mirror_zfs/include/sys/space_map.h
Serapheim Dimitropoulos 93e28d661e Log Spacemap Project
= 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
2019-07-16 10:11:49 -07:00

238 lines
8.4 KiB
C

/*
* CDDL HEADER START
*
* The contents of this file are subject to the terms of the
* Common Development and Distribution License (the "License").
* You may not use this file except in compliance with the License.
*
* You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
* or http://www.opensolaris.org/os/licensing.
* See the License for the specific language governing permissions
* and limitations under the License.
*
* When distributing Covered Code, include this CDDL HEADER in each
* file and include the License file at usr/src/OPENSOLARIS.LICENSE.
* If applicable, add the following below this CDDL HEADER, with the
* fields enclosed by brackets "[]" replaced with your own identifying
* information: Portions Copyright [yyyy] [name of copyright owner]
*
* CDDL HEADER END
*/
/*
* Copyright 2009 Sun Microsystems, Inc. All rights reserved.
* Use is subject to license terms.
*/
/*
* Copyright (c) 2012, 2019 by Delphix. All rights reserved.
*/
#ifndef _SYS_SPACE_MAP_H
#define _SYS_SPACE_MAP_H
#include <sys/avl.h>
#include <sys/range_tree.h>
#include <sys/dmu.h>
#ifdef __cplusplus
extern "C" {
#endif
/*
* The size of the space map object has increased to include a histogram.
* The SPACE_MAP_SIZE_V0 designates the original size and is used to
* maintain backward compatibility.
*/
#define SPACE_MAP_SIZE_V0 (3 * sizeof (uint64_t))
#define SPACE_MAP_HISTOGRAM_SIZE 32
/*
* The space_map_phys is the on-disk representation of the space map.
* Consumers of space maps should never reference any of the members of this
* structure directly. These members may only be updated in syncing context.
*
* Note the smp_object is no longer used but remains in the structure
* for backward compatibility.
*/
typedef struct space_map_phys {
/* object number: not needed but kept for backwards compatibility */
uint64_t smp_object;
/* length of the object in bytes */
uint64_t smp_length;
/* space allocated from the map */
int64_t smp_alloc;
/* reserved */
uint64_t smp_pad[5];
/*
* The smp_histogram maintains a histogram of free regions. Each
* bucket, smp_histogram[i], contains the number of free regions
* whose size is:
* 2^(i+sm_shift) <= size of free region in bytes < 2^(i+sm_shift+1)
*
* Note that, if log space map feature is enabled, histograms of
* space maps that belong to metaslabs will take into account any
* unflushed changes for their metaslabs, even though the actual
* space map doesn't have entries for these changes.
*/
uint64_t smp_histogram[SPACE_MAP_HISTOGRAM_SIZE];
} space_map_phys_t;
/*
* The space map object defines a region of space, its size, how much is
* allocated, and the on-disk object that stores this information.
* Consumers of space maps may only access the members of this structure.
*
* Note: the space_map may not be accessed concurrently; consumers
* must provide external locking if required.
*/
typedef struct space_map {
uint64_t sm_start; /* start of map */
uint64_t sm_size; /* size of map */
uint8_t sm_shift; /* unit shift */
objset_t *sm_os; /* objset for this map */
uint64_t sm_object; /* object id for this map */
uint32_t sm_blksz; /* block size for space map */
dmu_buf_t *sm_dbuf; /* space_map_phys_t dbuf */
space_map_phys_t *sm_phys; /* on-disk space map */
} space_map_t;
/*
* debug entry
*
* 2 2 10 50
* +-----+-----+------------+----------------------------------+
* | 1 0 | act | syncpass | txg (lower bits) |
* +-----+-----+------------+----------------------------------+
* 63 62 61 60 59 50 49 0
*
*
* one-word entry
*
* 1 47 1 15
* +-----------------------------------------------------------+
* | 0 | offset (sm_shift units) | type | run |
* +-----------------------------------------------------------+
* 63 62 16 15 14 0
*
*
* two-word entry
*
* 2 2 36 24
* +-----+-----+---------------------------+-------------------+
* | 1 1 | pad | run | vdev |
* +-----+-----+---------------------------+-------------------+
* 63 62 61 60 59 24 23 0
*
* 1 63
* +------+----------------------------------------------------+
* | type | offset |
* +------+----------------------------------------------------+
* 63 62 0
*
* Note that a two-word entry will not straddle a block boundary.
* If necessary, the last word of a block will be padded with a
* debug entry (with act = syncpass = txg = 0).
*/
typedef enum {
SM_ALLOC,
SM_FREE
} maptype_t;
typedef struct space_map_entry {
maptype_t sme_type;
uint32_t sme_vdev; /* max is 2^24-1; SM_NO_VDEVID if not present */
uint64_t sme_offset; /* max is 2^63-1; units of sm_shift */
uint64_t sme_run; /* max is 2^36; units of sm_shift */
} space_map_entry_t;
#define SM_NO_VDEVID (1 << SPA_VDEVBITS)
/* one-word entry constants */
#define SM_DEBUG_PREFIX 2
#define SM_OFFSET_BITS 47
#define SM_RUN_BITS 15
/* two-word entry constants */
#define SM2_PREFIX 3
#define SM2_OFFSET_BITS 63
#define SM2_RUN_BITS 36
#define SM_PREFIX_DECODE(x) BF64_DECODE(x, 62, 2)
#define SM_PREFIX_ENCODE(x) BF64_ENCODE(x, 62, 2)
#define SM_DEBUG_ACTION_DECODE(x) BF64_DECODE(x, 60, 2)
#define SM_DEBUG_ACTION_ENCODE(x) BF64_ENCODE(x, 60, 2)
#define SM_DEBUG_SYNCPASS_DECODE(x) BF64_DECODE(x, 50, 10)
#define SM_DEBUG_SYNCPASS_ENCODE(x) BF64_ENCODE(x, 50, 10)
#define SM_DEBUG_TXG_DECODE(x) BF64_DECODE(x, 0, 50)
#define SM_DEBUG_TXG_ENCODE(x) BF64_ENCODE(x, 0, 50)
#define SM_OFFSET_DECODE(x) BF64_DECODE(x, 16, SM_OFFSET_BITS)
#define SM_OFFSET_ENCODE(x) BF64_ENCODE(x, 16, SM_OFFSET_BITS)
#define SM_TYPE_DECODE(x) BF64_DECODE(x, 15, 1)
#define SM_TYPE_ENCODE(x) BF64_ENCODE(x, 15, 1)
#define SM_RUN_DECODE(x) (BF64_DECODE(x, 0, SM_RUN_BITS) + 1)
#define SM_RUN_ENCODE(x) BF64_ENCODE((x) - 1, 0, SM_RUN_BITS)
#define SM_RUN_MAX SM_RUN_DECODE(~0ULL)
#define SM_OFFSET_MAX SM_OFFSET_DECODE(~0ULL)
#define SM2_RUN_DECODE(x) (BF64_DECODE(x, SPA_VDEVBITS, SM2_RUN_BITS) + 1)
#define SM2_RUN_ENCODE(x) BF64_ENCODE((x) - 1, SPA_VDEVBITS, SM2_RUN_BITS)
#define SM2_VDEV_DECODE(x) BF64_DECODE(x, 0, SPA_VDEVBITS)
#define SM2_VDEV_ENCODE(x) BF64_ENCODE(x, 0, SPA_VDEVBITS)
#define SM2_TYPE_DECODE(x) BF64_DECODE(x, SM2_OFFSET_BITS, 1)
#define SM2_TYPE_ENCODE(x) BF64_ENCODE(x, SM2_OFFSET_BITS, 1)
#define SM2_OFFSET_DECODE(x) BF64_DECODE(x, 0, SM2_OFFSET_BITS)
#define SM2_OFFSET_ENCODE(x) BF64_ENCODE(x, 0, SM2_OFFSET_BITS)
#define SM2_RUN_MAX SM2_RUN_DECODE(~0ULL)
#define SM2_OFFSET_MAX SM2_OFFSET_DECODE(~0ULL)
boolean_t sm_entry_is_debug(uint64_t e);
boolean_t sm_entry_is_single_word(uint64_t e);
boolean_t sm_entry_is_double_word(uint64_t e);
typedef int (*sm_cb_t)(space_map_entry_t *sme, void *arg);
int space_map_load(space_map_t *sm, range_tree_t *rt, maptype_t maptype);
int space_map_load_length(space_map_t *sm, range_tree_t *rt, maptype_t maptype,
uint64_t length);
int space_map_iterate(space_map_t *sm, uint64_t length,
sm_cb_t callback, void *arg);
int space_map_incremental_destroy(space_map_t *sm, sm_cb_t callback, void *arg,
dmu_tx_t *tx);
boolean_t space_map_histogram_verify(space_map_t *sm, range_tree_t *rt);
void space_map_histogram_clear(space_map_t *sm);
void space_map_histogram_add(space_map_t *sm, range_tree_t *rt,
dmu_tx_t *tx);
uint64_t space_map_object(space_map_t *sm);
int64_t space_map_allocated(space_map_t *sm);
uint64_t space_map_length(space_map_t *sm);
uint64_t space_map_entries(space_map_t *sm, range_tree_t *rt);
uint64_t space_map_nblocks(space_map_t *sm);
void space_map_write(space_map_t *sm, range_tree_t *rt, maptype_t maptype,
uint64_t vdev_id, dmu_tx_t *tx);
uint64_t space_map_estimate_optimal_size(space_map_t *sm, range_tree_t *rt,
uint64_t vdev_id);
void space_map_truncate(space_map_t *sm, int blocksize, dmu_tx_t *tx);
uint64_t space_map_alloc(objset_t *os, int blocksize, dmu_tx_t *tx);
void space_map_free(space_map_t *sm, dmu_tx_t *tx);
void space_map_free_obj(objset_t *os, uint64_t smobj, dmu_tx_t *tx);
int space_map_open(space_map_t **smp, objset_t *os, uint64_t object,
uint64_t start, uint64_t size, uint8_t shift);
void space_map_close(space_map_t *sm);
#ifdef __cplusplus
}
#endif
#endif /* _SYS_SPACE_MAP_H */