= Motivation
We've noticed several zloop crashes within Delphix generated
due to the following sequence of events:
- A device gets expanded and new metaslabas are allocated for
it. These metaslabs go through `metaslab_init()` but haven't
gone through `metaslab_sync_done()` yet. This meas that the
only range tree that's actually set is the `ms_allocatable`.
All the others are NULL.
- A vdev_initialization is issues and `vdev_initialize_thread`
starts processing one of these new metaslabs of the expanded
vdev.
- As part of `vdev_initialize_calculate_progress()` we call
into `metaslab_load()` and `metaslab_load_impl()` which
in turn tries to dereference the metaslabs trees that
are still NULL and therefore we crash.
The same failure can come up from the `vdev_trim` code paths.
= This Patch
We considered the following solutions to deal with this issue:
[A] Add logic to `vdev_initialize/trim` to skip those new
metaslabs. We decided against this as it would be good
to avoid exposing this lower-level detail to higer-level
operations.
[B] Have `metaslab_load_impl()` return early for new metaslabs
and thus never touch those range_trees that are NULL at
that time. This seemed more of a work-around for the bug
and not a clear-cut solution.
[C] Refactor our logic so all metaslabs have their range_trees
created at the time of their creatin in `metaslab_init()`.
In this patch we decided to go with [C] because:
(1) It doesn't expose more metaslab details to higher level
operations such as vdev initialize and trim.
(2) The current behavior of creating the range trees lazily
in `metaslab_sync_done()` is unnecessarily complicated.
(3) Always initializing the metaslab range_trees makes other
parts of the codebase cleaner. For example, we used to
use `ms_freed` as the reference value for knowing whether
all the range_trees have been initialized. Now we no
longer need to do that check in most places (and in the
few that we do we use the `ms_new` boolean field now
which is more readable).
= Side Changes
Probably due to a mismerge we set `ms_loaded` to `B_TRUE` twice
in `metasloab_load_impl()`. In this patch we remove the extraneous
assignment.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Matthew Ahrens <mahrens@delphix.com>
Signed-off-by: Serapheim Dimitropoulos <serapheim@delphix.com>
Closes#11737
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
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
This patch adds a new top-level vdev type called dRAID, which stands
for Distributed parity RAID. This pool configuration allows all dRAID
vdevs to participate when rebuilding to a distributed hot spare device.
This can substantially reduce the total time required to restore full
parity to pool with a failed device.
A dRAID pool can be created using the new top-level `draid` type.
Like `raidz`, the desired redundancy is specified after the type:
`draid[1,2,3]`. No additional information is required to create the
pool and reasonable default values will be chosen based on the number
of child vdevs in the dRAID vdev.
zpool create <pool> draid[1,2,3] <vdevs...>
Unlike raidz, additional optional dRAID configuration values can be
provided as part of the draid type as colon separated values. This
allows administrators to fully specify a layout for either performance
or capacity reasons. The supported options include:
zpool create <pool> \
draid[<parity>][:<data>d][:<children>c][:<spares>s] \
<vdevs...>
- draid[parity] - Parity level (default 1)
- draid[:<data>d] - Data devices per group (default 8)
- draid[:<children>c] - Expected number of child vdevs
- draid[:<spares>s] - Distributed hot spares (default 0)
Abbreviated example `zpool status` output for a 68 disk dRAID pool
with two distributed spares using special allocation classes.
```
pool: tank
state: ONLINE
config:
NAME STATE READ WRITE CKSUM
slag7 ONLINE 0 0 0
draid2:8d:68c:2s-0 ONLINE 0 0 0
L0 ONLINE 0 0 0
L1 ONLINE 0 0 0
...
U25 ONLINE 0 0 0
U26 ONLINE 0 0 0
spare-53 ONLINE 0 0 0
U27 ONLINE 0 0 0
draid2-0-0 ONLINE 0 0 0
U28 ONLINE 0 0 0
U29 ONLINE 0 0 0
...
U42 ONLINE 0 0 0
U43 ONLINE 0 0 0
special
mirror-1 ONLINE 0 0 0
L5 ONLINE 0 0 0
U5 ONLINE 0 0 0
mirror-2 ONLINE 0 0 0
L6 ONLINE 0 0 0
U6 ONLINE 0 0 0
spares
draid2-0-0 INUSE currently in use
draid2-0-1 AVAIL
```
When adding test coverage for the new dRAID vdev type the following
options were added to the ztest command. These options are leverages
by zloop.sh to test a wide range of dRAID configurations.
-K draid|raidz|random - kind of RAID to test
-D <value> - dRAID data drives per group
-S <value> - dRAID distributed hot spares
-R <value> - RAID parity (raidz or dRAID)
The zpool_create, zpool_import, redundancy, replacement and fault
test groups have all been updated provide test coverage for the
dRAID feature.
Co-authored-by: Isaac Huang <he.huang@intel.com>
Co-authored-by: Mark Maybee <mmaybee@cray.com>
Co-authored-by: Don Brady <don.brady@delphix.com>
Co-authored-by: Matthew Ahrens <mahrens@delphix.com>
Co-authored-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Mark Maybee <mmaybee@cray.com>
Reviewed-by: Matt Ahrens <matt@delphix.com>
Reviewed-by: Tony Hutter <hutter2@llnl.gov>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes#10102
Commit d4a72f2 which introduced multi-phase scrubs and resilvers
continued the work presented by Nexenta at the 2016 ZFS developer
summit. Update the source to reflect their contribution.
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Metaslabs are now (usually) loaded and unloaded infrequently, but when
that is not the case, it is useful to have a log of when and why these
events happened.
This commit enables the zfs_dbgmsg() in metaslab_load(), and adds a
zfs_dbgmsg() in metaslab_unload().
Reviewed-by: Serapheim Dimitropoulos <serapheim@delphix.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Matthew Ahrens <mahrens@delphix.com>
Closes#10683
Livelists and spacemaps are data structures that are logs of allocations
and frees. Livelists entries are block pointers (blkptr_t). Spacemaps
entries are ranges of numbers, most often used as to track
allocated/freed regions of metaslabs/vdevs.
These data structures can become self-inconsistent, for example if a
block or range can be "double allocated" (two allocation records without
an intervening free) or "double freed" (two free records without an
intervening allocation).
ZDB (as well as zfs running in the kernel) can detect these
inconsistencies when loading livelists and metaslab. However, it
generally halts processing when the error is detected.
When analyzing an on-disk problem, we often want to know the entire set
of inconsistencies, which is not possible with the current behavior.
This commit adds a new flag, `zdb -y`, which analyzes the livelist and
metaslab data structures and displays all of their inconsistencies.
Note that this is different from the leak detection performed by
`zdb -b`, which checks for inconsistencies between the spacemaps and the
tree of block pointers, but assumes the spacemaps are self-consistent.
The specific checks added are:
Verify livelists by iterating through each sublivelists and:
- report leftover FREEs
- report double ALLOCs and double FREEs
- record leftover ALLOCs together with their TXG [see Cross Check]
Verify spacemaps by iterating over each metaslab and:
- iterate over spacemap and then the metaslab's entries in the
spacemap log, then report any double FREEs and double ALLOCs
Verify that livelists are consistenet with spacemaps. The space
referenced by livelists (after using the FREE's to cancel out
corresponding ALLOCs) should be allocated, according to the spacemaps.
Reviewed-by: Serapheim Dimitropoulos <serapheim@delphix.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Co-authored-by: Sara Hartse <sara.hartse@delphix.com>
Signed-off-by: Matthew Ahrens <mahrens@delphix.com>
External-issue: DLPX-66031
Closes#10515
Mark functions used only in the same translation unit as static. This
only includes functions that do not have a prototype in a header file
either.
Reviewed-by: Ryan Moeller <ryan@iXsystems.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Arvind Sankar <nivedita@alum.mit.edu>
Closes#10470
Each metaslab group (of which there is one per top-level vdev) has
several (4, by default) "metaslab group allocators". Each "allocator"
has its own metaslab that it prefers to allocate from (the "primary"
allocator), and each can perform allocations concurrently with the other
allocators. In addition to the primary metaslab, there are several
other fields that need to be tracked separately for each allocator.
These are currently stored as several arrays in the metaslab_group_t,
each array indexed by allocator number.
This change organizes all the metaslab-group-allocator-specific fields
into a new struct, metaslab_group_allocator_t. The metaslab_group_t now
needs only one array indexed by the allocator number - which contains
the metaslab_group_allocator_t's.
Reviewed-by: Paul Dagnelie <pcd@delphix.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Matthew Ahrens <mahrens@delphix.com>
Closes#10213
Remove the ASSERTV macro and handle suppressing unused
compiler warnings for variables only in ASSERTs using the
__attribute__((unused)) compiler annotation. The annotation
is understood by both gcc and clang.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Jorgen Lundman <lundman@lundman.net>
Signed-off-by: Matt Macy <mmacy@FreeBSD.org>
Closes#9671
This patch implements a new tree structure for ZFS, and uses it to
store range trees more efficiently.
The new structure is approximately a B-tree, though there are some
small differences from the usual characterizations. The tree has core
nodes and leaf nodes; each contain data elements, which the elements
in the core nodes acting as separators between its children. The
difference between core and leaf nodes is that the core nodes have an
array of children, while leaf nodes don't. Every node in the tree may
be only partially full; in most cases, they are all at least 50% full
(in terms of element count) except for the root node, which can be
less full. Underfull nodes will steal from their neighbors or merge to
remain full enough, while overfull nodes will split in two. The data
elements are contained in tree-controlled buffers; they are copied
into these on insertion, and overwritten on deletion. This means that
the elements are not independently allocated, which reduces overhead,
but also means they can't be shared between trees (and also that
pointers to them are only valid until a side-effectful tree operation
occurs). The overhead varies based on how dense the tree is, but is
usually on the order of about 50% of the element size; the per-node
overheads are very small, and so don't make a significant difference.
The trees can accept arbitrary records; they accept a size and a
comparator to allow them to be used for a variety of purposes.
The new trees replace the AVL trees used in the range trees today.
Currently, the range_seg_t structure contains three 8 byte integers
of payload and two 24 byte avl_tree_node_ts to handle its storage in
both an offset-sorted tree and a size-sorted tree (total size: 64
bytes). In the new model, the range seg structures are usually two 4
byte integers, but a separate one needs to exist for the size-sorted
and offset-sorted tree. Between the raw size, the 50% overhead, and
the double storage, the new btrees are expected to use 8*1.5*2 = 24
bytes per record, or 33.3% as much memory as the AVL trees (this is
for the purposes of storing metaslab range trees; for other purposes,
like scrubs, they use ~50% as much memory).
We reduced the size of the payload in the range segments by teaching
range trees about starting offsets and shifts; since metaslabs have a
fixed starting offset, and they all operate in terms of disk sectors,
we can store the ranges using 4-byte integers as long as the size of
the metaslab divided by the sector size is less than 2^32. For 512-byte
sectors, this is a 2^41 (or 2TB) metaslab, which with the default
settings corresponds to a 256PB disk. 4k sector disks can handle
metaslabs up to 2^46 bytes, or 2^63 byte disks. Since we do not
anticipate disks of this size in the near future, there should be
almost no cases where metaslabs need 64-byte integers to store their
ranges. We do still have the capability to store 64-byte integer ranges
to account for cases where we are storing per-vdev (or per-dnode) trees,
which could reasonably go above the limits discussed. We also do not
store fill information in the compact version of the node, since it
is only used for sorted scrub.
We also optimized the metaslab loading process in various other ways
to offset some inefficiencies in the btree model. While individual
operations (find, insert, remove_from) are faster for the btree than
they are for the avl tree, remove usually requires a find operation,
while in the AVL tree model the element itself suffices. Some clever
changes actually caused an overall speedup in metaslab loading; we use
approximately 40% less cpu to load metaslabs in our tests on Illumos.
Another memory and performance optimization was achieved by changing
what is stored in the size-sorted trees. When a disk is heavily
fragmented, the df algorithm used by default in ZFS will almost always
find a number of small regions in its initial cursor-based search; it
will usually only fall back to the size-sorted tree to find larger
regions. If we increase the size of the cursor-based search slightly,
and don't store segments that are smaller than a tunable size floor
in the size-sorted tree, we can further cut memory usage down to
below 20% of what the AVL trees store. This also results in further
reductions in CPU time spent loading metaslabs.
The 16KiB size floor was chosen because it results in substantial memory
usage reduction while not usually resulting in situations where we can't
find an appropriate chunk with the cursor and are forced to use an
oversized chunk from the size-sorted tree. In addition, even if we do
have to use an oversized chunk from the size-sorted tree, the chunk
would be too small to use for ZIL allocations, so it isn't as big of a
loss as it might otherwise be. And often, more small allocations will
follow the initial one, and the cursor search will now find the
remainder of the chunk we didn't use all of and use it for subsequent
allocations. Practical testing has shown little or no change in
fragmentation as a result of this change.
If the size-sorted tree becomes empty while the offset sorted one still
has entries, it will load all the entries from the offset sorted tree
and disregard the size floor until it is unloaded again. This operation
occurs rarely with the default setting, only on incredibly thoroughly
fragmented pools.
There are some other small changes to zdb to teach it to handle btrees,
but nothing major.
Reviewed-by: George Wilson <gwilson@delphix.com>
Reviewed-by: Matt Ahrens <matt@delphix.com>
Reviewed by: Sebastien Roy seb@delphix.com
Reviewed-by: Igor Kozhukhov <igor@dilos.org>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Paul Dagnelie <pcd@delphix.com>
Closes#9181
`metaslab_verify_weight_and_frag()` a verification function and
by the end of it there shouldn't be any side-effects.
The function calls `metaslab_weight()` which in turn calls
`metaslab_set_fragmentation()`. The latter can dirty and otherwise
not dirty metaslab fro the next TXGand set `metaslab_condense_wanted`
if the spacemaps were just upgraded (meaning we just enabled the
SPACEMAP_HISTOGRAM feature through upgrade).
This patch adds a new flag as a parameter to `metaslab_weight()` and
`metaslab_set_fragmentation()` making the dirtying of the metaslab
optional.
Reviewed-by: Matt Ahrens <matt@delphix.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Serapheim Dimitropoulos <serapheim@delphix.com>
Closes#9185Closes#9282
If a pool enables the SPACEMAP_HISTOGRAM feature shortly before being
exported, we can enter a situation that causes a kernel panic. Any metaslabs
that are loaded during the final dirty txg and haven't already been condensed
will cause metaslab_sync to proceed after the final dirty txg so that the
condense can be performed, which there are assertions to prevent. Because of
the nature of this issue, there are a number of ways we can enter this
state. Rather than try to prevent each of them one by one, potentially missing
some edge cases, we instead cut it off at the point of intersection; by
preventing metaslab_sync from proceeding if it would only do so to perform a
condense and we're past the final dirty txg, we preserve the utility of the
existing asserts while preventing this particular issue.
Reviewed-by: Matt Ahrens <matt@delphix.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Paul Dagnelie <pcd@delphix.com>
Closes#9185Closes#9186Closes#9231Closes#9253
With the other metaslab changes loaded onto a system, we can
significantly reduce the memory usage of each loaded metaslab and
unload them on demand if there is memory pressure. However, none
of those changes actually result in us keeping more metaslabs loaded.
If we don't keep more metaslabs loaded, we will still have to wait
for demand-loading to finish when no loaded metaslab can satisfy our
allocation, which can cause ZIL performance issues. In addition,
performance is traditionally measured by IOs per unit time, while
unloading is currently done on a txg-count basis. Txgs can take a
widely varying range of times, from tenths of a second to several
seconds. This can result in confusing, hard to predict behavior.
This change simply adds a time-based component to metaslab unloading.
A metaslab will remain loaded for one minute and 8 txgs (by default)
after it was last used, unless it is evicted due to memory pressure.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Matt Ahrens <mahrens@delphix.com>
Signed-off-by: Paul Dagnelie <pcd@delphix.com>
External-issue: DLPX-65016
External-issue: DLPX-65047
Closes#9197
On systems with large amounts of storage and high fragmentation, a huge
amount of space can be used by storing metaslab range trees. Since
metaslabs are only unloaded during a txg sync, and only if they have
been inactive for 8 txgs, it is possible to get into a state where all
of the system's memory is consumed by range trees and metaslabs, and
txgs cannot sync. While ZFS knows how to evict ARC data when needed,
it has no such mechanism for range tree data. This can result in boot
hangs for some system configurations.
First, we add the ability to unload metaslabs outside of syncing
context. Second, we store a multilist of all loaded metaslabs, sorted
by their selection txg, so we can quickly identify the oldest
metaslabs. We use a multilist to reduce lock contention during heavy
write workloads. Finally, we add logic that will unload a metaslab
when we're loading a new metaslab, if we're using more than a certain
fraction of the available memory on range trees.
Reviewed-by: Matt Ahrens <mahrens@delphix.com>
Reviewed-by: George Wilson <gwilson@delphix.com>
Reviewed-by: Sebastien Roy <sebastien.roy@delphix.com>
Reviewed-by: Serapheim Dimitropoulos <serapheim@delphix.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Paul Dagnelie <pcd@delphix.com>
Closes#9128
When we unload metaslabs today in ZFS, the cached max_size value is
discarded. We instead use the histogram to determine whether or not we
think we can satisfy an allocation from the metaslab. This can result in
situations where, if we're doing I/Os of a size not aligned to a
histogram bucket, a metaslab is loaded even though it cannot satisfy the
allocation we think it can. For example, a metaslab with 16 entries in
the 16k-32k bucket may have entirely 16kB entries. If we try to allocate
a 24kB buffer, we will load that metaslab because we think it should be
able to handle the allocation. Doing so is expensive in CPU time, disk
reads, and average IO latency. This is exacerbated if the write being
attempted is a sync write.
This change makes ZFS cache the max_size after the metaslab is
unloaded. If we ever get a free (or a coalesced group of frees) larger
than the max_size, we will update it. Otherwise, we leave it as is. When
attempting to allocate, we use the max_size as a lower bound, and
respect it unless we are in try_hard. However, we do age the max_size
out at some point, since we expect the actual max_size to increase as we
do more frees. A more sophisticated algorithm here might be helpful, but
this works reasonably well.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Matt Ahrens <mahrens@delphix.com>
Signed-off-by: Paul Dagnelie <pcd@delphix.com>
Closes#9055
When the log spacemap commit was merged in ZoL, the
metaslab_verify_unflushed_changes() debugging function
was deleted as the feature was pretty much stable by
then. Unfortunately though there was a reference to
it from a comment in metaslab_verify_weight_and_frag().
This patch deletes the reference and pastes that
comment as is.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Matt Ahrens <mahrens@delphix.com>
Reviewed-by: Igor Kozhukhov <igor@dilos.org>
Reviewed-by: George Melikov <mail@gmelikov.ru>
Signed-off-by: Serapheim Dimitropoulos <serapheim@delphix.com>
Closes#9097
metaslab_should_allocate() is used in two places:
[1] When trying to select a metaslab to allocate from
[2] When trying to allocate from a metaslab
In [2] we always expect the metaslab to be loaded, and after
the refactoring of the log spacemap changes, whenever we load
a metaslab we set ms_max_size to the biggest range in the
ms_allocatable tree. Thus, when it is used in [2], if that
field is 0, it means that the metaslab doesn't have any
segments that can be used for allocations now (though it may
have some free space but that space can be in the freeing,
freed, or deferred trees).
In [1] a metaslab can be loaded or unloaded at which point 0
can either mean the metaslab doesn't have any space or the
metaslab is just not loaded thus we go ahead and try to make
an estimation based on its weight.
The issue here is when we call the above function for [2] and
the metaslab doesn't have any allocatable space, we still go
ahead and check its ms_weight which may be out of date because
we haven't ran metaslab_sync_done() yet. At that point we are
allowing an allocation to be attempted even though we know
there is no range that is allocatable.
This patch fixes this issue by explicitly checking if the
metaslab is loaded and if it is, the ms_max_size is used.
Reviewed-by: Matt Ahrens <mahrens@delphix.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Serapheim Dimitropoulos <serapheim@delphix.com>
Closes#9045
= Motivation
At Delphix we've seen a lot of customer systems where fragmentation
is over 75% and random writes take a performance hit because a lot
of time is spend on I/Os that update on-disk space accounting metadata.
Specifically, we seen cases where 20% to 40% of sync time is spend
after sync pass 1 and ~30% of the I/Os on the system is spent updating
spacemaps.
The problem is that these pools have existed long enough that we've
touched almost every metaslab at least once, and random writes
scatter frees across all metaslabs every TXG, thus appending to
their spacemaps and resulting in many I/Os. To give an example,
assuming that every VDEV has 200 metaslabs and our writes fit within
a single spacemap block (generally 4K) we have 200 I/Os. Then if we
assume 2 levels of indirection, we need 400 additional I/Os and
since we are talking about metadata for which we keep 2 extra copies
for redundancy we need to triple that number, leading to a total of
1800 I/Os per VDEV every TXG.
We could try and decrease the number of metaslabs so we have less
I/Os per TXG but then each metaslab would cover a wider range on
disk and thus would take more time to be loaded in memory from disk.
In addition, after it's loaded, it's range tree would consume more
memory.
Another idea would be to just increase the spacemap block size
which would allow us to fit more entries within an I/O block
resulting in fewer I/Os per metaslab and a speedup in loading time.
The problem is still that we don't deal with the number of I/Os
going up as the number of metaslabs is increasing and the fact
is that we generally write a lot to a few metaslabs and a little
to the rest of them. Thus, just increasing the block size would
actually waste bandwidth because we won't be utilizing our bigger
block size.
= About this patch
This patch introduces the Log Spacemap project which provides the
solution to the above problem while taking into account all the
aforementioned tradeoffs. The details on how it achieves that can
be found in the references sections below and in the code (see
Big Theory Statement in spa_log_spacemap.c).
Even though the change is fairly constraint within the metaslab
and lower-level SPA codepaths, there is a side-change that is
user-facing. The change is that VDEV IDs from VDEV holes will no
longer be reused. To give some background and reasoning for this,
when a log device is removed and its VDEV structure was replaced
with a hole (or was compacted; if at the end of the vdev array),
its vdev_id could be reused by devices added after that. Now
with the pool-wide space maps recording the vdev ID, this behavior
can cause problems (e.g. is this entry referring to a segment in
the new vdev or the removed log?). Thus, to simplify things the
ID reuse behavior is gone and now vdev IDs for top-level vdevs
are truly unique within a pool.
= Testing
The illumos implementation of this feature has been used internally
for a year and has been in production for ~6 months. For this patch
specifically there don't seem to be any regressions introduced to
ZTS and I have been running zloop for a week without any related
problems.
= Performance Analysis (Linux Specific)
All performance results and analysis for illumos can be found in
the links of the references. Redoing the same experiments in Linux
gave similar results. Below are the specifics of the Linux run.
After the pool reached stable state the percentage of the time
spent in pass 1 per TXG was 64% on average for the stock bits
while the log spacemap bits stayed at 95% during the experiment
(graph: sdimitro.github.io/img/linux-lsm/PercOfSyncInPassOne.png).
Sync times per TXG were 37.6 seconds on average for the stock
bits and 22.7 seconds for the log spacemap bits (related graph:
sdimitro.github.io/img/linux-lsm/SyncTimePerTXG.png). As a result
the log spacemap bits were able to push more TXGs, which is also
the reason why all graphs quantified per TXG have more entries for
the log spacemap bits.
Another interesting aspect in terms of txg syncs is that the stock
bits had 22% of their TXGs reach sync pass 7, 55% reach sync pass 8,
and 20% reach 9. The log space map bits reached sync pass 4 in 79%
of their TXGs, sync pass 7 in 19%, and sync pass 8 at 1%. This
emphasizes the fact that not only we spend less time on metadata
but we also iterate less times to convergence in spa_sync() dirtying
objects.
[related graphs:
stock- sdimitro.github.io/img/linux-lsm/NumberOfPassesPerTXGStock.png
lsm- sdimitro.github.io/img/linux-lsm/NumberOfPassesPerTXGLSM.png]
Finally, the improvement in IOPs that the userland gains from the
change is approximately 40%. There is a consistent win in IOPS as
you can see from the graphs below but the absolute amount of
improvement that the log spacemap gives varies within each minute
interval.
sdimitro.github.io/img/linux-lsm/StockVsLog3Days.png
sdimitro.github.io/img/linux-lsm/StockVsLog10Hours.png
= Porting to Other Platforms
For people that want to port this commit to other platforms below
is a list of ZoL commits that this patch depends on:
Make zdb results for checkpoint tests consistent
db587941c5
Update vdev_is_spacemap_addressable() for new spacemap encoding
419ba59145
Simplify spa_sync by breaking it up to smaller functions
8dc2197b7b
Factor metaslab_load_wait() in metaslab_load()
b194fab0fb
Rename range_tree_verify to range_tree_verify_not_present
df72b8bebe
Change target size of metaslabs from 256GB to 16GB
c853f382db
zdb -L should skip leak detection altogether
21e7cf5da8
vs_alloc can underflow in L2ARC vdevs
7558997d2f
Simplify log vdev removal code
6c926f426a
Get rid of space_map_update() for ms_synced_length
425d3237ee
Introduce auxiliary metaslab histograms
928e8ad47d
Error path in metaslab_load_impl() forgets to drop ms_sync_lock
8eef997679
= References
Background, Motivation, and Internals of the Feature
- OpenZFS 2017 Presentation:
youtu.be/jj2IxRkl5bQ
- Slides:
slideshare.net/SerapheimNikolaosDim/zfs-log-spacemaps-project
Flushing Algorithm Internals & Performance Results
(Illumos Specific)
- Blogpost:
sdimitro.github.io/post/zfs-lsm-flushing/
- OpenZFS 2018 Presentation:
youtu.be/x6D2dHRjkxw
- Slides:
slideshare.net/SerapheimNikolaosDim/zfs-log-spacemap-flushing-algorithm
Upstream Delphix Issues:
DLPX-51539, DLPX-59659, DLPX-57783, DLPX-61438, DLPX-41227, DLPX-59320
DLPX-63385
Reviewed-by: Sean Eric Fagan <sef@ixsystems.com>
Reviewed-by: Matt Ahrens <matt@delphix.com>
Reviewed-by: George Wilson <gwilson@delphix.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Serapheim Dimitropoulos <serapheim@delphix.com>
Closes#8442
We return ENOSPC in metaslab_activate if the metaslab has weight 0,
to avoid activating a metaslab with no space available. For sanity
checking, we also assert that there is no free space in the range
tree in that case.
Reviewed-by: Igor Kozhukhov <igor@dilos.org>
Reviewed by: Matt Ahrens <matt@delphix.com>
Reviewed by: Serapheim Dimitropoulos <serapheim.dimitro@delphix.com>
Reviewed by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Paul Dagnelie <pcd@delphix.com>
Closes#8968
With the new parallel allocators scheme, there is a possibility for
a problem where two threads, allocating from the same allocator at
the same time, conflict with each other. There are two primary cases
to worry about. First, another thread working on another allocator
activates the same metaslab that the first thread was trying to
activate. This results in the first thread needing to go back and
reselect a new metaslab, even though it may have waited a long time
for this metaslab to load. Second, another thread working on the same
allocator may have activated a different metaslab while the first
thread was waiting for its metaslab to load. Both of these cases
can cause the first thread to be significantly delayed in issuing
its IOs. The second case can also cause metaslab load/unload churn;
because the metaslab is loaded but not fully activated, we never set
the selected_txg, which results in the metaslab being immediately
unloaded again. This process can repeat many times, wasting disk and
cpu resources. This is more likely to happen when the IO of the first
thread is a larger one (like a ZIL write) and the other thread is
doing a smaller write, because it is more likely to find an
acceptable metaslab quickly.
There are two primary changes. The first is to always proceed with
the allocation when returning from metaslab_activate if we were
preempted in either of the ways described in the previous section.
The second change is to set the selected_txg before we do the call
to activate so that even if the metaslab is not used for an
allocation, we won't immediately attempt to unload it.
Reviewed by: Jerry Jelinek <jerry.jelinek@joyent.com>
Reviewed by: Matt Ahrens <matt@delphix.com>
Reviewed by: Serapheim Dimitropoulos <serapheim.dimitro@delphix.com>
Reviewed by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Paul Dagnelie <pcd@delphix.com>
External-issue: DLPX-61314
Closes#8843
On fragmented pools with high-performance storage, the looping in
metaslab_block_picker() can become the performance-limiting bottleneck.
When looking for a larger block (e.g. a 128K block for the ZIL), we may
search through many free segments (up to hundreds of thousands) to find
one that is large enough to satisfy the allocation. This can take a long
time (up to dozens of ms), and is done while holding the ms_lock, which
other threads may spin waiting for.
When this performance problem is encountered, profiling will show
high CPU time in metaslab_block_picker, as well as in mutex_enter from
various callers.
The problem is very evident on a test system with a sync write workload
with 8K writes to a recordsize=8k filesystem, with 4TB of SSD storage,
84% full and 88% fragmented. It has also been observed on production
systems with 90TB of storage, 76% full and 87% fragmented.
The fix is to change metaslab_df_alloc() to search only up to 16MB from
the previous allocation (of this alignment). After that, we will pick a
segment that is of the exact size requested (or larger). This reduces
the number of iterations to a few hundred on fragmented pools (a ~100x
improvement).
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Paul Dagnelie <pcd@delphix.com>
Reviewed-by: Tony Nguyen <tony.nguyen@delphix.com>
Reviewed-by: George Wilson <george.wilson@delphix.com>
Reviewed-by: Serapheim Dimitropoulos <serapheim@delphix.com>
Signed-off-by: Matthew Ahrens <mahrens@delphix.com>
External-issue: DLPX-62324
Closes#8877
On large systems, the memory used by loaded metaslabs can become
a concern. While range trees are a fairly efficient data structure,
on heavily fragmented pools they can still consume a significant
amount of memory. This problem is amplified when we fail to unload
metaslabs that we aren't using. Currently, we only unload a metaslab
during metaslab_sync_done; in order for that function to be called
on a given metaslab in a given txg, we have to have dirtied that
metaslab in that txg. If the dirtying was the result of an allocation,
we wouldn't be unloading it (since it wouldn't be 8 txgs since it
was selected), so in effect we only unload a metaslab during txgs
where it's being freed from.
We move the unload logic from sync_done to a new function, and
call that function on all metaslabs in a given vdev during
vdev_sync_done().
Reviewed-by: Richard Elling <Richard.Elling@RichardElling.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Paul Dagnelie <pcd@delphix.com>
Closes#8837
Historically while doing performance testing we've noticed that IOPS
can be significantly reduced when all vdevs in the pool are hitting
the zfs_mg_fragmentation_threshold percentage. Specifically in a
hypothetical pool with two vdevs, what can happen is the following:
Vdev A would go above that threshold and only vdev B would be used.
Then vdev B would pass that threshold but vdev A would go below it
(we've been freeing from A to allocate to B). The allocations would
go back and forth utilizing one vdev at a time with IOPS taking a hit.
Empirically, we've seen that our vdev selection for allocations is
good enough that fragmentation increases uniformly across all vdevs
the majority of the time. Thus we set the threshold percentage high
enough to avoid hitting the speed bump on pools that are being pushed
to the edge. We effectively disable its effect in the majority of the
cases but we don't remove (at least for now) just in case we hit any
weird behavior in the future.
Reviewed-by: George Melikov <mail@gmelikov.ru>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Matt Ahrens <mahrens@delphix.com>
Signed-off-by: Serapheim Dimitropoulos <serapheim@delphix.com>
Closes#8859
There are several places where we use zfs_dbgmsg and %p to
print pointers. In the Linux kernel, these values obfuscated
to prevent information leaks which means the pointers aren't
very useful for debugging crash dumps. We decided to restrict
the permissions of dbgmsg (and some other kstats while we were
at it) and print pointers with %px in zfs_dbgmsg as well as
spl_dumpstack
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: John Gallagher <john.gallagher@delphix.com>
Signed-off-by: sara hartse <sara.hartse@delphix.com>
Closes#8467Closes#8476
UNMAP/TRIM support is a frequently-requested feature to help
prevent performance from degrading on SSDs and on various other
SAN-like storage back-ends. By issuing UNMAP/TRIM commands for
sectors which are no longer allocated the underlying device can
often more efficiently manage itself.
This TRIM implementation is modeled on the `zpool initialize`
feature which writes a pattern to all unallocated space in the
pool. The new `zpool trim` command uses the same vdev_xlate()
code to calculate what sectors are unallocated, the same per-
vdev TRIM thread model and locking, and the same basic CLI for
a consistent user experience. The core difference is that
instead of writing a pattern it will issue UNMAP/TRIM commands
for those extents.
The zio pipeline was updated to accommodate this by adding a new
ZIO_TYPE_TRIM type and associated spa taskq. This new type makes
is straight forward to add the platform specific TRIM/UNMAP calls
to vdev_disk.c and vdev_file.c. These new ZIO_TYPE_TRIM zios are
handled largely the same way as ZIO_TYPE_READs or ZIO_TYPE_WRITEs.
This makes it possible to largely avoid changing the pipieline,
one exception is that TRIM zio's may exceed the 16M block size
limit since they contain no data.
In addition to the manual `zpool trim` command, a background
automatic TRIM was added and is controlled by the 'autotrim'
property. It relies on the exact same infrastructure as the
manual TRIM. However, instead of relying on the extents in a
metaslab's ms_allocatable range tree, a ms_trim tree is kept
per metaslab. When 'autotrim=on', ranges added back to the
ms_allocatable tree are also added to the ms_free tree. The
ms_free tree is then periodically consumed by an autotrim
thread which systematically walks a top level vdev's metaslabs.
Since the automatic TRIM will skip ranges it considers too small
there is value in occasionally running a full `zpool trim`. This
may occur when the freed blocks are small and not enough time
was allowed to aggregate them. An automatic TRIM and a manual
`zpool trim` may be run concurrently, in which case the automatic
TRIM will yield to the manual TRIM.
Reviewed-by: Jorgen Lundman <lundman@lundman.net>
Reviewed-by: Tim Chase <tim@chase2k.com>
Reviewed-by: Matt Ahrens <mahrens@delphix.com>
Reviewed-by: George Wilson <george.wilson@delphix.com>
Reviewed-by: Serapheim Dimitropoulos <serapheim@delphix.com>
Contributions-by: Saso Kiselkov <saso.kiselkov@nexenta.com>
Contributions-by: Tim Chase <tim@chase2k.com>
Contributions-by: Chunwei Chen <tuxoko@gmail.com>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes#8419Closes#598
This patch introduces 3 new histograms per metaslab. These
histograms track segments that have made it to the metaslab's
space map histogram (and are part of the spacemap) but have
not yet reached the ms_allocatable tree on loaded metaslab's
because these metaslab's are currently syncing and haven't
gone through metaslab_sync_done() yet.
The histograms help when we decide whether to load an unloaded
metaslab in-order to allocate from it. When calculating the
weight of an unloaded metaslab traditionally, we look at the
highest bucket of its spacemap's histogram. The problem is
that we are not guaranteed to be able to allocated that
segment when we load the metaslab because it may still be at
the freeing, freed, or defer trees. The new histograms are
used when we try to calculate an unloaded metaslab's weight
to deal with this issue by removing segments that have would
not be in the allocatable tree at runtime. Note, that this
method of dealing with this is not completely accurate as
adjacent segments are not always consolidated in the space
map histogram of a metaslab.
In addition and to make things deterministic, we always reset
the weight of unloaded metaslabs based on their space map
weight (instead of doing that on a need basis). Thus, every
time a metaslab is loaded and its weight is reset again (from
the weight based on its space map to the one based on its
allocatable range tree) we expect (and assert) that this
change in weight can only get better if it doesn't stay the
same.
Reviewed by: Paul Dagnelie <pcd@delphix.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed by: Matt Ahrens <mahrens@delphix.com>
Signed-off-by: Serapheim Dimitropoulos <serapheim@delphix.com>
Closes#8358
Initially, metaslabs and space maps used to be the same thing
in ZFS. Later, we started differentiating them by referring
to the space map as the on-disk state of the metaslab, making
the metaslab a higher-level concept that is metadata that deals
with space accounting. Today we've managed to split that code
furthermore, with the space map being its own on-disk data
structure used in areas of ZFS besides metaslabs (e.g. the
vdev-wide space maps used for zpool checkpoint or vdev removal
features).
This patch refactors the space map code to further split the
space map code from the metaslab code. It does so by getting
rid of the idea that the space map can have a different in-core
and on-disk length (sm_length vs smp_length) which is something
that is only used for the metaslab code, and other consumers
of space maps just have to deal with. Instead, this patch
introduces changes that move the old in-core length of the
metaslab's space map to the metaslab structure itself (see
ms_synced_length field) while making the space map code only
care about the actual space map's length on-disk.
The result of this is that space map consumers no longer have
to deal with syncing two different lengths for the same
structure (e.g. space_map_update() goes away) while metaslab
specific behavior stays within the metaslab code. Specifically,
the ms_synced_length field keeps track of the amount of data
metaslab_load() can read from the metaslab's space map while
working concurrently with metaslab_sync() that may be
appending to that same space map.
As a side note, the patch also adds a few comments around
the metaslab code documenting some assumptions and expected
behavior.
Reviewed-by: Matt Ahrens <mahrens@delphix.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed by: Pavel Zakharov <pavel.zakharov@delphix.com>
Signed-off-by: Serapheim Dimitropoulos <serapheim@delphix.com>
Closes#8328
The range_tree_verify function looks for a segment in a
range tree and panics if the segment is present on the
tree. This patch gives the function a more descriptive
name.
Reviewed-by: Matt Ahrens <mahrens@delphix.com>
Reviewed-by: George Melikov <mail@gmelikov.ru>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Serapheim Dimitropoulos <serapheim@delphix.com>
Closes#8327
Most callers that need to operate on a loaded metaslab, always
call metaslab_load_wait() before loading the metaslab just in
case someone else is already doing the work.
Factoring metaslab_load_wait() within metaslab_load() makes the
later more robust, as callers won't have to do the load-wait
check explicitly every time they need to load a metaslab.
Reviewed-by: Matt Ahrens <mahrens@delphix.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Serapheim Dimitropoulos <serapheim@delphix.com>
Closes#8290
PROBLEM
========
The first access to a block incurs a performance penalty on some platforms
(e.g. AWS's EBS, VMware VMDKs). Therefore we recommend that volumes are
"thick provisioned", where supported by the platform (VMware). This can
create a large delay in getting a new virtual machines up and running (or
adding storage to an existing Engine). If the thick provision step is
omitted, write performance will be suboptimal until all blocks on the LUN
have been written.
SOLUTION
=========
This feature introduces a way to 'initialize' the disks at install or in the
background to make sure we don't incur this first read penalty.
When an entire LUN is added to ZFS, we make all space available immediately,
and allow ZFS to find unallocated space and zero it out. This works with
concurrent writes to arbitrary offsets, ensuring that we don't zero out
something that has been (or is in the middle of being) written. This scheme
can also be applied to existing pools (affecting only free regions on the
vdev). Detailed design:
- new subcommand:zpool initialize [-cs] <pool> [<vdev> ...]
- start, suspend, or cancel initialization
- Creates new open-context thread for each vdev
- Thread iterates through all metaslabs in this vdev
- Each metaslab:
- select a metaslab
- load the metaslab
- mark the metaslab as being zeroed
- walk all free ranges within that metaslab and translate
them to ranges on the leaf vdev
- issue a "zeroing" I/O on the leaf vdev that corresponds to
a free range on the metaslab we're working on
- continue until all free ranges for this metaslab have been
"zeroed"
- reset/unmark the metaslab being zeroed
- if more metaslabs exist, then repeat above tasks.
- if no more metaslabs, then we're done.
- progress for the initialization is stored on-disk in the vdev’s
leaf zap object. The following information is stored:
- the last offset that has been initialized
- the state of the initialization process (i.e. active,
suspended, or canceled)
- the start time for the initialization
- progress is reported via the zpool status command and shows
information for each of the vdevs that are initializing
Porting notes:
- Added zfs_initialize_value module parameter to set the pattern
written by "zpool initialize".
- Added zfs_vdev_{initializing,removal}_{min,max}_active module options.
Authored by: George Wilson <george.wilson@delphix.com>
Reviewed by: John Wren Kennedy <john.kennedy@delphix.com>
Reviewed by: Matthew Ahrens <mahrens@delphix.com>
Reviewed by: Pavel Zakharov <pavel.zakharov@delphix.com>
Reviewed by: Prakash Surya <prakash.surya@delphix.com>
Reviewed by: loli10K <ezomori.nozomu@gmail.com>
Reviewed by: Brian Behlendorf <behlendorf1@llnl.gov>
Approved by: Richard Lowe <richlowe@richlowe.net>
Signed-off-by: Tim Chase <tim@chase2k.com>
Ported-by: Tim Chase <tim@chase2k.com>
OpenZFS-issue: https://www.illumos.org/issues/9102
OpenZFS-commit: https://github.com/openzfs/openzfs/commit/c3963210ebCloses#8230
In order to validate the gang block code ztest is configured to
artificially force a fraction of large blocks to be written as
gang blocks. The default setting chosen for this was to
write 25% of all blocks 32k or larger using gang blocks.
The confluence of an unrealistically large number of gang blocks,
the aggressive fault injection done by ztest, and the split
segment reconstruction logic introduced by device removal has
resulted in the following type of failure:
zdb -bccsv -G -d ... exit code 3
Specifically, zdb was unable to open the pool because it was
unable to reconstruct a damaged block. Manual investigation
of multiple failures clearly showed that the block could be
reconstructed. However, due to the large number of damaged
segments (>35) it could not be done in the allotted time.
Furthermore, the large number of gang blocks was determined
to be the reason for the unrealistically large number of
damaged segments. In order to make this situation less
likely, this change both increases the forced gang block
size to 64k and reduces the frequency to 3% of blocks.
Reviewed-by: Matthew Ahrens <mahrens@delphix.com>
Reviewed-by: Tom Caputi <tcaputi@datto.com>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes#8080
This patch fixes a small issue where the zil_check_log_chain()
code path would hit an EBUSY error. This would occur when
2 threads attempted to call metaslab_activate() at the same time.
In this case, the "loser" would receive an error code which should
have been ignored, but was instead floated to the caller. This
ended up resulting in an ENXIO being returned from from
spa_ld_verify_logs().
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Serapheim Dimitropoulos <serapheim.dimitro@delphix.com>
Reviewed-by: Matthew Ahrens <mahrens@delphix.com>
Signed-off-by: Tom Caputi <tcaputi@datto.com>
Closes#8010
Recent changes in the Linux kernel made it necessary to prefix
the refcount_add() function with zfs_ due to a name collision.
To bring the other functions in line with that and to avoid future
collisions, prefix the other refcount functions as well.
Reviewed by: Matthew Ahrens <mahrens@delphix.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Tim Schumacher <timschumi@gmx.de>
Closes#7963
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
Relax allocation throttling for ditto blocks. Due to random imbalances
in allocation it tends to push block copies to one vdev, that looks
slightly better at the moment. Slightly less strict policy allows both
improve data security and surprisingly write performance, since we don't
need to touch extra metaslabs on each vdev to respect the min distance.
Sponsored by: iXsystems, Inc.
Authored by: mav <mav@FreeBSD.org>
Reviewed by: Matt Ahrens <mahrens@delphix.com>
Reviewed by: Pavel Zakharov <pavel.zakharov@delphix.com>
Ported-by: Brian Behlendorf <behlendorf1@llnl.gov>
OpenZFS-issue: https://illumos.org/issues/9751
FreeBSD-commit: https://github.com/freebsd/freebsd/commit/8253837ac3Closes#7857
Overview
========
We parallelize the allocation process by creating the concept of
"allocators". There are a certain number of allocators per metaslab
group, defined by the value of a tunable at pool open time. Each
allocator for a given metaslab group has up to 2 active metaslabs; one
"primary", and one "secondary". The primary and secondary weight mean
the same thing they did in in the pre-allocator world; primary metaslabs
are used for most allocations, secondary metaslabs are used for ditto
blocks being allocated in the same metaslab group. There is also the
CLAIM weight, which has been separated out from the other weights, but
that is less important to understanding the patch. The active metaslabs
for each allocator are moved from their normal place in the metaslab
tree for the group to the back of the tree. This way, they will not be
selected for use by other allocators searching for new metaslabs unless
all the passive metaslabs are unsuitable for allocations. If that does
happen, the allocators will "steal" from each other to ensure that IOs
don't fail until there is truly no space left to perform allocations.
In addition, the alloc queue for each metaslab group has been broken
into a separate queue for each allocator. We don't want to dramatically
increase the number of inflight IOs on low-end systems, because it can
significantly increase txg times. On the other hand, we want to ensure
that there are enough IOs for each allocator to allow for good
coalescing before sending the IOs to the disk. As a result, we take a
compromise path; each allocator's alloc queue max depth starts at a
certain value for every txg. Every time an IO completes, we increase the
max depth. This should hopefully provide a good balance between the two
failure modes, while not dramatically increasing complexity.
We also parallelize the spa_alloc_tree and spa_alloc_lock, which cause
very similar contention when selecting IOs to allocate. This
parallelization uses the same allocator scheme as metaslab selection.
Performance Results
===================
Performance improvements from this change can vary significantly based
on the number of CPUs in the system, whether or not the system has a
NUMA architecture, the speed of the drives, the values for the various
tunables, and the workload being performed. For an fio async sequential
write workload on a 24 core NUMA system with 256 GB of RAM and 8 128 GB
SSDs, there is a roughly 25% performance improvement.
Future Work
===========
Analysis of the performance of the system with this patch applied shows
that a significant new bottleneck is the vdev disk queues, which also
need to be parallelized. Prototyping of this change has occurred, and
there was a performance improvement, but more work needs to be done
before its stability has been verified and it is ready to be upstreamed.
Authored by: Paul Dagnelie <pcd@delphix.com>
Reviewed by: Matthew Ahrens <mahrens@delphix.com>
Reviewed by: George Wilson <george.wilson@delphix.com>
Reviewed by: Serapheim Dimitropoulos <serapheim.dimitro@delphix.com>
Reviewed by: Alexander Motin <mav@FreeBSD.org>
Reviewed by: Brian Behlendorf <behlendorf1@llnl.gov>
Approved by: Gordon Ross <gwr@nexenta.com>
Ported-by: Paul Dagnelie <pcd@delphix.com>
Signed-off-by: Paul Dagnelie <pcd@delphix.com>
Porting Notes:
* Fix reservation test failures by increasing tolerance.
OpenZFS-issue: https://illumos.org/issues/9112
OpenZFS-commit: https://github.com/openzfs/openzfs/commit/3f3cc3c3Closes#7682
Motivation
==========
The current space map encoding has the following disadvantages:
[1] Assuming 512 sector size each entry can represent at most 16MB for a segment.
This makes the encoding very inefficient for large regions of space.
[2] As vdev-wide space maps have started to be used by new features (i.e.
device removal, zpool checkpoint) we've started imposing limits in the
vdevs that can be used with them based on the maximum addressable offset
(currently 64PB for a top-level vdev).
New encoding
============
The layout can be found at space_map.h and it remains backwards compatible with
the old one. The introduced two-word entry format, besides extending the limits
imposed by the single-entry layout, also includes a vdev field and some extra
padding after its prefix.
The extra padding after the prefix should is reserved for future usage (e.g.
new prefixes for future encodings or new fields for flags). The new vdev field
not only makes the space maps more self-descriptive, but also opens the doors
for pool-wide space maps (expected to be used in the log spacemap project).
One final important note is that the number of bits used for vdevs is reduced
to 24 bits for blkptrs. That was decided as we don't know of any setups that
use more than 16M vdevs for the time being and we wanted to fit the vdev field
in the space map. In addition that gives us some extra bits in dva_t.
Other references:
=================
The new encoding is also discussed towards the end of the Log Space Map
presentation from 2017's OpenZFS summit.
Link: https://www.youtube.com/watch?v=jj2IxRkl5bQ
Authored by: Serapheim Dimitropoulos <serapheim@delphix.com>
Reviewed by: Matt Ahrens <mahrens@delphix.com>
Reviewed by: George Wilson <gwilson@zfsmail.com>
Reviewed by: Brian Behlendorf <behlendorf1@llnl.gov>
Approved by: Gordon Ross <gwr@nexenta.com>
Ported-by: Tim Chase <tim@chase2k.com>
Signed-off-by: Tim Chase <tim@chase2k.com>
OpenZFS-commit: https://github.com/openzfs/openzfs/commit/90a56e6d
OpenZFS-issue: https://www.illumos.org/issues/9238Closes#7665
Details about the motivation of this feature and its usage can
be found in this blogpost:
https://sdimitro.github.io/post/zpool-checkpoint/
A lightning talk of this feature can be found here:
https://www.youtube.com/watch?v=fPQA8K40jAM
Implementation details can be found in big block comment of
spa_checkpoint.c
Side-changes that are relevant to this commit but not explained
elsewhere:
* renames members of "struct metaslab trees to be shorter without
losing meaning
* space_map_{alloc,truncate}() accept a block size as a
parameter. The reason is that in the current state all space
maps that we allocate through the DMU use a global tunable
(space_map_blksz) which defauls to 4KB. This is ok for metaslab
space maps in terms of bandwirdth since they are scattered all
over the disk. But for other space maps this default is probably
not what we want. Examples are device removal's vdev_obsolete_sm
or vdev_chedkpoint_sm from this review. Both of these have a
1:1 relationship with each vdev and could benefit from a bigger
block size.
Porting notes:
* The part of dsl_scan_sync() which handles async destroys has
been moved into the new dsl_process_async_destroys() function.
* Remove "VERIFY(!(flags & FWRITE))" in "kernel.c" so zhack can write
to block device backed pools.
* ZTS:
* Fix get_txg() in zpool_sync_001_pos due to "checkpoint_txg".
* Don't use large dd block sizes on /dev/urandom under Linux in
checkpoint_capacity.
* Adopt Delphix-OS's setting of 4 (spa_asize_inflation =
SPA_DVAS_PER_BP + 1) for the checkpoint_capacity test to speed
its attempts to fill the pool
* Create the base and nested pools with sync=disabled to speed up
the "setup" phase.
* Clear labels in test pool between checkpoint tests to avoid
duplicate pool issues.
* The import_rewind_device_replaced test has been marked as "known
to fail" for the reasons listed in its DISCLAIMER.
* New module parameters:
zfs_spa_discard_memory_limit,
zfs_remove_max_bytes_pause (not documented - debugging only)
vdev_max_ms_count (formerly metaslabs_per_vdev)
vdev_min_ms_count
Authored by: Serapheim Dimitropoulos <serapheim.dimitro@delphix.com>
Reviewed by: Matthew Ahrens <mahrens@delphix.com>
Reviewed by: John Kennedy <john.kennedy@delphix.com>
Reviewed by: Dan Kimmel <dan.kimmel@delphix.com>
Reviewed by: Brian Behlendorf <behlendorf1@llnl.gov>
Approved by: Richard Lowe <richlowe@richlowe.net>
Ported-by: Tim Chase <tim@chase2k.com>
Signed-off-by: Tim Chase <tim@chase2k.com>
OpenZFS-issue: https://illumos.org/issues/9166
OpenZFS-commit: https://github.com/openzfs/openzfs/commit/7159fdb8Closes#7570
Serapheim Dimitropoulos