Multi-modifier protection (MMP)

Add multihost=on|off pool property to control MMP.  When enabled
a new thread writes uberblocks to the last slot in each label, at a
set frequency, to indicate to other hosts the pool is actively imported.
These uberblocks are the last synced uberblock with an updated
timestamp.  Property defaults to off.

During tryimport, find the "best" uberblock (newest txg and timestamp)
repeatedly, checking for change in the found uberblock.  Include the
results of the activity test in the config returned by tryimport.
These results are reported to user in "zpool import".

Allow the user to control the period between MMP writes, and the
duration of the activity test on import, via a new module parameter
zfs_multihost_interval.  The period is specified in milliseconds.  The
activity test duration is calculated from this value, and from the
mmp_delay in the "best" uberblock found initially.

Add a kstat interface to export statistics about Multiple Modifier
Protection (MMP) updates. Include the last synced txg number, the
timestamp, the delay since the last MMP update, the VDEV GUID, the VDEV
label that received the last MMP update, and the VDEV path.  Abbreviated
output below.

$ cat /proc/spl/kstat/zfs/mypool/multihost
31 0 0x01 10 880 105092382393521 105144180101111
txg   timestamp  mmp_delay   vdev_guid   vdev_label vdev_path
20468    261337  250274925   68396651780       3    /dev/sda
20468    261339  252023374   6267402363293     1    /dev/sdc
20468    261340  252000858   6698080955233     1    /dev/sdx
20468    261341  251980635   783892869810      2    /dev/sdy
20468    261342  253385953   8923255792467     3    /dev/sdd
20468    261344  253336622   042125143176      0    /dev/sdab
20468    261345  253310522   1200778101278     2    /dev/sde
20468    261346  253286429   0950576198362     2    /dev/sdt
20468    261347  253261545   96209817917       3    /dev/sds
20468    261349  253238188   8555725937673     3    /dev/sdb

Add a new tunable zfs_multihost_history to specify the number of MMP
updates to store history for. By default it is set to zero meaning that
no MMP statistics are stored.

When using ztest to generate activity, for automated tests of the MMP
function, some test functions interfere with the test.  For example, the
pool is exported to run zdb and then imported again.  Add a new ztest
function, "-M", to alter ztest behavior to prevent this.

Add new tests to verify the new functionality.  Tests provided by
Giuseppe Di Natale.

Reviewed by: Matthew Ahrens <[email protected]>
Reviewed-by: Giuseppe Di Natale <[email protected]>
Reviewed-by: Ned Bass <[email protected]>
Reviewed-by: Andreas Dilger <[email protected]>
Reviewed-by: Brian Behlendorf <[email protected]>
Signed-off-by: Olaf Faaland <[email protected]>
Closes #745
Closes #6279

1 files changed, 123 insertions, 30 deletions

diff --git a/cmd/zpool/zpool_main.c b/cmd/zpool/zpool_main.c
index 14181b083..052b429cc 100644
--- a/cmd/zpool/zpool_main.c
+++ b/cmd/zpool/zpool_main.c
@@ -1743,6 +1743,9 @@ print_status_config(zpool_handle_t *zhp, status_cbdata_t *cb, const char *name,
 			(void) printf(gettext("split into new pool"));
 			break;
 
+		case VDEV_AUX_ACTIVE:
+			(void) printf(gettext("currently in use"));
+
 		default:
 			(void) printf(gettext("corrupted data"));
 			break;
@@ -1838,6 +1841,10 @@ print_import_config(status_cbdata_t *cb, const char *name, nvlist_t *nv,
 			(void) printf(gettext("too many errors"));
 			break;
 
+		case VDEV_AUX_ACTIVE:
+			(void) printf(gettext("currently in use"));
+			break;
+
 		default:
 			(void) printf(gettext("corrupted data"));
 			break;
@@ -1935,8 +1942,10 @@ show_import(nvlist_t *config)
 	vdev_stat_t *vs;
 	char *name;
 	uint64_t guid;
+	uint64_t hostid = 0;
 	char *msgid;
-	nvlist_t *nvroot;
+	char *hostname = "unknown";
+	nvlist_t *nvroot, *nvinfo;
 	zpool_status_t reason;
 	zpool_errata_t errata;
 	const char *health;
@@ -2024,6 +2033,17 @@ show_import(nvlist_t *config)
 		zpool_print_unsup_feat(config);
 		break;
 
+	case ZPOOL_STATUS_HOSTID_ACTIVE:
+		(void) printf(gettext(" status: The pool is currently "
+		    "imported by another system.\n"));
+		break;
+
+	case ZPOOL_STATUS_HOSTID_REQUIRED:
+		(void) printf(gettext(" status: The pool has the "
+		    "multihost property on.  It cannot\n\tbe safely imported "
+		    "when the system hostid is not set.\n"));
+		break;
+
 	case ZPOOL_STATUS_HOSTID_MISMATCH:
 		(void) printf(gettext(" status: The pool was last accessed by "
 		    "another system.\n"));
@@ -2137,6 +2157,27 @@ show_import(nvlist_t *config)
 			    "imported. Attach the missing\n\tdevices and try "
 			    "again.\n"));
 			break;
+		case ZPOOL_STATUS_HOSTID_ACTIVE:
+			VERIFY0(nvlist_lookup_nvlist(config,
+			    ZPOOL_CONFIG_LOAD_INFO, &nvinfo));
+
+			if (nvlist_exists(nvinfo, ZPOOL_CONFIG_MMP_HOSTNAME))
+				hostname = fnvlist_lookup_string(nvinfo,
+				    ZPOOL_CONFIG_MMP_HOSTNAME);
+
+			if (nvlist_exists(nvinfo, ZPOOL_CONFIG_MMP_HOSTID))
+				hostid = fnvlist_lookup_uint64(nvinfo,
+				    ZPOOL_CONFIG_MMP_HOSTID);
+
+			(void) printf(gettext(" action: The pool must be "
+			    "exported from %s (hostid=%lx)\n\tbefore it "
+			    "can be safely imported.\n"), hostname,
+			    (unsigned long) hostid);
+			break;
+		case ZPOOL_STATUS_HOSTID_REQUIRED:
+			(void) printf(gettext(" action: Set a unique system "
+			    "hostid with the genhostid(1) command.\n"));
+			break;
 		default:
 			(void) printf(gettext(" action: The pool cannot be "
 			    "imported due to damaged devices or data.\n"));
@@ -2184,6 +2225,31 @@ show_import(nvlist_t *config)
 	}
 }
 
+static boolean_t
+zfs_force_import_required(nvlist_t *config)
+{
+	uint64_t state;
+	uint64_t hostid = 0;
+	nvlist_t *nvinfo;
+
+	state = fnvlist_lookup_uint64(config, ZPOOL_CONFIG_POOL_STATE);
+	(void) nvlist_lookup_uint64(config, ZPOOL_CONFIG_HOSTID, &hostid);
+
+	if (state != POOL_STATE_EXPORTED && hostid != get_system_hostid())
+		return (B_TRUE);
+
+	nvinfo = fnvlist_lookup_nvlist(config, ZPOOL_CONFIG_LOAD_INFO);
+	if (nvlist_exists(nvinfo, ZPOOL_CONFIG_MMP_STATE)) {
+		mmp_state_t mmp_state = fnvlist_lookup_uint64(nvinfo,
+		    ZPOOL_CONFIG_MMP_STATE);
+
+		if (mmp_state != MMP_STATE_INACTIVE)
+			return (B_TRUE);
+	}
+
+	return (B_FALSE);
+}
+
 /*
  * Perform the import for the given configuration.  This passes the heavy
  * lifting off to zpool_import_props(), and then mounts the datasets contained
@@ -2198,45 +2264,72 @@ do_import(nvlist_t *config, const char *newname, const char *mntopts,
 	uint64_t state;
 	uint64_t version;
 
-	verify(nvlist_lookup_string(config, ZPOOL_CONFIG_POOL_NAME,
-	    &name) == 0);
+	name = fnvlist_lookup_string(config, ZPOOL_CONFIG_POOL_NAME);
+	state = fnvlist_lookup_uint64(config, ZPOOL_CONFIG_POOL_STATE);
+	version = fnvlist_lookup_uint64(config, ZPOOL_CONFIG_VERSION);
 
-	verify(nvlist_lookup_uint64(config,
-	    ZPOOL_CONFIG_POOL_STATE, &state) == 0);
-	verify(nvlist_lookup_uint64(config,
-	    ZPOOL_CONFIG_VERSION, &version) == 0);
 	if (!SPA_VERSION_IS_SUPPORTED(version)) {
 		(void) fprintf(stderr, gettext("cannot import '%s': pool "
 		    "is formatted using an unsupported ZFS version\n"), name);
 		return (1);
-	} else if (state != POOL_STATE_EXPORTED &&
+	} else if (zfs_force_import_required(config) &&
 	    !(flags & ZFS_IMPORT_ANY_HOST)) {
-		uint64_t hostid = 0;
-		unsigned long system_hostid = get_system_hostid();
+		mmp_state_t mmp_state = MMP_STATE_INACTIVE;
+		nvlist_t *nvinfo;
 
-		(void) nvlist_lookup_uint64(config, ZPOOL_CONFIG_HOSTID,
-		    &hostid);
+		nvinfo = fnvlist_lookup_nvlist(config, ZPOOL_CONFIG_LOAD_INFO);
+		if (nvlist_exists(nvinfo, ZPOOL_CONFIG_MMP_STATE))
+			mmp_state = fnvlist_lookup_uint64(nvinfo,
+			    ZPOOL_CONFIG_MMP_STATE);
 
-		if (hostid != 0 && (unsigned long)hostid != system_hostid) {
-			char *hostname;
-			uint64_t timestamp;
-			time_t t;
+		if (mmp_state == MMP_STATE_ACTIVE) {
+			char *hostname = "<unknown>";
+			uint64_t hostid = 0;
 
-			verify(nvlist_lookup_string(config,
-			    ZPOOL_CONFIG_HOSTNAME, &hostname) == 0);
-			verify(nvlist_lookup_uint64(config,
-			    ZPOOL_CONFIG_TIMESTAMP, &timestamp) == 0);
-			t = timestamp;
-			(void) fprintf(stderr, gettext("cannot import "
-			    "'%s': pool may be in use from other "
-			    "system, it was last accessed by %s "
-			    "(hostid: 0x%lx) on %s"), name, hostname,
-			    (unsigned long)hostid,
-			    asctime(localtime(&t)));
-			(void) fprintf(stderr, gettext("use '-f' to "
-			    "import anyway\n"));
-			return (1);
+			if (nvlist_exists(nvinfo, ZPOOL_CONFIG_MMP_HOSTNAME))
+				hostname = fnvlist_lookup_string(nvinfo,
+				    ZPOOL_CONFIG_MMP_HOSTNAME);
+
+			if (nvlist_exists(nvinfo, ZPOOL_CONFIG_MMP_HOSTID))
+				hostid = fnvlist_lookup_uint64(nvinfo,
+				    ZPOOL_CONFIG_MMP_HOSTID);
+
+			(void) fprintf(stderr, gettext("cannot import '%s': "
+			    "pool is imported on %s (hostid: "
+			    "0x%lx)\nExport the pool on the other system, "
+			    "then run 'zpool import'.\n"),
+			    name, hostname, (unsigned long) hostid);
+		} else if (mmp_state == MMP_STATE_NO_HOSTID) {
+			(void) fprintf(stderr, gettext("Cannot import '%s': "
+			    "pool has the multihost property on and the\n"
+			    "system's hostid is not set. Set a unique hostid "
+			    "with the genhostid(1) command.\n"), name);
+		} else {
+			char *hostname = "<unknown>";
+			uint64_t timestamp = 0;
+			uint64_t hostid = 0;
+
+			if (nvlist_exists(config, ZPOOL_CONFIG_HOSTNAME))
+				hostname = fnvlist_lookup_string(config,
+				    ZPOOL_CONFIG_HOSTNAME);
+
+			if (nvlist_exists(config, ZPOOL_CONFIG_TIMESTAMP))
+				timestamp = fnvlist_lookup_uint64(config,
+				    ZPOOL_CONFIG_TIMESTAMP);
+
+			if (nvlist_exists(config, ZPOOL_CONFIG_HOSTID))
+				hostid = fnvlist_lookup_uint64(config,
+				    ZPOOL_CONFIG_HOSTID);
+
+			(void) fprintf(stderr, gettext("cannot import '%s': "
+			    "pool was previously in use from another system.\n"
+			    "Last accessed by %s (hostid=%lx) at %s"
+			    "The pool can be imported, use 'zpool import -f' "
+			    "to import the pool.\n"), name, hostname,
+			    (unsigned long)hostid, ctime((time_t *)&timestamp));
 		}
+
+		return (1);
 	}
 
 	if (zpool_import_props(g_zfs, config, newname, props, flags) != 0)