scrub: add option to scrub only recent data

Sponsored-By: Wasabi Technology, Inc.
Sponsored-By: Klara Inc.
Signed-off-by: Mariusz Zaborski <mariusz.zaborski@klarasystems.com>

cmd/zpool/zpool_main.c

@@ -506,7 +506,7 @@ get_usage(zpool_help_t idx)
 		return (gettext("\tinitialize [-c | -s | -u] [-w] <pool> "
 		    "[<device> ...]\n"));
 	case HELP_SCRUB:
-		return (gettext("\tscrub [-s | -p] [-w] [-e] [-C] "
+		return (gettext("\tscrub [-s | -p] [-w] [-e] [-C] [-R] "
 		    "<pool> ...\n"));
 	case HELP_RESILVER:
 		return (gettext("\tresilver <pool> ...\n"));
@@ -8398,6 +8398,7 @@ wait_callback(zpool_handle_t *zhp, void *data)
  *	-e	Only scrub blocks in the error log.
  *	-s	Stop.  Stops any in-progress scrub.
  *	-p	Pause. Pause in-progress scrub.
+ *	-R	Scrub only recent data.
  *	-w	Wait.  Blocks until scrub has completed.
  */
 int
@@ -8415,9 +8416,10 @@ zpool_do_scrub(int argc, char **argv)
 	boolean_t is_pause = B_FALSE;
 	boolean_t is_stop = B_FALSE;
 	boolean_t is_txg_continue = B_FALSE;
+	boolean_t is_recent_scrub = B_FALSE;
 
 	/* check options */
-	while ((c = getopt(argc, argv, "spweC")) != -1) {
+	while ((c = getopt(argc, argv, "spweCR")) != -1) {
 		switch (c) {
 		case 'C':
 			is_txg_continue = B_TRUE;
@@ -8431,6 +8433,9 @@ zpool_do_scrub(int argc, char **argv)
 		case 'p':
 			is_pause = B_TRUE;
 			break;
+		case 'R':
+			is_recent_scrub = B_TRUE;
+			break;
 		case 'w':
 			wait = B_TRUE;
 			break;
@@ -8457,11 +8462,26 @@ zpool_do_scrub(int argc, char **argv)
 		(void) fprintf(stderr, gettext("invalid option "
 		    "combination :-e and -C are mutually exclusive\n"));
 		usage(B_FALSE);
+	} else if (is_pause && is_recent_scrub) {
+		(void) fprintf(stderr, gettext("invalid option "
+		    "combination :-p and -R are mutually exclusive\n"));
+		usage(B_FALSE);
+	} else if (is_stop && is_recent_scrub) {
+		(void) fprintf(stderr, gettext("invalid option "
+		    "combination :-s and -R are mutually exclusive\n"));
+		usage(B_FALSE);
+	} else if (is_error_scrub && is_recent_scrub) {
+		(void) fprintf(stderr, gettext("invalid option "
+		    "combination :-e and -R are mutually exclusive\n"));
+		usage(B_FALSE);
 	} else {
-		if (is_error_scrub)
+		if (is_error_scrub) {
 			cb.cb_type = POOL_SCAN_ERRORSCRUB;
+		}
 
-		if (is_pause) {
+		if (is_recent_scrub) {
+			cb.cb_scrub_cmd = POOL_SCRUB_RECENT_TXGS;
+		} else if (is_pause) {
 			cb.cb_scrub_cmd = POOL_SCRUB_PAUSE;
 		} else if (is_stop) {
 			cb.cb_type = POOL_SCAN_NONE;

include/sys/dsl_scan.h

@@ -44,6 +44,7 @@ struct dsl_pool;
 struct dmu_tx;
 
 extern int zfs_scan_suspend_progress;
+extern uint_t zfs_scrub_recent_txgs;
 
 /*
  * All members of this structure must be uint64_t, for byteswap

include/sys/fs/zfs.h

@@ -1077,6 +1077,7 @@ typedef enum pool_scrub_cmd {
 	POOL_SCRUB_NORMAL = 0,
 	POOL_SCRUB_PAUSE,
 	POOL_SCRUB_FROM_LAST_TXG,
+	POOL_SCRUB_RECENT_TXGS,
 	POOL_SCRUB_FLAGS_END
 } pool_scrub_cmd_t;
 

lib/libzfs/libzfs.abi

@@ -183,8 +183,8 @@
     <elf-symbol name='fsleep' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/>
     <elf-symbol name='get_dataset_depth' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/>
     <elf-symbol name='get_system_hostid' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/>
-    <elf-symbol name='getexecname' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/>
     <elf-symbol name='get_timestamp' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/>
+    <elf-symbol name='getexecname' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/>
     <elf-symbol name='getextmntent' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/>
     <elf-symbol name='getmntany' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/>
     <elf-symbol name='getprop_uint64' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/>
@@ -466,6 +466,7 @@
     <elf-symbol name='zpool_clear' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/>
     <elf-symbol name='zpool_clear_label' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/>
     <elf-symbol name='zpool_close' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/>
+    <elf-symbol name='zpool_collect_unsup_feat' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/>
     <elf-symbol name='zpool_create' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/>
     <elf-symbol name='zpool_default_search_paths' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/>
     <elf-symbol name='zpool_destroy' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/>
@@ -485,8 +486,8 @@
     <elf-symbol name='zpool_export_force' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/>
     <elf-symbol name='zpool_feature_init' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/>
     <elf-symbol name='zpool_find_config' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/>
-    <elf-symbol name='zpool_find_vdev' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/>
     <elf-symbol name='zpool_find_parent_vdev' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/>
+    <elf-symbol name='zpool_find_vdev' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/>
     <elf-symbol name='zpool_find_vdev_by_physpath' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/>
     <elf-symbol name='zpool_free_handles' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/>
     <elf-symbol name='zpool_get_all_vdev_props' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/>
@@ -529,7 +530,6 @@
     <elf-symbol name='zpool_prefetch' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/>
     <elf-symbol name='zpool_prepare_and_label_disk' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/>
     <elf-symbol name='zpool_prepare_disk' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/>
-    <elf-symbol name='zpool_collect_unsup_feat' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/>
     <elf-symbol name='zpool_prop_align_right' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/>
     <elf-symbol name='zpool_prop_column_name' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/>
     <elf-symbol name='zpool_prop_default_numeric' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/>
@@ -5809,7 +5809,8 @@
       <enumerator name='POOL_SCRUB_NORMAL' value='0'/>
       <enumerator name='POOL_SCRUB_PAUSE' value='1'/>
       <enumerator name='POOL_SCRUB_FROM_LAST_TXG' value='2'/>
-      <enumerator name='POOL_SCRUB_FLAGS_END' value='3'/>
+      <enumerator name='POOL_SCRUB_RECENT_TXGS' value='3'/>
+      <enumerator name='POOL_SCRUB_FLAGS_END' value='4'/>
     </enum-decl>
     <typedef-decl name='pool_scrub_cmd_t' type-id='a1474cbd' id='b51cf3c2'/>
     <enum-decl name='zpool_errata' id='d9abbf54'>
@@ -7839,7 +7840,7 @@
       </data-member>
     </class-decl>
     <typedef-decl name='vdev_cbdata_t' type-id='b8006be8' id='a9679c94'/>
-    <class-decl name='zprop_get_cbdata' size-in-bits='832' is-struct='yes' visibility='default' id='f3d3c319'>
+    <class-decl name='zprop_get_cbdata' size-in-bits='960' is-struct='yes' visibility='default' id='f3d3c319'>
       <data-member access='public' layout-offset-in-bits='0'>
         <var-decl name='cb_sources' type-id='95e97e5e' visibility='default'/>
       </data-member>
@@ -7858,6 +7859,9 @@
       <data-member access='public' layout-offset-in-bits='448'>
         <var-decl name='cb_first' type-id='c19b74c3' visibility='default'/>
       </data-member>
+      <data-member access='public' layout-offset-in-bits='480'>
+        <var-decl name='cb_json' type-id='c19b74c3' visibility='default'/>
+      </data-member>
       <data-member access='public' layout-offset-in-bits='512'>
         <var-decl name='cb_proplist' type-id='3a9b2288' visibility='default'/>
       </data-member>
@@ -7867,6 +7871,15 @@
       <data-member access='public' layout-offset-in-bits='640'>
         <var-decl name='cb_vdevs' type-id='a9679c94' visibility='default'/>
       </data-member>
+      <data-member access='public' layout-offset-in-bits='832'>
+        <var-decl name='cb_jsobj' type-id='5ce45b60' visibility='default'/>
+      </data-member>
+      <data-member access='public' layout-offset-in-bits='896'>
+        <var-decl name='cb_json_as_int' type-id='c19b74c3' visibility='default'/>
+      </data-member>
+      <data-member access='public' layout-offset-in-bits='928'>
+        <var-decl name='cb_json_pool_key_guid' type-id='c19b74c3' visibility='default'/>
+      </data-member>
     </class-decl>
     <typedef-decl name='zprop_get_cbdata_t' type-id='f3d3c319' id='f3d87113'/>
     <typedef-decl name='zprop_func' type-id='2e711a2a' id='1ec3747a'/>
@@ -7970,6 +7983,11 @@
     <qualified-type-def type-id='d33f11cb' restrict='yes' id='5c53ba29'/>
     <pointer-type-def type-id='ffa52b96' size-in-bits='64' id='76c8174b'/>
     <pointer-type-def type-id='f3d87113' size-in-bits='64' id='0d2a0670'/>
+    <function-decl name='nvlist_print_json' visibility='default' binding='global' size-in-bits='64'>
+      <parameter type-id='822cd80b'/>
+      <parameter type-id='5ce45b60'/>
+      <return type-id='95e97e5e'/>
+    </function-decl>
     <function-decl name='zpool_label_disk' mangled-name='zpool_label_disk' visibility='default' binding='global' size-in-bits='64' elf-symbol-id='zpool_label_disk'>
       <parameter type-id='b0382bb3'/>
       <parameter type-id='4c81de99'/>
@@ -8077,6 +8095,11 @@
       <parameter type-id='d33f11cb'/>
       <return type-id='48b5725f'/>
     </function-decl>
+    <function-decl name='putc' visibility='default' binding='global' size-in-bits='64'>
+      <parameter type-id='95e97e5e'/>
+      <parameter type-id='822cd80b'/>
+      <return type-id='95e97e5e'/>
+    </function-decl>
     <function-decl name='puts' visibility='default' binding='global' size-in-bits='64'>
       <parameter type-id='80f4b756'/>
       <return type-id='95e97e5e'/>
@@ -8095,6 +8118,11 @@
       <parameter type-id='95e97e5e'/>
       <return type-id='48b5725f'/>
     </function-decl>
+    <function-decl name='strspn' visibility='default' binding='global' size-in-bits='64'>
+      <parameter type-id='80f4b756'/>
+      <parameter type-id='80f4b756'/>
+      <return type-id='b59d7dce'/>
+    </function-decl>
     <function-decl name='strnlen' visibility='default' binding='global' size-in-bits='64'>
       <parameter type-id='80f4b756'/>
       <parameter type-id='b59d7dce'/>
@@ -8294,12 +8322,12 @@
     <function-decl name='zfs_version_print' mangled-name='zfs_version_print' visibility='default' binding='global' size-in-bits='64' elf-symbol-id='zfs_version_print'>
       <return type-id='95e97e5e'/>
     </function-decl>
-    <function-decl name='use_color' mangled-name='use_color' visibility='default' binding='global' size-in-bits='64' elf-symbol-id='use_color'>
-      <return type-id='95e97e5e'/>
-    </function-decl>
     <function-decl name='zfs_version_nvlist' mangled-name='zfs_version_nvlist' visibility='default' binding='global' size-in-bits='64' elf-symbol-id='zfs_version_nvlist'>
       <return type-id='5ce45b60'/>
     </function-decl>
+    <function-decl name='use_color' mangled-name='use_color' visibility='default' binding='global' size-in-bits='64' elf-symbol-id='use_color'>
+      <return type-id='95e97e5e'/>
+    </function-decl>
     <function-decl name='printf_color' mangled-name='printf_color' visibility='default' binding='global' size-in-bits='64' elf-symbol-id='printf_color'>
       <parameter type-id='80f4b756' name='color'/>
       <parameter type-id='80f4b756' name='format'/>
@@ -8804,11 +8832,6 @@
       <parameter type-id='78c01427'/>
       <return type-id='13956559'/>
     </function-decl>
-    <function-decl name='strspn' visibility='default' binding='global' size-in-bits='64'>
-      <parameter type-id='80f4b756'/>
-      <parameter type-id='80f4b756'/>
-      <return type-id='b59d7dce'/>
-    </function-decl>
     <function-decl name='zfs_dirnamelen' mangled-name='zfs_dirnamelen' visibility='default' binding='global' size-in-bits='64' elf-symbol-id='zfs_dirnamelen'>
       <parameter type-id='80f4b756' name='path'/>
       <return type-id='79a0948f'/>

man/man4/zfs.4

@@ -1986,6 +1986,9 @@ working on a scrub between TXG flushes.
 .It Sy zfs_scrub_error_blocks_per_txg Ns = Ns Sy 4096 Pq uint
 Error blocks to be scrubbed in one txg.
 .
+.It Sy zfs_scrub_recent_txgs Ns = Ns Sy 256 Pq uint
+Number of txgs to be considered as recent when performing a recent data scrub.
+.
 .It Sy zfs_scan_checkpoint_intval Ns = Ns Sy 7200 Ns s Po 2 hour Pc Pq uint
 To preserve progress across reboots, the sequential scan algorithm periodically
 needs to stop metadata scanning and issue all the verification I/O to disk.

man/man8/zpool-scrub.8

@@ -40,6 +40,7 @@
 .Op Fl w
 .Op Fl e
 .Op Fl C
+.Op Fl R
 .Ar pool Ns …
 .
 .Sh DESCRIPTION
@@ -119,6 +120,10 @@ resilvering, nor can it be run when a regular scrub is paused.
 Continue scrub from last saved txg (see zpool
 .Sy last_scrubbed_txg
 property).
+.It Fl R
+Scrub only recent data (this can be controlled by the
+.Sy zfs_scrub_recent_txgs
+parameter).
 .El
 .Sh EXAMPLES
 .Ss Example 1

module/zfs/dsl_scan.c

@@ -239,6 +239,9 @@ static int zfs_free_bpobj_enabled = 1;
 /* Error blocks to be scrubbed in one txg. */
 static uint_t zfs_scrub_error_blocks_per_txg = 1 << 12;
 
+/* The number of TXGs should be scrubbed while scrubbing recent data. */
+uint_t zfs_scrub_recent_txgs = 256;
+
 /* the order has to match pool_scan_type */
 static scan_cb_t *scan_funcs[POOL_SCAN_FUNCS] = {
 	NULL,
@@ -5323,4 +5326,7 @@ ZFS_MODULE_PARAM(zfs, zfs_, resilver_disable_defer, INT, ZMOD_RW,
 
 ZFS_MODULE_PARAM(zfs, zfs_, scrub_error_blocks_per_txg, UINT, ZMOD_RW,
 	"Error blocks to be scrubbed in one txg");
+
+ZFS_MODULE_PARAM(zfs, zfs_, scrub_recent_txgs, UINT, ZMOD_RW,
+	"The number of TXGs should be scrubbed while scrubbing recent data");
 /* END CSTYLED */

module/zfs/zfs_ioctl.c

@@ -1722,6 +1722,16 @@ zfs_ioc_pool_scrub(const char *poolname, nvlist_t *innvl, nvlist_t *outnvl)
 	} else if (scan_cmd == POOL_SCRUB_FROM_LAST_TXG) {
 		error = spa_scan_range(spa, scan_type,
 		    spa_get_last_scrubbed_txg(spa), 0);
+	} else if (scan_cmd == POOL_SCRUB_RECENT_TXGS) {
+		uint64_t start;
+
+		start = 0;
+		if (spa_last_synced_txg(spa) > zfs_scrub_recent_txgs) {
+			start = spa_last_synced_txg(spa) -
+			    zfs_scrub_recent_txgs;
+		}
+
+		error = spa_scan_range(spa, scan_type, start, 0);
 	} else {
 		error = spa_scan(spa, scan_type);
 	}