When performing concurrent object allocations using the new
multi-threaded allocator and large dnodes it's possible to
allocate overlapping large dnodes.
This case should have been handled by detecting an error
returned by dnode_hold_impl(). But that logic only checked
the returned dnp was not-NULL, and the dnp variable was not
reset to NULL when retrying. Resolve this issue by properly
checking the return value of dnode_hold_impl().
Additionally, it was possible that dnode_hold_impl() would
misreport a dnode as free when it was in fact in use. This
could occurs for two reasons:
* The per-slot zrl_lock must be held over the entire critical
section which includes the alloc/free until the new dnode
is assigned to children_dnodes. Additionally, all of the
zrl_lock's in the range must be held to protect moving
dnodes.
* The dn->dn_ot_type cannot be solely relied upon to check
the type. When allocating a new dnode its type will be
DMU_OT_NONE after dnode_create(). Only latter when
dnode_allocate() is called will it transition to the new
type. This means there's a window when allocating where
it can mistaken for a free dnode.
Reviewed-by: Giuseppe Di Natale <dinatale2@llnl.gov>
Reviewed-by: Ned Bass <bass6@llnl.gov>
Reviewed-by: Tony Hutter <hutter2@llnl.gov>
Reviewed-by: Olaf Faaland <faaland1@llnl.gov>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes#6414Closes#6439
In zfs/dmu_object and icp/core/kcf_sched, the CPU_SEQID macro
should be surrounded by `kpreempt_disable` and `kpreempt_enable`
calls to avoid a Linux kernel BUG warning. These code paths use
the cpuid to minimize lock contention and is is safe to reschedule
the process to a different processor at any time.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Morgan Jones <me@numin.it>
Closes#6239
dmu_object_alloc() is single-threaded, so when multiple threads are
creating files in a single filesystem, they spend a lot of time waiting
for the os_obj_lock. To improve performance of multi-threaded file
creation, we must make dmu_object_alloc() typically not grab any
filesystem-wide locks.
The solution is to have a "next object to allocate" for each CPU. Each
of these "next object"s is in a different block of the dnode object, so
that concurrent allocation holds dnodes in different dbufs. When a
thread's "next object" reaches the end of a chunk of objects (by default
4 blocks worth -- 128 dnodes), it will be reset to the per-objset
os_obj_next, which will be increased by a chunk of objects (128). Only
when manipulating the os_obj_next will we need to grab the os_obj_lock.
This decreases lock contention dramatically, because each thread only
needs to grab the os_obj_lock briefly, once per 128 allocations.
This results in a 70% performance improvement to multi-threaded object
creation (where each thread is creating objects in its own directory),
from 67,000/sec to 115,000/sec, with 8 CPUs.
Work sponsored by Intel Corp.
Authored by: Matthew Ahrens <mahrens@delphix.com>
Reviewed-by: Ned Bass <bass6@llnl.gov>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Ported-by: Matthew Ahrens <mahrens@delphix.com>
Signed-off-by: Matthew Ahrens <mahrens@delphix.com>
OpenZFS-issue: https://www.illumos.org/issues/8199
OpenZFS-commit: https://github.com/openzfs/openzfs/pull/374Closes#4703Closes#6117
0eef1bde31
introduced some changes which we slightly improved the style of when
porting to illumos.
There is also one minor error-handling fix, in zap_add() the "zap" may
become NULL in case of an error re-opening the ZAP.
Originally suggested at: https://github.com/openzfs/openzfs/pull/276
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed by: Pavel Zakharov <pavel.zakharov@delphix.com>
Signed-off-by: Matthew Ahrens <mahrens@delphix.com>
Closes#5805
Fix dmu_object_next() to correctly handle unallocated objects on
large_dnode datasets.
We implement this by scanning the dnode block until we find the correct
offset to be used in dnode_next_offset(). This is necessary because we
can't assume *objectp is a hole even if dmu_object_info() returns
ENOENT.
This fixes a couple of issues with zfs receive on large_dnode datasets.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Ned Bass <bass6@llnl.gov>
Signed-off-by: loli10K <ezomori.nozomu@gmail.com>
Closes#5027Closes#5532
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
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
Only attempt to backfill lower metadnode object numbers if at least
4096 objects have been freed since the last rescan, and at most once
per transaction group. This avoids a pathology in dmu_object_alloc()
that caused O(N^2) behavior for create-heavy workloads and
substantially improves object creation rates. As summarized by
@mahrens in #4636:
"Normally, the object allocator simply checks to see if the next
object is available. The slow calls happened when dmu_object_alloc()
checks to see if it can backfill lower object numbers. This happens
every time we move on to a new L1 indirect block (i.e. every 32 *
128 = 4096 objects). When re-checking lower object numbers, we use
the on-disk fill count (blkptr_t:blk_fill) to quickly skip over
indirect blocks that don’t have enough free dnodes (defined as an L2
with at least 393,216 of 524,288 dnodes free). Therefore, we may
find that a block of dnodes has a low (or zero) fill count, and yet
we can’t allocate any of its dnodes, because they've been allocated
in memory but not yet written to disk. In this case we have to hold
each of the dnodes and then notice that it has been allocated in
memory.
The end result is that allocating N objects in the same TXG can
require CPU usage proportional to N^2."
Add a tunable dmu_rescan_dnode_threshold to define the number of
objects that must be freed before a rescan is performed. Don't bother
to export this as a module option because testing doesn't show a
compelling reason to change it. The vast majority of the performance
gain comes from limit the rescan to at most once per TXG.
Signed-off-by: Ned Bass <bass6@llnl.gov>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
6370 ZFS send fails to transmit some holes
Reviewed by: Matthew Ahrens <mahrens@delphix.com>
Reviewed by: Chris Williamson <chris.williamson@delphix.com>
Reviewed by: Stefan Ring <stefanrin@gmail.com>
Reviewed by: Steven Burgess <sburgess@datto.com>
Reviewed by: Arne Jansen <sensille@gmx.net>
Approved by: Robert Mustacchi <rm@joyent.com>
References:
https://www.illumos.org/issues/6370https://github.com/illumos/illumos-gate/commit/286ef71
In certain circumstances, "zfs send -i" (incremental send) can produce
a stream which will result in incorrect sparse file contents on the
target.
The problem manifests as regions of the received file that should be
sparse (and read a zero-filled) actually contain data from a file that
was deleted (and which happened to share this file's object ID).
Note: this can happen only with filesystems (not zvols, because they do
not free (and thus can not reuse) object IDs).
Note: This can happen only if, since the incremental source (FromSnap),
a file was deleted and then another file was created, and the new file
is sparse (i.e. has areas that were never written to and should be
implicitly zero-filled).
We suspect that this was introduced by 4370 (applies only if hole_birth
feature is enabled), and made worse by 5243 (applies if hole_birth
feature is disabled, and we never send any holes).
The bug is caused by the hole birth feature. When an object is deleted
and replaced, all the holes in the object have birth time zero. However,
zfs send cannot tell that the holes are new since the file was replaced,
so it doesn't send them in an incremental. As a result, you can end up
with invalid data when you receive incremental send streams. As a
short-term fix, we can always send holes with birth time 0 (unless it's
a zvol or a dataset where we can guarantee that no objects have been
reused).
Ported-by: Steven Burgess <sburgess@datto.com>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes#4369Closes#4050
Restore_object should not use two transactions to restore an object:
* one transaction is used for dmu_object_claim
* another transaction is used to set compression, checksum and most
importantly bonus data
* furthermore dmu_object_reclaim internally uses multiple transactions
* dmu_free_long_range frees chunks in separate transactions
* dnode_reallocate is executed in a distinct transaction
The fact the dnode_allocate/dnode_reallocate are executed in one
transaction and bonus (re-)population is executed in a different
transaction may lead to violation of ZFS consistency assertions if the
transactions are assigned to different transaction groups. Also, if
the first transaction group is successfully written to a permanent
storage, but the second transaction is lost, then an invalid dnode may
be created on the stable storage.
3693 restore_object uses at least two transactions to restore an object
Reviewed by: Christopher Siden <christopher.siden@delphix.com>
Reviewed by: George Wilson <george.wilson@delphix.com>
Reviewed by: Andriy Gapon <andriy.gapon@hybridcluster.com>
Approved by: Robert Mustacchi <rm@joyent.com>
Original authors: Matthew Ahrens and Andriy Gapon
References:
https://www.illumos.org/issues/3693https://github.com/illumos/illumos-gate/commit/e77d42e
Ported by: Turbo Fredriksson <turbo@bayour.com>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes#2689
4171 clean up spa_feature_*() interfaces
4172 implement extensible_dataset feature for use by other zpool features
Reviewed by: Max Grossman <max.grossman@delphix.com>
Reviewed by: Christopher Siden <christopher.siden@delphix.com>
Reviewed by: George Wilson <george.wilson@delphix.com>
Reviewed by: Jerry Jelinek <jerry.jelinek@joyent.com>
Approved by: Garrett D'Amore <garrett@damore.org>a
References:
https://www.illumos.org/issues/4171https://www.illumos.org/issues/4172https://github.com/illumos/illumos-gate/commit/2acef22
Ported-by: Tim Chase <tim@chase2k.com>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes#2528
3598 want to dtrace when errors are generated in zfs
Reviewed by: Dan Kimmel <dan.kimmel@delphix.com>
Reviewed by: Adam Leventhal <ahl@delphix.com>
Reviewed by: Christopher Siden <christopher.siden@delphix.com>
Approved by: Garrett D'Amore <garrett@damore.org>
References:
https://www.illumos.org/issues/3598illumos/illumos-gate@be6fd75a69
Ported-by: Richard Yao <ryao@gentoo.org>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Issue #1775
Porting notes:
1. include/sys/zfs_context.h has been modified to render some new
macros inert until dtrace is available on Linux.
2. Linux-specific changes have been adapted to use SET_ERROR().
3. I'm NOT happy about this change. It does nothing but ugly
up the code under Linux. Unfortunately we need to take it to
avoid more merge conflicts in the future. -Brian
