mirror of
https://git.proxmox.com/git/mirror_zfs.git
synced 2024-11-17 18:11:00 +03:00
e8b96c6007
4045 zfs write throttle & i/o scheduler performance work 1. The ZFS i/o scheduler (vdev_queue.c) now divides i/os into 5 classes: sync read, sync write, async read, async write, and scrub/resilver. The scheduler issues a number of concurrent i/os from each class to the device. Once a class has been selected, an i/o is selected from this class using either an elevator algorithem (async, scrub classes) or FIFO (sync classes). The number of concurrent async write i/os is tuned dynamically based on i/o load, to achieve good sync i/o latency when there is not a high load of writes, and good write throughput when there is. See the block comment in vdev_queue.c (reproduced below) for more details. 2. The write throttle (dsl_pool_tempreserve_space() and txg_constrain_throughput()) is rewritten to produce much more consistent delays when under constant load. The new write throttle is based on the amount of dirty data, rather than guesses about future performance of the system. When there is a lot of dirty data, each transaction (e.g. write() syscall) will be delayed by the same small amount. This eliminates the "brick wall of wait" that the old write throttle could hit, causing all transactions to wait several seconds until the next txg opens. One of the keys to the new write throttle is decrementing the amount of dirty data as i/o completes, rather than at the end of spa_sync(). Note that the write throttle is only applied once the i/o scheduler is issuing the maximum number of outstanding async writes. See the block comments in dsl_pool.c and above dmu_tx_delay() (reproduced below) for more details. This diff has several other effects, including: * the commonly-tuned global variable zfs_vdev_max_pending has been removed; use per-class zfs_vdev_*_max_active values or zfs_vdev_max_active instead. * the size of each txg (meaning the amount of dirty data written, and thus the time it takes to write out) is now controlled differently. There is no longer an explicit time goal; the primary determinant is amount of dirty data. Systems that are under light or medium load will now often see that a txg is always syncing, but the impact to performance (e.g. read latency) is minimal. Tune zfs_dirty_data_max and zfs_dirty_data_sync to control this. * zio_taskq_batch_pct = 75 -- Only use 75% of all CPUs for compression, checksum, etc. This improves latency by not allowing these CPU-intensive tasks to consume all CPU (on machines with at least 4 CPU's; the percentage is rounded up). --matt APPENDIX: problems with the current i/o scheduler The current ZFS i/o scheduler (vdev_queue.c) is deadline based. The problem with this is that if there are always i/os pending, then certain classes of i/os can see very long delays. For example, if there are always synchronous reads outstanding, then no async writes will be serviced until they become "past due". One symptom of this situation is that each pass of the txg sync takes at least several seconds (typically 3 seconds). If many i/os become "past due" (their deadline is in the past), then we must service all of these overdue i/os before any new i/os. This happens when we enqueue a batch of async writes for the txg sync, with deadlines 2.5 seconds in the future. If we can't complete all the i/os in 2.5 seconds (e.g. because there were always reads pending), then these i/os will become past due. Now we must service all the "async" writes (which could be hundreds of megabytes) before we service any reads, introducing considerable latency to synchronous i/os (reads or ZIL writes). Notes on porting to ZFS on Linux: - zio_t gained new members io_physdone and io_phys_children. Because object caches in the Linux port call the constructor only once at allocation time, objects may contain residual data when retrieved from the cache. Therefore zio_create() was updated to zero out the two new fields. - vdev_mirror_pending() relied on the depth of the per-vdev pending queue (vq->vq_pending_tree) to select the least-busy leaf vdev to read from. This tree has been replaced by vq->vq_active_tree which is now used for the same purpose. - vdev_queue_init() used the value of zfs_vdev_max_pending to determine the number of vdev I/O buffers to pre-allocate. That global no longer exists, so we instead use the sum of the *_max_active values for each of the five I/O classes described above. - The Illumos implementation of dmu_tx_delay() delays a transaction by sleeping in condition variable embedded in the thread (curthread->t_delay_cv). We do not have an equivalent CV to use in Linux, so this change replaced the delay logic with a wrapper called zfs_sleep_until(). This wrapper could be adopted upstream and in other downstream ports to abstract away operating system-specific delay logic. - These tunables are added as module parameters, and descriptions added to the zfs-module-parameters.5 man page. spa_asize_inflation zfs_deadman_synctime_ms zfs_vdev_max_active zfs_vdev_async_write_active_min_dirty_percent zfs_vdev_async_write_active_max_dirty_percent zfs_vdev_async_read_max_active zfs_vdev_async_read_min_active zfs_vdev_async_write_max_active zfs_vdev_async_write_min_active zfs_vdev_scrub_max_active zfs_vdev_scrub_min_active zfs_vdev_sync_read_max_active zfs_vdev_sync_read_min_active zfs_vdev_sync_write_max_active zfs_vdev_sync_write_min_active zfs_dirty_data_max_percent zfs_delay_min_dirty_percent zfs_dirty_data_max_max_percent zfs_dirty_data_max zfs_dirty_data_max_max zfs_dirty_data_sync zfs_delay_scale The latter four have type unsigned long, whereas they are uint64_t in Illumos. This accommodates Linux's module_param() supported types, but means they may overflow on 32-bit architectures. The values zfs_dirty_data_max and zfs_dirty_data_max_max are the most likely to overflow on 32-bit systems, since they express physical RAM sizes in bytes. In fact, Illumos initializes zfs_dirty_data_max_max to 2^32 which does overflow. To resolve that, this port instead initializes it in arc_init() to 25% of physical RAM, and adds the tunable zfs_dirty_data_max_max_percent to override that percentage. While this solution doesn't completely avoid the overflow issue, it should be a reasonable default for most systems, and the minority of affected systems can work around the issue by overriding the defaults. - Fixed reversed logic in comment above zfs_delay_scale declaration. - Clarified comments in vdev_queue.c regarding when per-queue minimums take effect. - Replaced dmu_tx_write_limit in the dmu_tx kstat file with dmu_tx_dirty_delay and dmu_tx_dirty_over_max. The first counts how many times a transaction has been delayed because the pool dirty data has exceeded zfs_delay_min_dirty_percent. The latter counts how many times the pool dirty data has exceeded zfs_dirty_data_max (which we expect to never happen). - The original patch would have regressed the bug fixed in zfsonlinux/zfs@c418410, which prevented users from setting the zfs_vdev_aggregation_limit tuning larger than SPA_MAXBLOCKSIZE. A similar fix is added to vdev_queue_aggregate(). - In vdev_queue_io_to_issue(), dynamically allocate 'zio_t search' on the heap instead of the stack. In Linux we can't afford such large structures on the stack. Reviewed by: George Wilson <george.wilson@delphix.com> Reviewed by: Adam Leventhal <ahl@delphix.com> Reviewed by: Christopher Siden <christopher.siden@delphix.com> Reviewed by: Ned Bass <bass6@llnl.gov> Reviewed by: Brendan Gregg <brendan.gregg@joyent.com> Approved by: Robert Mustacchi <rm@joyent.com> References: http://www.illumos.org/issues/4045 illumos/illumos-gate@69962b5647 Ported-by: Ned Bass <bass6@llnl.gov> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #1913
288 lines
8.2 KiB
C
288 lines
8.2 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 (c) 2010, Oracle and/or its affiliates. All rights reserved.
|
|
* Copyright (c) 2013 by Delphix. All rights reserved.
|
|
*/
|
|
|
|
#ifndef _SYS_SA_IMPL_H
|
|
#define _SYS_SA_IMPL_H
|
|
|
|
#include <sys/dmu.h>
|
|
#include <sys/refcount.h>
|
|
#include <sys/list.h>
|
|
|
|
/*
|
|
* Array of known attributes and their
|
|
* various characteristics.
|
|
*/
|
|
typedef struct sa_attr_table {
|
|
sa_attr_type_t sa_attr;
|
|
uint8_t sa_registered;
|
|
uint16_t sa_length;
|
|
sa_bswap_type_t sa_byteswap;
|
|
char *sa_name;
|
|
} sa_attr_table_t;
|
|
|
|
/*
|
|
* Zap attribute format for attribute registration
|
|
*
|
|
* 64 56 48 40 32 24 16 8 0
|
|
* +-------+-------+-------+-------+-------+-------+-------+-------+
|
|
* | unused | len | bswap | attr num |
|
|
* +-------+-------+-------+-------+-------+-------+-------+-------+
|
|
*
|
|
* Zap attribute format for layout information.
|
|
*
|
|
* layout information is stored as an array of attribute numbers
|
|
* The name of the attribute is the layout number (0, 1, 2, ...)
|
|
*
|
|
* 16 0
|
|
* +---- ---+
|
|
* | attr # |
|
|
* +--------+
|
|
* | attr # |
|
|
* +--- ----+
|
|
* ......
|
|
*
|
|
*/
|
|
|
|
#define ATTR_BSWAP(x) BF32_GET(x, 16, 8)
|
|
#define ATTR_LENGTH(x) BF32_GET(x, 24, 16)
|
|
#define ATTR_NUM(x) BF32_GET(x, 0, 16)
|
|
#define ATTR_ENCODE(x, attr, length, bswap) \
|
|
{ \
|
|
BF64_SET(x, 24, 16, length); \
|
|
BF64_SET(x, 16, 8, bswap); \
|
|
BF64_SET(x, 0, 16, attr); \
|
|
}
|
|
|
|
#define TOC_OFF(x) BF32_GET(x, 0, 23)
|
|
#define TOC_ATTR_PRESENT(x) BF32_GET(x, 31, 1)
|
|
#define TOC_LEN_IDX(x) BF32_GET(x, 24, 4)
|
|
#define TOC_ATTR_ENCODE(x, len_idx, offset) \
|
|
{ \
|
|
BF32_SET(x, 31, 1, 1); \
|
|
BF32_SET(x, 24, 7, len_idx); \
|
|
BF32_SET(x, 0, 24, offset); \
|
|
}
|
|
|
|
#define SA_LAYOUTS "LAYOUTS"
|
|
#define SA_REGISTRY "REGISTRY"
|
|
|
|
/*
|
|
* Each unique layout will have their own table
|
|
* sa_lot (layout_table)
|
|
*/
|
|
typedef struct sa_lot {
|
|
avl_node_t lot_num_node;
|
|
avl_node_t lot_hash_node;
|
|
uint64_t lot_num;
|
|
uint64_t lot_hash;
|
|
sa_attr_type_t *lot_attrs; /* array of attr #'s */
|
|
uint32_t lot_var_sizes; /* how many aren't fixed size */
|
|
uint32_t lot_attr_count; /* total attr count */
|
|
list_t lot_idx_tab; /* should be only a couple of entries */
|
|
int lot_instance; /* used with lot_hash to identify entry */
|
|
} sa_lot_t;
|
|
|
|
/* index table of offsets */
|
|
typedef struct sa_idx_tab {
|
|
list_node_t sa_next;
|
|
sa_lot_t *sa_layout;
|
|
uint16_t *sa_variable_lengths;
|
|
refcount_t sa_refcount;
|
|
uint32_t *sa_idx_tab; /* array of offsets */
|
|
} sa_idx_tab_t;
|
|
|
|
/*
|
|
* Since the offset/index information into the actual data
|
|
* will usually be identical we can share that information with
|
|
* all handles that have the exact same offsets.
|
|
*
|
|
* You would typically only have a large number of different table of
|
|
* contents if you had a several variable sized attributes.
|
|
*
|
|
* Two AVL trees are used to track the attribute layout numbers.
|
|
* one is keyed by number and will be consulted when a DMU_OT_SA
|
|
* object is first read. The second tree is keyed by the hash signature
|
|
* of the attributes and will be consulted when an attribute is added
|
|
* to determine if we already have an instance of that layout. Both
|
|
* of these tree's are interconnected. The only difference is that
|
|
* when an entry is found in the "hash" tree the list of attributes will
|
|
* need to be compared against the list of attributes you have in hand.
|
|
* The assumption is that typically attributes will just be updated and
|
|
* adding a completely new attribute is a very rare operation.
|
|
*/
|
|
struct sa_os {
|
|
kmutex_t sa_lock;
|
|
boolean_t sa_need_attr_registration;
|
|
boolean_t sa_force_spill;
|
|
uint64_t sa_master_obj;
|
|
uint64_t sa_reg_attr_obj;
|
|
uint64_t sa_layout_attr_obj;
|
|
int sa_num_attrs;
|
|
sa_attr_table_t *sa_attr_table; /* private attr table */
|
|
sa_update_cb_t *sa_update_cb;
|
|
avl_tree_t sa_layout_num_tree; /* keyed by layout number */
|
|
avl_tree_t sa_layout_hash_tree; /* keyed by layout hash value */
|
|
int sa_user_table_sz;
|
|
sa_attr_type_t *sa_user_table; /* user name->attr mapping table */
|
|
};
|
|
|
|
/*
|
|
* header for all bonus and spill buffers.
|
|
*
|
|
* The header has a fixed portion with a variable number
|
|
* of "lengths" depending on the number of variable sized
|
|
* attributes which are determined by the "layout number"
|
|
*/
|
|
|
|
#define SA_MAGIC 0x2F505A /* ZFS SA */
|
|
typedef struct sa_hdr_phys {
|
|
uint32_t sa_magic;
|
|
/*
|
|
* Encoded with hdrsize and layout number as follows:
|
|
* 16 10 0
|
|
* +--------+-------+
|
|
* | hdrsz |layout |
|
|
* +--------+-------+
|
|
*
|
|
* Bits 0-10 are the layout number
|
|
* Bits 11-16 are the size of the header.
|
|
* The hdrsize is the number * 8
|
|
*
|
|
* For example.
|
|
* hdrsz of 1 ==> 8 byte header
|
|
* 2 ==> 16 byte header
|
|
*
|
|
*/
|
|
uint16_t sa_layout_info;
|
|
uint16_t sa_lengths[1]; /* optional sizes for variable length attrs */
|
|
/* ... Data follows the lengths. */
|
|
} sa_hdr_phys_t;
|
|
|
|
#define SA_HDR_LAYOUT_NUM(hdr) BF32_GET(hdr->sa_layout_info, 0, 10)
|
|
#define SA_HDR_SIZE(hdr) BF32_GET_SB(hdr->sa_layout_info, 10, 6, 3, 0)
|
|
#define SA_HDR_LAYOUT_INFO_ENCODE(x, num, size) \
|
|
{ \
|
|
BF32_SET_SB(x, 10, 6, 3, 0, size); \
|
|
BF32_SET(x, 0, 10, num); \
|
|
}
|
|
|
|
typedef enum sa_buf_type {
|
|
SA_BONUS = 1,
|
|
SA_SPILL = 2
|
|
} sa_buf_type_t;
|
|
|
|
typedef enum sa_data_op {
|
|
SA_LOOKUP,
|
|
SA_UPDATE,
|
|
SA_ADD,
|
|
SA_REPLACE,
|
|
SA_REMOVE
|
|
} sa_data_op_t;
|
|
|
|
/*
|
|
* Opaque handle used for most sa functions
|
|
*
|
|
* This needs to be kept as small as possible.
|
|
*/
|
|
|
|
struct sa_handle {
|
|
kmutex_t sa_lock;
|
|
dmu_buf_t *sa_bonus;
|
|
dmu_buf_t *sa_spill;
|
|
objset_t *sa_os;
|
|
void *sa_userp;
|
|
sa_idx_tab_t *sa_bonus_tab; /* idx of bonus */
|
|
sa_idx_tab_t *sa_spill_tab; /* only present if spill activated */
|
|
};
|
|
|
|
#define SA_GET_DB(hdl, type) \
|
|
(dmu_buf_impl_t *)((type == SA_BONUS) ? hdl->sa_bonus : hdl->sa_spill)
|
|
|
|
#define SA_GET_HDR(hdl, type) \
|
|
((sa_hdr_phys_t *)((dmu_buf_impl_t *)(SA_GET_DB(hdl, \
|
|
type))->db.db_data))
|
|
|
|
#define SA_IDX_TAB_GET(hdl, type) \
|
|
(type == SA_BONUS ? hdl->sa_bonus_tab : hdl->sa_spill_tab)
|
|
|
|
#define IS_SA_BONUSTYPE(a) \
|
|
((a == DMU_OT_SA) ? B_TRUE : B_FALSE)
|
|
|
|
#define SA_BONUSTYPE_FROM_DB(db) \
|
|
(dmu_get_bonustype((dmu_buf_t *)db))
|
|
|
|
#define SA_BLKPTR_SPACE (DN_MAX_BONUSLEN - sizeof (blkptr_t))
|
|
|
|
#define SA_LAYOUT_NUM(x, type) \
|
|
((!IS_SA_BONUSTYPE(type) ? 0 : (((IS_SA_BONUSTYPE(type)) && \
|
|
((SA_HDR_LAYOUT_NUM(x)) == 0)) ? 1 : SA_HDR_LAYOUT_NUM(x))))
|
|
|
|
|
|
#define SA_REGISTERED_LEN(sa, attr) sa->sa_attr_table[attr].sa_length
|
|
|
|
#define SA_ATTR_LEN(sa, idx, attr, hdr) ((SA_REGISTERED_LEN(sa, attr) == 0) ?\
|
|
hdr->sa_lengths[TOC_LEN_IDX(idx->sa_idx_tab[attr])] : \
|
|
SA_REGISTERED_LEN(sa, attr))
|
|
|
|
#define SA_SET_HDR(hdr, num, size) \
|
|
{ \
|
|
hdr->sa_magic = SA_MAGIC; \
|
|
SA_HDR_LAYOUT_INFO_ENCODE(hdr->sa_layout_info, num, size); \
|
|
}
|
|
|
|
#define SA_ATTR_INFO(sa, idx, hdr, attr, bulk, type, hdl) \
|
|
{ \
|
|
bulk.sa_size = SA_ATTR_LEN(sa, idx, attr, hdr); \
|
|
bulk.sa_buftype = type; \
|
|
bulk.sa_addr = \
|
|
(void *)((uintptr_t)TOC_OFF(idx->sa_idx_tab[attr]) + \
|
|
(uintptr_t)hdr); \
|
|
}
|
|
|
|
#define SA_HDR_SIZE_MATCH_LAYOUT(hdr, tb) \
|
|
(SA_HDR_SIZE(hdr) == (sizeof (sa_hdr_phys_t) + \
|
|
(tb->lot_var_sizes > 1 ? P2ROUNDUP((tb->lot_var_sizes - 1) * \
|
|
sizeof (uint16_t), 8) : 0)))
|
|
|
|
int sa_add_impl(sa_handle_t *, sa_attr_type_t,
|
|
uint32_t, sa_data_locator_t, void *, dmu_tx_t *);
|
|
|
|
void sa_register_update_callback_locked(objset_t *, sa_update_cb_t *);
|
|
int sa_size_locked(sa_handle_t *, sa_attr_type_t, int *);
|
|
|
|
void sa_default_locator(void **, uint32_t *, uint32_t, boolean_t, void *);
|
|
int sa_attr_size(sa_os_t *, sa_idx_tab_t *, sa_attr_type_t,
|
|
uint16_t *, sa_hdr_phys_t *);
|
|
|
|
#ifdef __cplusplus
|
|
extern "C" {
|
|
#endif
|
|
|
|
#ifdef __cplusplus
|
|
}
|
|
#endif
|
|
|
|
#endif /* _SYS_SA_IMPL_H */
|