Fast Clone Deletion

Deleting a clone requires finding blocks are clone-only, not shared
with the snapshot. This was done by traversing the entire block tree
which results in a large performance penalty for sparsely
written clones.

This is new method keeps track of clone blocks when they are
modified in a "Livelist" so that, when it’s time to delete,
the clone-specific blocks are already at hand.

We see performance improvements because now deletion work is
proportional to the number of clone-modified blocks, not the size
of the original dataset.

Reviewed-by: Sean Eric Fagan <sef@ixsystems.com>
Reviewed-by: Matt Ahrens <matt@delphix.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Serapheim Dimitropoulos <serapheim@delphix.com>
Signed-off-by: Sara Hartse <sara.hartse@delphix.com>
Closes #8416
This commit is contained in:
Sara Hartse
2019-07-26 10:54:14 -07:00
committed by Brian Behlendorf
parent d274ac5460
commit 37f03da8ba
38 changed files with 2583 additions and 205 deletions
+29 -2
View File
@@ -20,7 +20,7 @@
*/
/*
* Copyright (c) 2010, Oracle and/or its affiliates. All rights reserved.
* Copyright (c) 2015 by Delphix. All rights reserved.
* Copyright (c) 2018, 2019 by Delphix. All rights reserved.
*/
#ifndef _SYS_DSL_DEADLIST_H
@@ -28,12 +28,14 @@
#include <sys/bpobj.h>
#include <sys/zfs_context.h>
#include <sys/zthr.h>
#ifdef __cplusplus
extern "C" {
#endif
struct dmu_buf;
struct dsl_pool;
struct dsl_dataset;
typedef struct dsl_deadlist_phys {
@@ -63,13 +65,34 @@ typedef struct dsl_deadlist_entry {
bpobj_t dle_bpobj;
} dsl_deadlist_entry_t;
typedef struct livelist_condense_entry {
struct dsl_dataset *ds;
dsl_deadlist_entry_t *first;
dsl_deadlist_entry_t *next;
boolean_t syncing;
boolean_t cancelled;
} livelist_condense_entry_t;
extern unsigned long zfs_livelist_max_entries;
extern int zfs_livelist_min_percent_shared;
typedef int deadlist_iter_t(void *args, dsl_deadlist_entry_t *dle);
void dsl_deadlist_open(dsl_deadlist_t *dl, objset_t *os, uint64_t object);
void dsl_deadlist_close(dsl_deadlist_t *dl);
void dsl_deadlist_iterate(dsl_deadlist_t *dl, deadlist_iter_t func, void *arg);
uint64_t dsl_deadlist_alloc(objset_t *os, dmu_tx_t *tx);
void dsl_deadlist_free(objset_t *os, uint64_t dlobj, dmu_tx_t *tx);
void dsl_deadlist_insert(dsl_deadlist_t *dl, const blkptr_t *bp, dmu_tx_t *tx);
void dsl_deadlist_insert(dsl_deadlist_t *dl, const blkptr_t *bp,
boolean_t free, dmu_tx_t *tx);
int dsl_deadlist_insert_alloc_cb(void *arg, const blkptr_t *bp, dmu_tx_t *tx);
int dsl_deadlist_insert_free_cb(void *arg, const blkptr_t *bp, dmu_tx_t *tx);
void dsl_deadlist_add_key(dsl_deadlist_t *dl, uint64_t mintxg, dmu_tx_t *tx);
void dsl_deadlist_remove_key(dsl_deadlist_t *dl, uint64_t mintxg, dmu_tx_t *tx);
void dsl_deadlist_remove_entry(dsl_deadlist_t *dl, uint64_t mintxg,
dmu_tx_t *tx);
dsl_deadlist_entry_t *dsl_deadlist_first(dsl_deadlist_t *dl);
dsl_deadlist_entry_t *dsl_deadlist_last(dsl_deadlist_t *dl);
uint64_t dsl_deadlist_clone(dsl_deadlist_t *dl, uint64_t maxtxg,
uint64_t mrs_obj, dmu_tx_t *tx);
void dsl_deadlist_space(dsl_deadlist_t *dl,
@@ -81,6 +104,10 @@ void dsl_deadlist_merge(dsl_deadlist_t *dl, uint64_t obj, dmu_tx_t *tx);
void dsl_deadlist_move_bpobj(dsl_deadlist_t *dl, bpobj_t *bpo, uint64_t mintxg,
dmu_tx_t *tx);
boolean_t dsl_deadlist_is_open(dsl_deadlist_t *dl);
int dsl_process_sub_livelist(bpobj_t *bpobj, struct bplist *to_free,
zthr_t *t, uint64_t *size);
void dsl_deadlist_clear_entry(dsl_deadlist_entry_t *dle, dsl_deadlist_t *dl,
dmu_tx_t *tx);
#ifdef __cplusplus
}