_BASE_DIR instead of /tmp
Increase HZ from 100 to 1000 which is more common on Linux
removal_multiple_indirection.ksh
Reduce iterations in order to not time out on the code
coverage builders.
removal_resume_export:
Functionally, the test case is correct but there exists a race
where the kernel thread hasn't been fully started yet and is
not visible. Wait for up to 1 second for the removal thread
to be started before giving up on it. Also, increase the
amount of data copied in order that the removal not finish
before the export has a chance to fail.
* MMP compatibility, the concept of concrete versus non-concrete devices
has slightly changed the semantics of vdev_writeable(). Update
mmp_random_leaf_impl() accordingly.
* Updated dbuf_remap() to handle the org.zfsonlinux:large_dnode pool
feature which is not supported by OpenZFS.
* Added support for new vdev removal tracepoints.
* Test cases removal_with_zdb and removal_condense_export have been
intentionally disabled. When run manually they pass as intended,
but when running in the automated test environment they produce
unreliable results on the latest Fedora release.
They may work better once the upstream pool import refectoring is
merged into ZoL at which point they will be re-enabled.
Authored by: Matthew Ahrens <mahrens@delphix.com>
Reviewed-by: Alex Reece <alex@delphix.com>
Reviewed-by: George Wilson <george.wilson@delphix.com>
Reviewed-by: John Kennedy <john.kennedy@delphix.com>
Reviewed-by: Prakash Surya <prakash.surya@delphix.com>
Reviewed by: Richard Laager <rlaager@wiktel.com>
Reviewed by: Tim Chase <tim@chase2k.com>
Reviewed by: Brian Behlendorf <behlendorf1@llnl.gov>
Approved by: Garrett D'Amore <garrett@damore.org>
Ported-by: Tim Chase <tim@chase2k.com>
Signed-off-by: Tim Chase <tim@chase2k.com>
OpenZFS-issue: https://www.illumos.org/issues/7614
OpenZFS-commit: https://github.com/openzfs/openzfs/commit/f539f1ebCloses#6900
Currently, dnode_check_slots_free() works by checking dn->dn_type
in the dnode to determine if the dnode is reclaimable. However,
there is a small window of time between dnode_free_sync() in the
first call to dsl_dataset_sync() and when the useraccounting code
is run when the type is set DMU_OT_NONE, but the dnode is not yet
evictable, leading to crashes. This patch adds the ability for
dnodes to track which txg they were last dirtied in and adds a
check for this before performing the reclaim.
This patch also corrects several instances when dn_dirty_link was
treated as a list_node_t when it is technically a multilist_node_t.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Tom Caputi <tcaputi@datto.com>
Closes#7147Closes#7388
b1d21733 made it possible for empty metadnode blocks to be
compressed to a hole, fixing a bug that would cause invalid
metadnode MACs when a send stream attempted to free objects
and allowing the blocks to be reclaimed when they were no
longer needed. However, this patch also introduced a race
condition; if a txg sync occurred after a DRR_OBJECT_RANGE
record was received but before any objects were added, the
metadnode block would be compressed to a hole and lose all
of its encryption parameters. This would cause subsequent
DRR_OBJECT records to fail when they attempted to write
their data into an unencrypted block. This patch defers the
DRR_OBJECT_RANGE handling to receive_object() so that the
encryption parameters are set with each object that is
written into that block.
Reviewed-by: Kash Pande <kash@tripleback.net>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Tom Caputi <tcaputi@datto.com>
Closes#7215Closes#7236
Currently, the DMU relies on ZIO layer compression to free LO
dnode blocks that no longer have objects in them. However,
raw receives disable all compression, meaning that these blocks
can never be freed. In addition to the obvious space concerns,
this could also cause incremental raw receives to fail to mount
since the MAC of a hole is different from that of a completely
zeroed block.
This patch corrects this issue by adding a special case in
zio_write_compress() which will attempt to compress these blocks
to a hole even if ZIO_FLAG_RAW_ENCRYPT is set. This patch also
removes the zfs_mdcomp_disable tunable, since tuning it could
cause these same issues.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Tom Caputi <tcaputi@datto.com>
Closes#7198
This one line patch adds adds a set to os->os_next_write_raw
that was omitted when the code was updated in 1b66810. Without
it, the code (in some instances) could attempt to write raw
encrypted data as regular unencrypted data without the keys
being loaded, triggering an ASSERT in zio_encrypt().
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Tom Caputi <tcaputi@datto.com>
Closes#7196
Project quota is a new ZFS system space/object usage accounting
and enforcement mechanism. Similar as user/group quota, project
quota is another dimension of system quota. It bases on the new
object attribute - project ID.
Project ID is a numerical value to indicate to which project an
object belongs. An object only can belong to one project though
you (the object owner or privileged user) can change the object
project ID via 'chattr -p' or 'zfs project [-s] -p' explicitly.
The object also can inherit the project ID from its parent when
created if the parent has the project inherit flag (that can be
set via 'chattr +P' or 'zfs project -s [-p]').
By accounting the spaces/objects belong to the same project, we
can know how many spaces/objects used by the project. And if we
set the upper limit then we can control the spaces/objects that
are consumed by such project. It is useful when multiple groups
and users cooperate for the same project, or a user/group needs
to participate in multiple projects.
Support the following commands and functionalities:
zfs set projectquota@project
zfs set projectobjquota@project
zfs get projectquota@project
zfs get projectobjquota@project
zfs get projectused@project
zfs get projectobjused@project
zfs projectspace
zfs allow projectquota
zfs allow projectobjquota
zfs allow projectused
zfs allow projectobjused
zfs unallow projectquota
zfs unallow projectobjquota
zfs unallow projectused
zfs unallow projectobjused
chattr +/-P
chattr -p project_id
lsattr -p
This patch also supports tree quota based on the project quota via
"zfs project" commands set as following:
zfs project [-d|-r] <file|directory ...>
zfs project -C [-k] [-r] <file|directory ...>
zfs project -c [-0] [-d|-r] [-p id] <file|directory ...>
zfs project [-p id] [-r] [-s] <file|directory ...>
For "df [-i] $DIR" command, if we set INHERIT (project ID) flag on
the $DIR, then the proejct [obj]quota and [obj]used values for the
$DIR's project ID will be shown as the total/free (avail) resource.
Keep the same behavior as EXT4/XFS does.
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by Ned Bass <bass6@llnl.gov>
Reviewed-by: Matthew Ahrens <mahrens@delphix.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Fan Yong <fan.yong@intel.com>
TEST_ZIMPORT_POOLS="zol-0.6.1 zol-0.6.2 master"
Change-Id: Ib4f0544602e03fb61fd46a849d7ba51a6005693c
Closes#6290
The on-disk format for encrypted datasets protects not only
the encrypted and authenticated blocks themselves, but also
the order and interpretation of these blocks. In order to
make this work while maintaining the ability to do raw
sends, the indirect bps maintain a secure checksum of all
the MACs in the block below it along with a few other
fields that determine how the data is interpreted.
Unfortunately, the current on-disk format erroneously
includes some fields which are not portable and thus cannot
support raw sends. It is not possible to easily work around
this issue due to a separate and much smaller bug which
causes indirect blocks for encrypted dnodes to not be
compressed, which conflicts with the previous bug. In
addition, the current code generates incompatible on-disk
formats on big endian and little endian systems due to an
issue with how block pointers are authenticated. Finally,
raw send streams do not currently include dn_maxblkid when
sending both the metadnode and normal dnodes which are
needed in order to ensure that we are correctly maintaining
the portable objset MAC.
This patch zero's out the offending fields when computing
the bp MAC and ensures that these MACs are always
calculated in little endian order (regardless of the host
system's byte order). This patch also registers an errata
for the old on-disk format, which we detect by adding a
"version" field to newly created DSL Crypto Keys. We allow
datasets without a version (version 0) to only be mounted
for read so that they can easily be migrated. We also now
include dn_maxblkid in raw send streams to ensure the MAC
can be maintained correctly.
This patch also contains minor bug fixes and cleanups.
Reviewed-by: Jorgen Lundman <lundman@lundman.net>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed by: Matthew Ahrens <mahrens@delphix.com>
Signed-off-by: Tom Caputi <tcaputi@datto.com>
Closes#6845Closes#6864Closes#7052
Authored by: Prakash Surya <prakash.surya@delphix.com>
Reviewed by: Brad Lewis <brad.lewis@delphix.com>
Reviewed by: Matt Ahrens <mahrens@delphix.com>
Reviewed by: George Wilson <george.wilson@delphix.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Approved by: Dan McDonald <danmcd@joyent.com>
Ported-by: Prakash Surya <prakash.surya@delphix.com>
Problem
=======
The current implementation of zil_commit() can introduce significant
latency, beyond what is inherent due to the latency of the underlying
storage. The additional latency comes from two main problems:
1. When there's outstanding ZIL blocks being written (i.e. there's
already a "writer thread" in progress), then any new calls to
zil_commit() will block waiting for the currently oustanding ZIL
blocks to complete. The blocks written for each "writer thread" is
coined a "batch", and there can only ever be a single "batch" being
written at a time. When a batch is being written, any new ZIL
transactions will have to wait for the next batch to be written,
which won't occur until the current batch finishes.
As a result, the underlying storage may not be used as efficiently
as possible. While "new" threads enter zil_commit() and are blocked
waiting for the next batch, it's possible that the underlying
storage isn't fully utilized by the current batch of ZIL blocks. In
that case, it'd be better to allow these new threads to generate
(and issue) a new ZIL block, such that it could be serviced by the
underlying storage concurrently with the other ZIL blocks that are
being serviced.
2. Any call to zil_commit() must wait for all ZIL blocks in its "batch"
to complete, prior to zil_commit() returning. The size of any given
batch is proportional to the number of ZIL transaction in the queue
at the time that the batch starts processing the queue; which
doesn't occur until the previous batch completes. Thus, if there's a
lot of transactions in the queue, the batch could be composed of
many ZIL blocks, and each call to zil_commit() will have to wait for
all of these writes to complete (even if the thread calling
zil_commit() only cared about one of the transactions in the batch).
To further complicate the situation, these two issues result in the
following side effect:
3. If a given batch takes longer to complete than normal, this results
in larger batch sizes, which then take longer to complete and
further drive up the latency of zil_commit(). This can occur for a
number of reasons, including (but not limited to): transient changes
in the workload, and storage latency irregularites.
Solution
========
The solution attempted by this change has the following goals:
1. no on-disk changes; maintain current on-disk format.
2. modify the "batch size" to be equal to the "ZIL block size".
3. allow new batches to be generated and issued to disk, while there's
already batches being serviced by the disk.
4. allow zil_commit() to wait for as few ZIL blocks as possible.
5. use as few ZIL blocks as possible, for the same amount of ZIL
transactions, without introducing significant latency to any
individual ZIL transaction. i.e. use fewer, but larger, ZIL blocks.
In theory, with these goals met, the new allgorithm will allow the
following improvements:
1. new ZIL blocks can be generated and issued, while there's already
oustanding ZIL blocks being serviced by the storage.
2. the latency of zil_commit() should be proportional to the underlying
storage latency, rather than the incoming synchronous workload.
Porting Notes
=============
Due to the changes made in commit 119a394ab0, the lifetime of an itx
structure differs than in OpenZFS. Specifically, the itx structure is
kept around until the data associated with the itx is considered to be
safe on disk; this is so that the itx's callback can be called after the
data is committed to stable storage. Since OpenZFS doesn't have this itx
callback mechanism, it's able to destroy the itx structure immediately
after the itx is committed to an lwb (before the lwb is written to
disk).
To support this difference, and to ensure the itx's callbacks can still
be called after the itx's data is on disk, a few changes had to be made:
* A list of itxs was added to the lwb structure. This list contains
all of the itxs that have been committed to the lwb, such that the
callbacks for these itxs can be called from zil_lwb_flush_vdevs_done(),
after the data for the itxs is committed to disk.
* A list of itxs was added on the stack of the zil_process_commit_list()
function; the "nolwb_itxs" list. In some circumstances, an itx may
not be committed to an lwb (e.g. if allocating the "next" ZIL block
on disk fails), so this list is used to keep track of which itxs
fall into this state, such that their callbacks can be called after
the ZIL's writer pipeline is "stalled".
* The logic to actually call the itx's callback was moved into the
zil_itx_destroy() function. Since all consumers of zil_itx_destroy()
were effectively performing the same logic (i.e. if callback is
non-null, call the callback), it seemed like useful code cleanup to
consolidate this logic into a single function.
Additionally, the existing Linux tracepoint infrastructure dealing with
the ZIL's probes and structures had to be updated to reflect these code
changes. Specifically:
* The "zil__cw1" and "zil__cw2" probes were removed, so they had to be
removed from "trace_zil.h" as well.
* Some of the zilog structure's fields were removed, which affected
the tracepoint definitions of the structure.
* New tracepoints had to be added for the following 3 new probes:
* zil__process__commit__itx
* zil__process__normal__itx
* zil__commit__io__error
OpenZFS-issue: https://www.illumos.org/issues/8585
OpenZFS-commit: https://github.com/openzfs/openzfs/commit/5d95a3aCloses#6566
The correct way to determine if a dnode is dirty is to check
if any of the dn->dn_dirty_link's are active. Relying solely
on the dn->dn_dirtyctx can result in the dnode being mistakenly
reported as clean.
Reviewed-by: Chunwei Chen <tuxoko@gmail.com>
Reviewed-by: George Melikov <mail@gmelikov.ru>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes#3125Closes#6867
With PR 5756 the zfs module now supports c99 and the
remaining past c89 workarounds can be undone.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: George Melikov <mail@gmelikov.ru>
Signed-off-by: Don Brady <don.brady@delphix.com>
Closes#6816
When dumping objects larger than 128PiB it's possible for do_dump() to
miscalculate the FREE_RECORD offset due to an integer overflow
condition: this prevents the receiving end from correctly restoring
the dumped object.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
Signed-off-by: loli10K <ezomori.nozomu@gmail.com>
Closes#6760
This small patch fixes an issue where dmu_free_long_object_raw()
calls dnode_hold() after freeing the dnode a line above.
Reviewed-by: Jorgen Lundman <lundman@lundman.net>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Tom Caputi <tcaputi@datto.com>
Closes#6766
Several issues were uncovered by running stress tests with zfs
encryption and raw sends in particular. The issues and their
associated fixes are as follows:
* arc_read_done() has the ability to chain several requests for
the same block of data via the arc_callback_t struct. In these
cases, the ARC would only use the first request's dsobj from
the bookmark to decrypt the data. This is problematic because
the first request might be a prefetch zio which is able to
handle the key not being loaded, while the second might use a
different key that it is sure will work. The fix here is to
pass the dsobj with each individual arc_callback_t so that each
request can attempt to decrypt the data separately.
* DRR_FREE and DRR_FREEOBJECT records in a send file were not
having their transactions properly tagged as raw during raw
sends, which caused a panic when the dbuf code attempted to
decrypt these blocks.
* traverse_prefetch_metadata() did not properly set
ZIO_FLAG_SPECULATIVE when issuing prefetch IOs.
* Added a few asserts and code cleanups to ensure these issues
are more detectable in the future.
Signed-off-by: Tom Caputi <tcaputi@datto.com>
This patch fixes several issues discovered after
the encryption patch was merged:
* Fixed a bug where encrypted datasets could attempt
to receive embedded data records.
* Fixed a bug where dirty records created by the recv
code wasn't properly setting the dr_raw flag.
* Fixed a typo where a dmu_tx_commit() was changed to
dmu_tx_abort()
* Fixed a few error handling bugs unrelated to the
encryption patch in dmu_recv_stream()
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Tom Caputi <tcaputi@datto.com>
Closes#6512Closes#6524Closes#6545
Since OpenZFS 7578 (1b7c1e5) if we have a ZVOL with logbias=throughput
we will force WR_INDIRECT itxs in zvol_log_write() setting itx->itx_lr
offset and length to the offset and length of the BIO from
zvol_write()->zvol_log_write(): these offset and length are later used
to take a range lock in zillog->zl_get_data function: zvol_get_data().
Now suppose we have a ZVOL with blocksize=8K and push 4K writes to
offset 0: we will only be range-locking 0-4096. This means the
ASSERTion we make in dbuf_unoverride() is no longer valid because now
dmu_sync() is called from zilog's get_data functions holding a partial
lock on the dbuf.
Fix this by taking a range lock on the whole block in zvol_get_data().
Reviewed-by: Chunwei Chen <tuxoko@gmail.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: loli10K <ezomori.nozomu@gmail.com>
Closes#6238Closes#6315Closes#6356Closes#6477
This change incorporates three major pieces:
The first change is a keystore that manages wrapping
and encryption keys for encrypted datasets. These
commands mostly involve manipulating the new
DSL Crypto Key ZAP Objects that live in the MOS. Each
encrypted dataset has its own DSL Crypto Key that is
protected with a user's key. This level of indirection
allows users to change their keys without re-encrypting
their entire datasets. The change implements the new
subcommands "zfs load-key", "zfs unload-key" and
"zfs change-key" which allow the user to manage their
encryption keys and settings. In addition, several new
flags and properties have been added to allow dataset
creation and to make mounting and unmounting more
convenient.
The second piece of this patch provides the ability to
encrypt, decyrpt, and authenticate protected datasets.
Each object set maintains a Merkel tree of Message
Authentication Codes that protect the lower layers,
similarly to how checksums are maintained. This part
impacts the zio layer, which handles the actual
encryption and generation of MACs, as well as the ARC
and DMU, which need to be able to handle encrypted
buffers and protected data.
The last addition is the ability to do raw, encrypted
sends and receives. The idea here is to send raw
encrypted and compressed data and receive it exactly
as is on a backup system. This means that the dataset
on the receiving system is protected using the same
user key that is in use on the sending side. By doing
so, datasets can be efficiently backed up to an
untrusted system without fear of data being
compromised.
Reviewed by: Matthew Ahrens <mahrens@delphix.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Jorgen Lundman <lundman@lundman.net>
Signed-off-by: Tom Caputi <tcaputi@datto.com>
Closes#494Closes#5769
Authored by: Matthew Ahrens <mahrens@delphix.com>
Reviewed by: Prakash Surya <prakash.surya@delphix.com>
Reviewed by: George Wilson <george.wilson@delphix.com>
Approved by: Robert Mustacchi <rm@joyent.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Ported-by: Giuseppe Di Natale <dinatale2@llnl.gov>
The problem is that zfs_get_data() supplies a stale zgd_bp to
dmu_sync(), which we then nopwrite against.
zfs_get_data() doesn't hold any DMU-related locks, so after it
copies db_blkptr to zgd_bp, dbuf_write_ready() could change
db_blkptr, and dbuf_write_done() could remove the dirty record.
dmu_sync() then sees the stale BP and that the dbuf it not dirty,
so it is eligible for nop-writing.
The fix is for dmu_sync() to copy db_blkptr to zgd_bp after
acquiring the db_mtx. We could still see a stale db_blkptr,
but if it is stale then the dirty record will still exist and
thus we won't attempt to nopwrite.
OpenZFS-issue: https://www.illumos.org/issues/8378
OpenZFS-commit: https://github.com/openzfs/openzfs/commit/3127742Closes#6293
This continues what was started in
0eef1bde31 by fully converting zvols
to avoid unnecessary dnode_hold() calls. This saves a small amount
of CPU time and slightly improves latencies of operations on zvols.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Richard Yao <richard.yao@prophetstor.com>
Closes#6058
Authored by: Matthew Ahrens <mahrens@delphix.com>
Reviewed by: Dan Kimmel <dan.kimmel@delphix.com>
Reviewed by: George Wilson <george.wilson@delphix.com>
Approved by: Robert Mustacchi <rm@joyent.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Ported-by: Giuseppe Di Natale <dinatale2@llnl.gov>
When writing pre-compressed buffers, arc_write() requires that
the compression algorithm used to compress the buffer matches
the compression algorithm requested by the zio_prop_t, which is
set by dmu_write_policy(). This makes dmu_write_policy() and its
callers a bit more complicated.
We simplify this by making arc_write() trust the caller to supply
the type of pre-compressed buffer that it wants to write,
and override the compression setting in the zio_prop_t.
OpenZFS-issue: https://www.illumos.org/issues/8155
OpenZFS-commit: https://github.com/openzfs/openzfs/commit/b55ff58Closes#6200
Remove the lz4_ac local variable from dmu_write_policy() to resolve
the following unused variable warning on non-debug builds.
dmu.c: In function ‘dmu_write_policy’:
dmu.c:1892:12: warning: unused variable ‘lz4_ac’ [-Wunused-variable]
boolean_t lz4_ac = spa_feature_is_active(os->os_spa,
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
OpenZFS 7252 - compressed zfs send / receive
OpenZFS 7628 - create long versions of ZFS send / receive options
Authored by: Dan Kimmel <dan.kimmel@delphix.com>
Reviewed by: George Wilson <george.wilson@delphix.com>
Reviewed by: John Kennedy <john.kennedy@delphix.com>
Reviewed by: Matthew Ahrens <mahrens@delphix.com>
Reviewed by: Paul Dagnelie <pcd@delphix.com>
Reviewed by: Pavel Zakharov <pavel.zakharov@delphix.com>
Reviewed by: Sebastien Roy <sebastien.roy@delphix.com>
Reviewed by: David Quigley <dpquigl@davequigley.com>
Reviewed by: Thomas Caputi <tcaputi@datto.com>
Approved by: Dan McDonald <danmcd@omniti.com>
Reviewed by: David Quigley <dpquigl@davequigley.com>
Reviewed-by: loli10K <ezomori.nozomu@gmail.com>
Ported-by: bunder2015 <omfgbunder@gmail.com>
Ported-by: Don Brady <don.brady@intel.com>
Ported-by: Brian Behlendorf <behlendorf1@llnl.gov>
Porting Notes:
- Most of 7252 was already picked up during ABD work. This
commit represents the gap from the final commit to openzfs.
- Fixed split_large_blocks check in do_dump()
- An alternate version of the write_compressible() function was
implemented for Linux which does not depend on fio. The behavior
of fio differs significantly based on the exact version.
- mkholes was replaced with truncate for Linux.
OpenZFS-issue: https://www.illumos.org/issues/7252
OpenZFS-commit: https://github.com/openzfs/openzfs/commit/5602294Closes#6067
Force flushing of txg's can be painfully slow when competing for disk
IO, since this is a process meant to execute asynchronously. Optimize
this path via allowing data/hole seeking if the file is clean, but if
dirty fall back to old logic. This is a compromise to disabling the
feature entirely.
Reviewed-by: Giuseppe Di Natale <dinatale2@llnl.gov>
Reviewed-by: George Melikov <mail@gmelikov.ru>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Debabrata Banerjee <dbanerje@akamai.com>
Closes#4306Closes#5962
When the code was added this tunable was not exposed via module params. Also it
was not documented. This patch changes the type from a uint32 to a ulong as
done with other percentage tunables and also documents it in the
zfs-module-parameters man page.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: David Quigley <david.quigley@intel.com>
Closes#5750
Add *_by_dnode() routines for accessing objects given their
dnode_t *, this is more efficient than accessing the object by
(objset_t *, uint64_t object). This change converts some but
not all of the existing consumers. As performance-sensitive
code paths are discovered they should be converted to use
these routines.
Reviewed-by: Matthew Ahrens <mahrens@delphix.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Alex Zhuravlev <bzzz@whamcloud.com>
Closes#5534
Issue #4802
Reviewed by: George Wilson <george.wilson@delphix.com>
Reviewed by: Prakash Surya <prakash.surya@delphix.com>
Reviewed by: Saso Kiselkov <saso.kiselkov@nexenta.com>
Reviewed by: Richard Lowe <richlowe@richlowe.net>
Approved by: Garrett D'Amore <garrett@damore.org>
Ported by: Tony Hutter <hutter2@llnl.gov>
OpenZFS-issue: https://www.illumos.org/issues/4185
OpenZFS-commit: https://github.com/openzfs/openzfs/commit/45818ee
Porting Notes:
This code is ported on top of the Illumos Crypto Framework code:
b5e030c8db
The list of porting changes includes:
- Copied module/icp/include/sha2/sha2.h directly from illumos
- Removed from module/icp/algs/sha2/sha2.c:
#pragma inline(SHA256Init, SHA384Init, SHA512Init)
- Added 'ctx' to lib/libzfs/libzfs_sendrecv.c:zio_checksum_SHA256() since
it now takes in an extra parameter.
- Added CTASSERT() to assert.h from for module/zfs/edonr_zfs.c
- Added skein & edonr to libicp/Makefile.am
- Added sha512.S. It was generated from sha512-x86_64.pl in Illumos.
- Updated ztest.c with new fletcher_4_*() args; used NULL for new CTX argument.
- In icp/algs/edonr/edonr_byteorder.h, Removed the #if defined(__linux) section
to not #include the non-existant endian.h.
- In skein_test.c, renane NULL to 0 in "no test vector" array entries to get
around a compiler warning.
- Fixup test files:
- Rename <sys/varargs.h> -> <varargs.h>, <strings.h> -> <string.h>,
- Remove <note.h> and define NOTE() as NOP.
- Define u_longlong_t
- Rename "#!/usr/bin/ksh" -> "#!/bin/ksh -p"
- Rename NULL to 0 in "no test vector" array entries to get around a
compiler warning.
- Remove "for isa in $($ISAINFO); do" stuff
- Add/update Makefiles
- Add some userspace headers like stdio.h/stdlib.h in places of
sys/types.h.
- EXPORT_SYMBOL *_Init/*_Update/*_Final... routines in ICP modules.
- Update scripts/zfs2zol-patch.sed
- include <sys/sha2.h> in sha2_impl.h
- Add sha2.h to include/sys/Makefile.am
- Add skein and edonr dirs to icp Makefile
- Add new checksums to zpool_get.cfg
- Move checksum switch block from zfs_secpolicy_setprop() to
zfs_check_settable()
- Fix -Wuninitialized error in edonr_byteorder.h on PPC
- Fix stack frame size errors on ARM32
- Don't unroll loops in Skein on 32-bit to save stack space
- Add memory barriers in sha2.c on 32-bit to save stack space
- Add filetest_001_pos.ksh checksum sanity test
- Add option to write psudorandom data in file_write utility
Authored by: George Wilson <george.wilson@delphix.com>
Reviewed by: Prakash Surya <prakash.surya@delphix.com>
Reviewed by: Dan Kimmel <dan.kimmel@delphix.com>
Reviewed by: Matt Ahrens <mahrens@delphix.com>
Reviewed by: Paul Dagnelie <pcd@delphix.com>
Reviewed by: Tom Caputi <tcaputi@datto.com>
Reviewed by: Brian Behlendorf <behlendorf1@llnl.gov>
Ported by: David Quigley <david.quigley@intel.com>
This review covers the reading and writing of compressed arc headers, sharing
data between the arc_hdr_t and the arc_buf_t, and the implementation of a new
dbuf cache to keep frequently access data uncompressed.
I've added a new member to l1 arc hdr called b_pdata. The b_pdata always hangs
off the arc_buf_hdr_t (if an L1 hdr is in use) and points to the physical block
for that DVA. The physical block may or may not be compressed. If compressed
arc is enabled and the block on-disk is compressed, then the b_pdata will match
the block on-disk and remain compressed in memory. If the block on disk is not
compressed, then neither will the b_pdata. Lastly, if compressed arc is
disabled, then b_pdata will always be an uncompressed version of the on-disk
block.
Typically the arc will cache only the arc_buf_hdr_t and will aggressively evict
any arc_buf_t's that are no longer referenced. This means that the arc will
primarily have compressed blocks as the arc_buf_t's are considered overhead and
are always uncompressed. When a consumer reads a block we first look to see if
the arc_buf_hdr_t is cached. If the hdr is cached then we allocate a new
arc_buf_t and decompress the b_pdata contents into the arc_buf_t's b_data. If
the hdr already has a arc_buf_t, then we will allocate an additional arc_buf_t
and bcopy the uncompressed contents from the first arc_buf_t to the new one.
Writing to the compressed arc requires that we first discard the b_pdata since
the physical block is about to be rewritten. The new data contents will be
passed in via an arc_buf_t (uncompressed) and during the I/O pipeline stages we
will copy the physical block contents to a newly allocated b_pdata.
When an l2arc is inuse it will also take advantage of the b_pdata. Now the
l2arc will always write the contents of b_pdata to the l2arc. This means that
when compressed arc is enabled that the l2arc blocks are identical to those
stored in the main data pool. This provides a significant advantage since we
can leverage the bp's checksum when reading from the l2arc to determine if the
contents are valid. If the compressed arc is disabled, then we must first
transform the read block to look like the physical block in the main data pool
before comparing the checksum and determining it's valid.
OpenZFS-issue: https://www.illumos.org/issues/6950
OpenZFS-commit: https://github.com/openzfs/openzfs/commit/7fc10f0
Issue #5078
For quite some time I was thinking about possibility to prefetch
ZFS indirection tables while doing sequential reads or writes.
Recent changes in predictive prefetcher made that much easier to
do. My tests on zvol with 16KB block size on 5x striped and 2x
mirrored pool of 10 disks show almost double throughput on sequential
read, and almost tripple on sequential rewrite. While for read alike
effect can be received from increasing maximal prefetch distance
(though at higher memory cost), for rewrite there is no other
solution so far.
Authored by: Alexander Motin <mav@freebsd.org>
Reviewed by: Matthew Ahrens <mahrens@delphix.com>
Reviewed by: Paul Dagnelie <pcd@delphix.com>
Approved by: Robert Mustacchi <rm@joyent.com>
Ported-by: kernelOfTruth kerneloftruth@gmail.com
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
OpenZFS-issue: https://www.illumos.org/issues/6322
OpenZFS-commit: https://github.com/illumos/illumos-gate/commit/cb92f413Closes#5040
Porting notes:
- Change from upstream in module/zfs/dbuf.c in 'int dbuf_read' due
to commit 5f6d0b6 'Handle block pointers with a corrupt logical size'
- Difference from upstream in module/zfs/dmu_zfetch.c,
uint32_t zfetch_max_idistance -> unsigned int zfetch_max_idistance
- Variables have been initialized at the beginning of the function
(void dmu_zfetch) to resemble the order of occurrence and account
for C99, C11 mode errors.
Using a benchmark which has 32 threads creating 2 million files in the
same directory, on a machine with 16 CPU cores, I observed poor
performance. I noticed that dmu_tx_hold_zap() was using about 30% of
all CPU, and doing dnode_hold() 7 times on the same object (the ZAP
object that is being held).
dmu_tx_hold_zap() keeps a hold on the dnode_t the entire time it is
running, in dmu_tx_hold_t:txh_dnode, so it would be nice to use the
dnode_t that we already have in hand, rather than repeatedly calling
dnode_hold(). To do this, we need to pass the dnode_t down through
all the intermediate calls that dmu_tx_hold_zap() makes, making these
routines take the dnode_t* rather than an objset_t* and a uint64_t
object number. In particular, the following routines will need to have
analogous *_by_dnode() variants created:
dmu_buf_hold_noread()
dmu_buf_hold()
zap_lookup()
zap_lookup_norm()
zap_count_write()
zap_lockdir()
zap_count_write()
This can improve performance on the benchmark described above by 100%,
from 30,000 file creations per second to 60,000. (This improvement is on
top of that provided by working around the object allocation issue. Peak
performance of ~90,000 creations per second was observed with 8 CPUs;
adding CPUs past that decreased performance due to lock contention.) The
CPU used by dmu_tx_hold_zap() was reduced by 88%, from 340 CPU-seconds
to 40 CPU-seconds.
Sponsored by: Intel Corp.
Signed-off-by: Matthew Ahrens <mahrens@delphix.com>
Signed-off-by: Ned Bass <bass6@llnl.gov>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
OpenZFS-issue: https://www.illumos.org/issues/7004
OpenZFS-commit: https://github.com/openzfs/openzfs/pull/109Closes#4641Closes#4972
Justification
-------------
This feature adds support for variable length dnodes. Our motivation is
to eliminate the overhead associated with using spill blocks. Spill
blocks are used to store system attribute data (i.e. file metadata) that
does not fit in the dnode's bonus buffer. By allowing a larger bonus
buffer area the use of a spill block can be avoided. Spill blocks
potentially incur an additional read I/O for every dnode in a dnode
block. As a worst case example, reading 32 dnodes from a 16k dnode block
and all of the spill blocks could issue 33 separate reads. Now suppose
those dnodes have size 1024 and therefore don't need spill blocks. Then
the worst case number of blocks read is reduced to from 33 to two--one
per dnode block. In practice spill blocks may tend to be co-located on
disk with the dnode blocks so the reduction in I/O would not be this
drastic. In a badly fragmented pool, however, the improvement could be
significant.
ZFS-on-Linux systems that make heavy use of extended attributes would
benefit from this feature. In particular, ZFS-on-Linux supports the
xattr=sa dataset property which allows file extended attribute data
to be stored in the dnode bonus buffer as an alternative to the
traditional directory-based format. Workloads such as SELinux and the
Lustre distributed filesystem often store enough xattr data to force
spill bocks when xattr=sa is in effect. Large dnodes may therefore
provide a performance benefit to such systems.
Other use cases that may benefit from this feature include files with
large ACLs and symbolic links with long target names. Furthermore,
this feature may be desirable on other platforms in case future
applications or features are developed that could make use of a
larger bonus buffer area.
Implementation
--------------
The size of a dnode may be a multiple of 512 bytes up to the size of
a dnode block (currently 16384 bytes). A dn_extra_slots field was
added to the current on-disk dnode_phys_t structure to describe the
size of the physical dnode on disk. The 8 bits for this field were
taken from the zero filled dn_pad2 field. The field represents how
many "extra" dnode_phys_t slots a dnode consumes in its dnode block.
This convention results in a value of 0 for 512 byte dnodes which
preserves on-disk format compatibility with older software.
Similarly, the in-memory dnode_t structure has a new dn_num_slots field
to represent the total number of dnode_phys_t slots consumed on disk.
Thus dn->dn_num_slots is 1 greater than the corresponding
dnp->dn_extra_slots. This difference in convention was adopted
because, unlike on-disk structures, backward compatibility is not a
concern for in-memory objects, so we used a more natural way to
represent size for a dnode_t.
The default size for newly created dnodes is determined by the value of
a new "dnodesize" dataset property. By default the property is set to
"legacy" which is compatible with older software. Setting the property
to "auto" will allow the filesystem to choose the most suitable dnode
size. Currently this just sets the default dnode size to 1k, but future
code improvements could dynamically choose a size based on observed
workload patterns. Dnodes of varying sizes can coexist within the same
dataset and even within the same dnode block. For example, to enable
automatically-sized dnodes, run
# zfs set dnodesize=auto tank/fish
The user can also specify literal values for the dnodesize property.
These are currently limited to powers of two from 1k to 16k. The
power-of-2 limitation is only for simplicity of the user interface.
Internally the implementation can handle any multiple of 512 up to 16k,
and consumers of the DMU API can specify any legal dnode value.
The size of a new dnode is determined at object allocation time and
stored as a new field in the znode in-memory structure. New DMU
interfaces are added to allow the consumer to specify the dnode size
that a newly allocated object should use. Existing interfaces are
unchanged to avoid having to update every call site and to preserve
compatibility with external consumers such as Lustre. The new
interfaces names are given below. The versions of these functions that
don't take a dnodesize parameter now just call the _dnsize() versions
with a dnodesize of 0, which means use the legacy dnode size.
New DMU interfaces:
dmu_object_alloc_dnsize()
dmu_object_claim_dnsize()
dmu_object_reclaim_dnsize()
New ZAP interfaces:
zap_create_dnsize()
zap_create_norm_dnsize()
zap_create_flags_dnsize()
zap_create_claim_norm_dnsize()
zap_create_link_dnsize()
The constant DN_MAX_BONUSLEN is renamed to DN_OLD_MAX_BONUSLEN. The
spa_maxdnodesize() function should be used to determine the maximum
bonus length for a pool.
These are a few noteworthy changes to key functions:
* The prototype for dnode_hold_impl() now takes a "slots" parameter.
When the DNODE_MUST_BE_FREE flag is set, this parameter is used to
ensure the hole at the specified object offset is large enough to
hold the dnode being created. The slots parameter is also used
to ensure a dnode does not span multiple dnode blocks. In both of
these cases, if a failure occurs, ENOSPC is returned. Keep in mind,
these failure cases are only possible when using DNODE_MUST_BE_FREE.
If the DNODE_MUST_BE_ALLOCATED flag is set, "slots" must be 0.
dnode_hold_impl() will check if the requested dnode is already
consumed as an extra dnode slot by an large dnode, in which case
it returns ENOENT.
* The function dmu_object_alloc() advances to the next dnode block
if dnode_hold_impl() returns an error for a requested object.
This is because the beginning of the next dnode block is the only
location it can safely assume to either be a hole or a valid
starting point for a dnode.
* dnode_next_offset_level() and other functions that iterate
through dnode blocks may no longer use a simple array indexing
scheme. These now use the current dnode's dn_num_slots field to
advance to the next dnode in the block. This is to ensure we
properly skip the current dnode's bonus area and don't interpret it
as a valid dnode.
zdb
---
The zdb command was updated to display a dnode's size under the
"dnsize" column when the object is dumped.
For ZIL create log records, zdb will now display the slot count for
the object.
ztest
-----
Ztest chooses a random dnodesize for every newly created object. The
random distribution is more heavily weighted toward small dnodes to
better simulate real-world datasets.
Unused bonus buffer space is filled with non-zero values computed from
the object number, dataset id, offset, and generation number. This
helps ensure that the dnode traversal code properly skips the interior
regions of large dnodes, and that these interior regions are not
overwritten by data belonging to other dnodes. A new test visits each
object in a dataset. It verifies that the actual dnode size matches what
was stored in the ztest block tag when it was created. It also verifies
that the unused bonus buffer space is filled with the expected data
patterns.
ZFS Test Suite
--------------
Added six new large dnode-specific tests, and integrated the dnodesize
property into existing tests for zfs allow and send/recv.
Send/Receive
------------
ZFS send streams for datasets containing large dnodes cannot be received
on pools that don't support the large_dnode feature. A send stream with
large dnodes sets a DMU_BACKUP_FEATURE_LARGE_DNODE flag which will be
unrecognized by an incompatible receiving pool so that the zfs receive
will fail gracefully.
While not implemented here, it may be possible to generate a
backward-compatible send stream from a dataset containing large
dnodes. The implementation may be tricky, however, because the send
object record for a large dnode would need to be resized to a 512
byte dnode, possibly kicking in a spill block in the process. This
means we would need to construct a new SA layout and possibly
register it in the SA layout object. The SA layout is normally just
sent as an ordinary object record. But if we are constructing new
layouts while generating the send stream we'd have to build the SA
layout object dynamically and send it at the end of the stream.
For sending and receiving between pools that do support large dnodes,
the drr_object send record type is extended with a new field to store
the dnode slot count. This field was repurposed from unused padding
in the structure.
ZIL Replay
----------
The dnode slot count is stored in the uppermost 8 bits of the lr_foid
field. The bits were unused as the object id is currently capped at
48 bits.
Resizing Dnodes
---------------
It should be possible to resize a dnode when it is dirtied if the
current dnodesize dataset property differs from the dnode's size, but
this functionality is not currently implemented. Clearly a dnode can
only grow if there are sufficient contiguous unused slots in the
dnode block, but it should always be possible to shrink a dnode.
Growing dnodes may be useful to reduce fragmentation in a pool with
many spill blocks in use. Shrinking dnodes may be useful to allow
sending a dataset to a pool that doesn't support the large_dnode
feature.
Feature Reference Counting
--------------------------
The reference count for the large_dnode pool feature tracks the
number of datasets that have ever contained a dnode of size larger
than 512 bytes. The first time a large dnode is created in a dataset
the dataset is converted to an extensible dataset. This is a one-way
operation and the only way to decrement the feature count is to
destroy the dataset, even if the dataset no longer contains any large
dnodes. The complexity of reference counting on a per-dnode basis was
too high, so we chose to track it on a per-dataset basis similarly to
the large_block feature.
Signed-off-by: Ned Bass <bass6@llnl.gov>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes#3542
