mirror_zfs/module/zcommon/zfeature_common.c
Tom Caputi 80a91e7469 Defer new resilvers until the current one ends
Currently, if a resilver is triggered for any reason while an
existing one is running, zfs will immediately restart the existing
resilver from the beginning to include the new drive. This causes
problems for system administrators when a drive fails while another
is already resilvering. In this case, the optimal thing to do to
reduce risk of data loss is to wait for the current resilver to end
before immediately replacing the second failed drive, which allows
the system to operate with two incomplete drives for the minimum
amount of time.

This patch introduces the resilver_defer feature that essentially
does this for the admin without forcing them to wait and monitor
the resilver manually. The change requires an on-disk feature
since we must mark drives that are part of a deferred resilver in
the vdev config to ensure that we do not assume they are done
resilvering when an existing resilver completes.

Reviewed-by: Matthew Ahrens <mahrens@delphix.com>
Reviewed-by: John Kennedy <john.kennedy@delphix.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: @mmaybee 
Signed-off-by: Tom Caputi <tcaputi@datto.com>
Closes #7732
2018-10-18 21:06:18 -07:00

464 lines
13 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) 2011, 2018 by Delphix. All rights reserved.
* Copyright (c) 2013 by Saso Kiselkov. All rights reserved.
* Copyright (c) 2013, Joyent, Inc. All rights reserved.
* Copyright (c) 2014, Nexenta Systems, Inc. All rights reserved.
* Copyright (c) 2017, Intel Corporation.
*/
#ifndef _KERNEL
#include <errno.h>
#include <string.h>
#include <sys/stat.h>
#endif
#include <sys/debug.h>
#include <sys/fs/zfs.h>
#include <sys/inttypes.h>
#include <sys/types.h>
#include <sys/zfs_sysfs.h>
#include "zfeature_common.h"
/*
* Set to disable all feature checks while opening pools, allowing pools with
* unsupported features to be opened. Set for testing only.
*/
boolean_t zfeature_checks_disable = B_FALSE;
zfeature_info_t spa_feature_table[SPA_FEATURES];
/*
* Valid characters for feature guids. This list is mainly for aesthetic
* purposes and could be expanded in the future. There are different allowed
* characters in the guids reverse dns portion (before the colon) and its
* short name (after the colon).
*/
static int
valid_char(char c, boolean_t after_colon)
{
return ((c >= 'a' && c <= 'z') ||
(c >= '0' && c <= '9') ||
(after_colon && c == '_') ||
(!after_colon && (c == '.' || c == '-')));
}
/*
* Every feature guid must contain exactly one colon which separates a reverse
* dns organization name from the feature's "short" name (e.g.
* "com.company:feature_name").
*/
boolean_t
zfeature_is_valid_guid(const char *name)
{
int i;
boolean_t has_colon = B_FALSE;
i = 0;
while (name[i] != '\0') {
char c = name[i++];
if (c == ':') {
if (has_colon)
return (B_FALSE);
has_colon = B_TRUE;
continue;
}
if (!valid_char(c, has_colon))
return (B_FALSE);
}
return (has_colon);
}
boolean_t
zfeature_is_supported(const char *guid)
{
if (zfeature_checks_disable)
return (B_TRUE);
for (spa_feature_t i = 0; i < SPA_FEATURES; i++) {
zfeature_info_t *feature = &spa_feature_table[i];
if (strcmp(guid, feature->fi_guid) == 0)
return (B_TRUE);
}
return (B_FALSE);
}
int
zfeature_lookup_guid(const char *guid, spa_feature_t *res)
{
for (spa_feature_t i = 0; i < SPA_FEATURES; i++) {
zfeature_info_t *feature = &spa_feature_table[i];
if (!feature->fi_zfs_mod_supported)
continue;
if (strcmp(guid, feature->fi_guid) == 0) {
if (res != NULL)
*res = i;
return (0);
}
}
return (ENOENT);
}
int
zfeature_lookup_name(const char *name, spa_feature_t *res)
{
for (spa_feature_t i = 0; i < SPA_FEATURES; i++) {
zfeature_info_t *feature = &spa_feature_table[i];
if (!feature->fi_zfs_mod_supported)
continue;
if (strcmp(name, feature->fi_uname) == 0) {
if (res != NULL)
*res = i;
return (0);
}
}
return (ENOENT);
}
boolean_t
zfeature_depends_on(spa_feature_t fid, spa_feature_t check)
{
zfeature_info_t *feature = &spa_feature_table[fid];
for (int i = 0; feature->fi_depends[i] != SPA_FEATURE_NONE; i++) {
if (feature->fi_depends[i] == check)
return (B_TRUE);
}
return (B_FALSE);
}
static boolean_t
deps_contains_feature(const spa_feature_t *deps, const spa_feature_t feature)
{
for (int i = 0; deps[i] != SPA_FEATURE_NONE; i++)
if (deps[i] == feature)
return (B_TRUE);
return (B_FALSE);
}
#if !defined(_KERNEL) && !defined(LIB_ZPOOL_BUILD)
static boolean_t
zfs_mod_supported_impl(const char *scope, const char *name, const char *sysfs)
{
struct stat64 statbuf;
char *path;
boolean_t supported = B_FALSE;
int len;
len = asprintf(&path, "%s/%s/%s", sysfs, scope, name);
if (len > 0) {
supported = !!(stat64(path, &statbuf) == 0);
free(path);
}
return (supported);
}
boolean_t
zfs_mod_supported(const char *scope, const char *name)
{
return (zfs_mod_supported_impl(scope, name, ZFS_SYSFS_DIR) ||
zfs_mod_supported_impl(scope, name, ZFS_SYSFS_ALT_DIR));
}
#endif
static boolean_t
zfs_mod_supported_feature(const char *name)
{
/*
* The zfs module spa_feature_table[], whether in-kernel or in
* libzpool, always supports all the features. libzfs needs to
* query the running module, via sysfs, to determine which
* features are supported.
*/
#if defined(_KERNEL) || defined(LIB_ZPOOL_BUILD)
return (B_TRUE);
#else
return (zfs_mod_supported(ZFS_SYSFS_POOL_FEATURES, name));
#endif
}
static void
zfeature_register(spa_feature_t fid, const char *guid, const char *name,
const char *desc, zfeature_flags_t flags, zfeature_type_t type,
const spa_feature_t *deps)
{
zfeature_info_t *feature = &spa_feature_table[fid];
static spa_feature_t nodeps[] = { SPA_FEATURE_NONE };
ASSERT(name != NULL);
ASSERT(desc != NULL);
ASSERT((flags & ZFEATURE_FLAG_READONLY_COMPAT) == 0 ||
(flags & ZFEATURE_FLAG_MOS) == 0);
ASSERT3U(fid, <, SPA_FEATURES);
ASSERT(zfeature_is_valid_guid(guid));
if (deps == NULL)
deps = nodeps;
VERIFY(((flags & ZFEATURE_FLAG_PER_DATASET) == 0) ||
(deps_contains_feature(deps, SPA_FEATURE_EXTENSIBLE_DATASET)));
feature->fi_feature = fid;
feature->fi_guid = guid;
feature->fi_uname = name;
feature->fi_desc = desc;
feature->fi_flags = flags;
feature->fi_type = type;
feature->fi_depends = deps;
feature->fi_zfs_mod_supported = zfs_mod_supported_feature(guid);
}
void
zpool_feature_init(void)
{
zfeature_register(SPA_FEATURE_ASYNC_DESTROY,
"com.delphix:async_destroy", "async_destroy",
"Destroy filesystems asynchronously.",
ZFEATURE_FLAG_READONLY_COMPAT, ZFEATURE_TYPE_BOOLEAN, NULL);
zfeature_register(SPA_FEATURE_EMPTY_BPOBJ,
"com.delphix:empty_bpobj", "empty_bpobj",
"Snapshots use less space.",
ZFEATURE_FLAG_READONLY_COMPAT, ZFEATURE_TYPE_BOOLEAN, NULL);
zfeature_register(SPA_FEATURE_LZ4_COMPRESS,
"org.illumos:lz4_compress", "lz4_compress",
"LZ4 compression algorithm support.",
ZFEATURE_FLAG_ACTIVATE_ON_ENABLE, ZFEATURE_TYPE_BOOLEAN, NULL);
zfeature_register(SPA_FEATURE_MULTI_VDEV_CRASH_DUMP,
"com.joyent:multi_vdev_crash_dump", "multi_vdev_crash_dump",
"Crash dumps to multiple vdev pools.",
0, ZFEATURE_TYPE_BOOLEAN, NULL);
zfeature_register(SPA_FEATURE_SPACEMAP_HISTOGRAM,
"com.delphix:spacemap_histogram", "spacemap_histogram",
"Spacemaps maintain space histograms.",
ZFEATURE_FLAG_READONLY_COMPAT, ZFEATURE_TYPE_BOOLEAN, NULL);
zfeature_register(SPA_FEATURE_ENABLED_TXG,
"com.delphix:enabled_txg", "enabled_txg",
"Record txg at which a feature is enabled",
ZFEATURE_FLAG_READONLY_COMPAT, ZFEATURE_TYPE_BOOLEAN, NULL);
{
static const spa_feature_t hole_birth_deps[] = {
SPA_FEATURE_ENABLED_TXG,
SPA_FEATURE_NONE
};
zfeature_register(SPA_FEATURE_HOLE_BIRTH,
"com.delphix:hole_birth", "hole_birth",
"Retain hole birth txg for more precise zfs send",
ZFEATURE_FLAG_MOS | ZFEATURE_FLAG_ACTIVATE_ON_ENABLE,
ZFEATURE_TYPE_BOOLEAN, hole_birth_deps);
}
zfeature_register(SPA_FEATURE_POOL_CHECKPOINT,
"com.delphix:zpool_checkpoint", "zpool_checkpoint",
"Pool state can be checkpointed, allowing rewind later.",
ZFEATURE_FLAG_READONLY_COMPAT, ZFEATURE_TYPE_BOOLEAN, NULL);
zfeature_register(SPA_FEATURE_SPACEMAP_V2,
"com.delphix:spacemap_v2", "spacemap_v2",
"Space maps representing large segments are more efficient.",
ZFEATURE_FLAG_READONLY_COMPAT | ZFEATURE_FLAG_ACTIVATE_ON_ENABLE,
ZFEATURE_TYPE_BOOLEAN, NULL);
zfeature_register(SPA_FEATURE_EXTENSIBLE_DATASET,
"com.delphix:extensible_dataset", "extensible_dataset",
"Enhanced dataset functionality, used by other features.",
0, ZFEATURE_TYPE_BOOLEAN, NULL);
{
static const spa_feature_t bookmarks_deps[] = {
SPA_FEATURE_EXTENSIBLE_DATASET,
SPA_FEATURE_NONE
};
zfeature_register(SPA_FEATURE_BOOKMARKS,
"com.delphix:bookmarks", "bookmarks",
"\"zfs bookmark\" command",
ZFEATURE_FLAG_READONLY_COMPAT, ZFEATURE_TYPE_BOOLEAN,
bookmarks_deps);
}
{
static const spa_feature_t filesystem_limits_deps[] = {
SPA_FEATURE_EXTENSIBLE_DATASET,
SPA_FEATURE_NONE
};
zfeature_register(SPA_FEATURE_FS_SS_LIMIT,
"com.joyent:filesystem_limits", "filesystem_limits",
"Filesystem and snapshot limits.",
ZFEATURE_FLAG_READONLY_COMPAT, ZFEATURE_TYPE_BOOLEAN,
filesystem_limits_deps);
}
zfeature_register(SPA_FEATURE_EMBEDDED_DATA,
"com.delphix:embedded_data", "embedded_data",
"Blocks which compress very well use even less space.",
ZFEATURE_FLAG_MOS | ZFEATURE_FLAG_ACTIVATE_ON_ENABLE,
ZFEATURE_TYPE_BOOLEAN, NULL);
{
static const spa_feature_t large_blocks_deps[] = {
SPA_FEATURE_EXTENSIBLE_DATASET,
SPA_FEATURE_NONE
};
zfeature_register(SPA_FEATURE_LARGE_BLOCKS,
"org.open-zfs:large_blocks", "large_blocks",
"Support for blocks larger than 128KB.",
ZFEATURE_FLAG_PER_DATASET, ZFEATURE_TYPE_BOOLEAN,
large_blocks_deps);
}
{
static const spa_feature_t large_dnode_deps[] = {
SPA_FEATURE_EXTENSIBLE_DATASET,
SPA_FEATURE_NONE
};
zfeature_register(SPA_FEATURE_LARGE_DNODE,
"org.zfsonlinux:large_dnode", "large_dnode",
"Variable on-disk size of dnodes.",
ZFEATURE_FLAG_PER_DATASET, ZFEATURE_TYPE_BOOLEAN,
large_dnode_deps);
}
{
static const spa_feature_t sha512_deps[] = {
SPA_FEATURE_EXTENSIBLE_DATASET,
SPA_FEATURE_NONE
};
zfeature_register(SPA_FEATURE_SHA512,
"org.illumos:sha512", "sha512",
"SHA-512/256 hash algorithm.",
ZFEATURE_FLAG_PER_DATASET, ZFEATURE_TYPE_BOOLEAN,
sha512_deps);
}
{
static const spa_feature_t skein_deps[] = {
SPA_FEATURE_EXTENSIBLE_DATASET,
SPA_FEATURE_NONE
};
zfeature_register(SPA_FEATURE_SKEIN,
"org.illumos:skein", "skein",
"Skein hash algorithm.",
ZFEATURE_FLAG_PER_DATASET, ZFEATURE_TYPE_BOOLEAN,
skein_deps);
}
{
static const spa_feature_t edonr_deps[] = {
SPA_FEATURE_EXTENSIBLE_DATASET,
SPA_FEATURE_NONE
};
zfeature_register(SPA_FEATURE_EDONR,
"org.illumos:edonr", "edonr",
"Edon-R hash algorithm.",
ZFEATURE_FLAG_PER_DATASET, ZFEATURE_TYPE_BOOLEAN,
edonr_deps);
}
zfeature_register(SPA_FEATURE_DEVICE_REMOVAL,
"com.delphix:device_removal", "device_removal",
"Top-level vdevs can be removed, reducing logical pool size.",
ZFEATURE_FLAG_MOS, ZFEATURE_TYPE_BOOLEAN, NULL);
{
static const spa_feature_t obsolete_counts_deps[] = {
SPA_FEATURE_EXTENSIBLE_DATASET,
SPA_FEATURE_DEVICE_REMOVAL,
SPA_FEATURE_NONE
};
zfeature_register(SPA_FEATURE_OBSOLETE_COUNTS,
"com.delphix:obsolete_counts", "obsolete_counts",
"Reduce memory used by removed devices when their blocks are "
"freed or remapped.",
ZFEATURE_FLAG_READONLY_COMPAT, ZFEATURE_TYPE_BOOLEAN,
obsolete_counts_deps);
}
{
static const spa_feature_t userobj_accounting_deps[] = {
SPA_FEATURE_EXTENSIBLE_DATASET,
SPA_FEATURE_NONE
};
zfeature_register(SPA_FEATURE_USEROBJ_ACCOUNTING,
"org.zfsonlinux:userobj_accounting", "userobj_accounting",
"User/Group object accounting.",
ZFEATURE_FLAG_READONLY_COMPAT | ZFEATURE_FLAG_PER_DATASET,
ZFEATURE_TYPE_BOOLEAN, userobj_accounting_deps);
}
{
static const spa_feature_t encryption_deps[] = {
SPA_FEATURE_EXTENSIBLE_DATASET,
SPA_FEATURE_NONE
};
zfeature_register(SPA_FEATURE_ENCRYPTION,
"com.datto:encryption", "encryption",
"Support for dataset level encryption",
ZFEATURE_FLAG_PER_DATASET, ZFEATURE_TYPE_BOOLEAN,
encryption_deps);
}
{
static const spa_feature_t project_quota_deps[] = {
SPA_FEATURE_EXTENSIBLE_DATASET,
SPA_FEATURE_NONE
};
zfeature_register(SPA_FEATURE_PROJECT_QUOTA,
"org.zfsonlinux:project_quota", "project_quota",
"space/object accounting based on project ID.",
ZFEATURE_FLAG_READONLY_COMPAT | ZFEATURE_FLAG_PER_DATASET,
ZFEATURE_TYPE_BOOLEAN, project_quota_deps);
}
{
zfeature_register(SPA_FEATURE_ALLOCATION_CLASSES,
"org.zfsonlinux:allocation_classes", "allocation_classes",
"Support for separate allocation classes.",
ZFEATURE_FLAG_READONLY_COMPAT, ZFEATURE_TYPE_BOOLEAN, NULL);
}
zfeature_register(SPA_FEATURE_RESILVER_DEFER,
"com.datto:resilver_defer", "resilver_defer",
"Support for defering new resilvers when one is already running.",
ZFEATURE_FLAG_READONLY_COMPAT, ZFEATURE_TYPE_BOOLEAN, NULL);
}
#if defined(_KERNEL)
EXPORT_SYMBOL(zfeature_lookup_guid);
EXPORT_SYMBOL(zfeature_lookup_name);
EXPORT_SYMBOL(zfeature_is_supported);
EXPORT_SYMBOL(zfeature_is_valid_guid);
EXPORT_SYMBOL(zfeature_depends_on);
EXPORT_SYMBOL(zpool_feature_init);
EXPORT_SYMBOL(spa_feature_table);
#endif