c/mirror_zfs
1
0
Fork 0
mirror of https://git.proxmox.com/git/mirror_zfs.git synced 2024-12-26 03:09:34 +03:00
Code Issues Packages Projects Releases Wiki Activity
e7a06356c1
mirror_zfs/include/sys/dmu_objset.h
Matthew Ahrens ba67d82142
Improve zfs receive performance with lightweight write 
The performance of `zfs receive` can be bottlenecked on the CPU consumed
by the `receive_writer` thread, especially when receiving streams with
small compressed block sizes.  Much of the CPU is spent creating and
destroying dbuf's and arc buf's, one for each `WRITE` record in the send
stream.

This commit introduces the concept of "lightweight writes", which allows
`zfs receive` to write to the DMU by providing an ABD, and instantiating
only a new type of `dbuf_dirty_record_t`.  The dbuf and arc buf for this
"dirty leaf block" are not instantiated.

Because there is no dbuf with the dirty data, this mechanism doesn't
support reading from "lightweight-dirty" blocks (they would see the
on-disk state rather than the dirty data).  Since the dedup-receive code
has been removed, `zfs receive` is write-only, so this works fine.

Because there are no arc bufs for the received data, the received data
is no longer cached in the ARC.

Testing a receive of a stream with average compressed block size of 4KB,
this commit improves performance by 50%, while also reducing CPU usage
by 50% of a CPU.  On a per-block basis, CPU consumed by receive_writer()
and dbuf_evict() is now 1/7th (14%) of what it was.

Baseline: 450MB/s, CPU in receive_writer() 40% + dbuf_evict() 35%
New: 670MB/s, CPU in receive_writer() 17% + dbuf_evict() 0%

The code is also restructured in a few ways:

Added a `dr_dnode` field to the dbuf_dirty_record_t.  This simplifies
some existing code that no longer needs `DB_DNODE_ENTER()` and related
routines.  The new field is needed by the lightweight-type dirty record.

To ensure that the `dr_dnode` field remains valid until the dirty record
is freed, we have to ensure that the `dnode_move()` doesn't relocate the
dnode_t.  To do this we keep a hold on the dnode until it's zio's have
completed.  This is already done by the user-accounting code
(`userquota_updates_task()`), this commit extends that so that it always
keeps the dnode hold until zio completion (see `dnode_rele_task()`).

`dn_dirty_txg` was previously zeroed when the dnode was synced.  This
was not necessary, since its meaning can be "when was this dnode last
dirtied".  This change simplifies the new `dnode_rele_task()` code.

Removed some dead code related to `DRR_WRITE_BYREF` (dedup receive).

Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Paul Dagnelie <pcd@delphix.com>
Reviewed-by: George Wilson <gwilson@delphix.com>
Signed-off-by: Matthew Ahrens <mahrens@delphix.com>
Closes #11105
2020-12-11 10:26:02 -08:00

274 lines
9.2 KiB
C
Raw Blame History

/*
* 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) 2005, 2010, Oracle and/or its affiliates. All rights reserved.
* Copyright (c) 2012, 2018 by Delphix. All rights reserved.
* Copyright (c) 2013 by Saso Kiselkov. All rights reserved.
* Copyright (c) 2014 Spectra Logic Corporation, All rights reserved.
*/
/* Portions Copyright 2010 Robert Milkowski */
#ifndef _SYS_DMU_OBJSET_H
#define _SYS_DMU_OBJSET_H
#include <sys/spa.h>
#include <sys/arc.h>
#include <sys/txg.h>
#include <sys/zfs_context.h>
#include <sys/dnode.h>
#include <sys/zio.h>
#include <sys/zil.h>
#include <sys/sa.h>
#include <sys/zfs_ioctl.h>
#ifdef __cplusplus
extern "C" {
#endif
extern krwlock_t os_lock;
struct dsl_pool;
struct dsl_dataset;
struct dmu_tx;
#define OBJSET_PHYS_SIZE_V1 1024
#define OBJSET_PHYS_SIZE_V2 2048
#define OBJSET_PHYS_SIZE_V3 4096
#define OBJSET_BUF_HAS_USERUSED(buf) \
(arc_buf_size(buf) >= OBJSET_PHYS_SIZE_V2)
#define OBJSET_BUF_HAS_PROJECTUSED(buf) \
(arc_buf_size(buf) >= OBJSET_PHYS_SIZE_V3)
#define OBJSET_FLAG_USERACCOUNTING_COMPLETE (1ULL << 0)
#define OBJSET_FLAG_USEROBJACCOUNTING_COMPLETE (1ULL << 1)
#define OBJSET_FLAG_PROJECTQUOTA_COMPLETE (1ULL << 2)
/*
* This mask defines the set of flags which are "portable", meaning
* that they can be preserved when doing a raw encrypted zfs send.
* Flags included in this mask will be protected by os_portable_mac
* when the block of dnodes is encrypted. No portable flags currently
* exist.
*/
#define OBJSET_CRYPT_PORTABLE_FLAGS_MASK (0)
typedef struct objset_phys {
dnode_phys_t os_meta_dnode;
zil_header_t os_zil_header;
uint64_t os_type;
uint64_t os_flags;
uint8_t os_portable_mac[ZIO_OBJSET_MAC_LEN];
uint8_t os_local_mac[ZIO_OBJSET_MAC_LEN];
char os_pad0[OBJSET_PHYS_SIZE_V2 - sizeof (dnode_phys_t)*3 -
sizeof (zil_header_t) - sizeof (uint64_t)*2 -
2*ZIO_OBJSET_MAC_LEN];
dnode_phys_t os_userused_dnode;
dnode_phys_t os_groupused_dnode;
dnode_phys_t os_projectused_dnode;
char os_pad1[OBJSET_PHYS_SIZE_V3 - OBJSET_PHYS_SIZE_V2 -
sizeof (dnode_phys_t)];
} objset_phys_t;
typedef int (*dmu_objset_upgrade_cb_t)(objset_t *);
#define OBJSET_PROP_UNINITIALIZED ((uint64_t)-1)
struct objset {
/* Immutable: */
struct dsl_dataset *os_dsl_dataset;
spa_t *os_spa;
arc_buf_t *os_phys_buf;
objset_phys_t *os_phys;
boolean_t os_encrypted;
/*
* The following "special" dnodes have no parent, are exempt
* from dnode_move(), and are not recorded in os_dnodes, but they
* root their descendents in this objset using handles anyway, so
* that all access to dnodes from dbufs consistently uses handles.
*/
dnode_handle_t os_meta_dnode;
dnode_handle_t os_userused_dnode;
dnode_handle_t os_groupused_dnode;
dnode_handle_t os_projectused_dnode;
zilog_t *os_zil;
list_node_t os_evicting_node;
/* can change, under dsl_dir's locks: */
uint64_t os_dnodesize; /* default dnode size for new objects */
enum zio_checksum os_checksum;
enum zio_compress os_compress;
uint8_t os_complevel;
uint8_t os_copies;
enum zio_checksum os_dedup_checksum;
boolean_t os_dedup_verify;
zfs_logbias_op_t os_logbias;
zfs_cache_type_t os_primary_cache;
zfs_cache_type_t os_secondary_cache;
zfs_sync_type_t os_sync;
zfs_redundant_metadata_type_t os_redundant_metadata;
uint64_t os_recordsize;
/*
* The next four values are used as a cache of whatever's on disk, and
* are initialized the first time these properties are queried. Before
* being initialized with their real values, their values are
* OBJSET_PROP_UNINITIALIZED.
*/
uint64_t os_version;
uint64_t os_normalization;
uint64_t os_utf8only;
uint64_t os_casesensitivity;
/*
* The largest zpl file block allowed in special class.
* cached here instead of zfsvfs for easier access.
*/
int os_zpl_special_smallblock;
/*
* Pointer is constant; the blkptr it points to is protected by
* os_dsl_dataset->ds_bp_rwlock
*/
blkptr_t *os_rootbp;
/* no lock needed: */
struct dmu_tx *os_synctx; /* XXX sketchy */
zil_header_t os_zil_header;
multilist_t *os_synced_dnodes;
uint64_t os_flags;
uint64_t os_freed_dnodes;
boolean_t os_rescan_dnodes;
boolean_t os_raw_receive;
/* os_phys_buf should be written raw next txg */
boolean_t os_next_write_raw[TXG_SIZE];
/* Protected by os_obj_lock */
kmutex_t os_obj_lock;
uint64_t os_obj_next_chunk;
/* Per-CPU next object to allocate, protected by atomic ops. */
uint64_t *os_obj_next_percpu;
int os_obj_next_percpu_len;
/* Protected by os_lock */
kmutex_t os_lock;
multilist_t *os_dirty_dnodes[TXG_SIZE];
list_t os_dnodes;
list_t os_downgraded_dbufs;
/* Protects changes to DMU_{USER,GROUP,PROJECT}USED_OBJECT */
kmutex_t os_userused_lock;
/* stuff we store for the user */
kmutex_t os_user_ptr_lock;
void *os_user_ptr;
sa_os_t *os_sa;
/* kernel thread to upgrade this dataset */
kmutex_t os_upgrade_lock;
taskqid_t os_upgrade_id;
dmu_objset_upgrade_cb_t os_upgrade_cb;
boolean_t os_upgrade_exit;
int os_upgrade_status;
};
#define DMU_META_OBJSET 0
#define DMU_META_DNODE_OBJECT 0
#define DMU_OBJECT_IS_SPECIAL(obj) ((int64_t)(obj) <= 0)
#define DMU_META_DNODE(os) ((os)->os_meta_dnode.dnh_dnode)
#define DMU_USERUSED_DNODE(os) ((os)->os_userused_dnode.dnh_dnode)
#define DMU_GROUPUSED_DNODE(os) ((os)->os_groupused_dnode.dnh_dnode)
#define DMU_PROJECTUSED_DNODE(os) ((os)->os_projectused_dnode.dnh_dnode)
#define DMU_OS_IS_L2CACHEABLE(os) \
((os)->os_secondary_cache == ZFS_CACHE_ALL || \
(os)->os_secondary_cache == ZFS_CACHE_METADATA)
/* called from zpl */
int dmu_objset_hold(const char *name, void *tag, objset_t **osp);
int dmu_objset_hold_flags(const char *name, boolean_t decrypt, void *tag,
objset_t **osp);
int dmu_objset_own(const char *name, dmu_objset_type_t type,
boolean_t readonly, boolean_t decrypt, void *tag, objset_t **osp);
int dmu_objset_own_obj(struct dsl_pool *dp, uint64_t obj,
dmu_objset_type_t type, boolean_t readonly, boolean_t decrypt,
void *tag, objset_t **osp);
void dmu_objset_refresh_ownership(struct dsl_dataset *ds,
struct dsl_dataset **newds, boolean_t decrypt, void *tag);
void dmu_objset_rele(objset_t *os, void *tag);
void dmu_objset_rele_flags(objset_t *os, boolean_t decrypt, void *tag);
void dmu_objset_disown(objset_t *os, boolean_t decrypt, void *tag);
int dmu_objset_from_ds(struct dsl_dataset *ds, objset_t **osp);
void dmu_objset_stats(objset_t *os, nvlist_t *nv);
void dmu_objset_fast_stat(objset_t *os, dmu_objset_stats_t *stat);
void dmu_objset_space(objset_t *os, uint64_t *refdbytesp, uint64_t *availbytesp,
uint64_t *usedobjsp, uint64_t *availobjsp);
uint64_t dmu_objset_fsid_guid(objset_t *os);
int dmu_objset_find_dp(struct dsl_pool *dp, uint64_t ddobj,
int func(struct dsl_pool *, struct dsl_dataset *, void *),
void *arg, int flags);
void dmu_objset_evict_dbufs(objset_t *os);
inode_timespec_t dmu_objset_snap_cmtime(objset_t *os);
/* called from dsl */
void dmu_objset_sync(objset_t *os, zio_t *zio, dmu_tx_t *tx);
boolean_t dmu_objset_is_dirty(objset_t *os, uint64_t txg);
objset_t *dmu_objset_create_impl_dnstats(spa_t *spa, struct dsl_dataset *ds,
blkptr_t *bp, dmu_objset_type_t type, int levels, int blksz, int ibs,
dmu_tx_t *tx);
objset_t *dmu_objset_create_impl(spa_t *spa, struct dsl_dataset *ds,
blkptr_t *bp, dmu_objset_type_t type, dmu_tx_t *tx);
int dmu_objset_open_impl(spa_t *spa, struct dsl_dataset *ds, blkptr_t *bp,
objset_t **osp);
void dmu_objset_evict(objset_t *os);
void dmu_objset_sync_done(objset_t *os, dmu_tx_t *tx);
void dmu_objset_userquota_get_ids(dnode_t *dn, boolean_t before, dmu_tx_t *tx);
boolean_t dmu_objset_userused_enabled(objset_t *os);
void dmu_objset_userspace_upgrade(objset_t *os);
boolean_t dmu_objset_userspace_present(objset_t *os);
boolean_t dmu_objset_userobjused_enabled(objset_t *os);
boolean_t dmu_objset_userobjspace_upgradable(objset_t *os);
boolean_t dmu_objset_userobjspace_present(objset_t *os);
boolean_t dmu_objset_incompatible_encryption_version(objset_t *os);
boolean_t dmu_objset_projectquota_enabled(objset_t *os);
boolean_t dmu_objset_projectquota_present(objset_t *os);
boolean_t dmu_objset_projectquota_upgradable(objset_t *os);
void dmu_objset_id_quota_upgrade(objset_t *os);
int dmu_get_file_info(objset_t *os, dmu_object_type_t bonustype,
const void *data, zfs_file_info_t *zfi);
int dmu_fsname(const char *snapname, char *buf);
void dmu_objset_evict_done(objset_t *os);
void dmu_objset_willuse_space(objset_t *os, int64_t space, dmu_tx_t *tx);
void dmu_objset_init(void);
void dmu_objset_fini(void);
#ifdef __cplusplus
}
#endif
#endif /* _SYS_DMU_OBJSET_H */
Reference in New Issue View Git Blame Copy Permalink