diff --git a/cmd/ztest/ztest.c b/cmd/ztest/ztest.c
index b4cedbdbaf66..b6102c8a5aba 100644
--- a/cmd/ztest/ztest.c
+++ b/cmd/ztest/ztest.c
@@ -326,6 +326,7 @@ ztest_func_t ztest_spa_create_destroy;
 ztest_func_t ztest_fault_inject;
 ztest_func_t ztest_ddt_repair;
 ztest_func_t ztest_dmu_snapshot_hold;
+ztest_func_t ztest_mmp_enable_disable;
 ztest_func_t ztest_spa_rename;
 ztest_func_t ztest_scrub;
 ztest_func_t ztest_dsl_dataset_promote_busy;
@@ -375,6 +376,7 @@ ztest_info_t ztest_info[] = {
 	ZTI_INIT(ztest_fault_inject, 1, &zopt_sometimes),
 	ZTI_INIT(ztest_ddt_repair, 1, &zopt_sometimes),
 	ZTI_INIT(ztest_dmu_snapshot_hold, 1, &zopt_sometimes),
+	ZTI_INIT(ztest_mmp_enable_disable, 1, &zopt_often),
 	ZTI_INIT(ztest_reguid, 1, &zopt_rarely),
 	ZTI_INIT(ztest_spa_rename, 1, &zopt_rarely),
 	ZTI_INIT(ztest_scrub, 1, &zopt_rarely),
@@ -2660,6 +2662,68 @@ ztest_spa_create_destroy(ztest_ds_t *zd, uint64_t id)
 	(void) rw_unlock(&ztest_name_lock);
 }
 
+/*
+ * Start and then stop the MMP threads to ensure the startup and shutdown code
+ * works properly.  Actual protection and property-related code tested via ZTS.
+ */
+/* ARGSUSED */
+void
+ztest_mmp_enable_disable(ztest_ds_t *zd, uint64_t id)
+{
+	ztest_shared_opts_t *zo = &ztest_opts;
+	spa_t *spa = NULL;
+
+	if (zo->zo_mmp_test)
+		return;
+
+	mutex_enter(&spa_namespace_lock);
+	if (ztest_spa) {
+		spa = ztest_spa;
+		spa_open_ref(spa, FTAG);
+	}
+	mutex_exit(&spa_namespace_lock);
+
+	if (spa) {
+		if (ztest_opts.zo_verbose >= 2)
+			printf("Possibly starting MMP thread.\n");
+		mutex_enter(&spa_namespace_lock);
+		spa_config_enter(spa, SCL_CONFIG, FTAG, RW_READER);
+		mutex_enter(&spa->spa_props_lock);
+		if (! spa_multihost(spa)) {
+			spa->spa_multihost = B_TRUE;
+			mmp_thread_start(spa);
+		}
+		mutex_exit(&spa->spa_props_lock);
+		spa_config_exit(spa, SCL_CONFIG, FTAG);
+		mutex_exit(&spa_namespace_lock);
+
+		if (ztest_opts.zo_verbose >= 2)
+			printf("Signalling all MMP threads.\n");
+		txg_wait_synced(spa_get_dsl(spa), 0);
+		mmp_signal_all_threads();
+		txg_wait_synced(spa_get_dsl(spa), 0);
+
+		if (ztest_opts.zo_verbose >= 2)
+			printf("Possibly stopping MMP thread.\n");
+		mutex_enter(&spa_namespace_lock);
+		spa_config_enter(spa, SCL_CONFIG, FTAG, RW_READER);
+		mutex_enter(&spa->spa_props_lock);
+		if (spa_multihost(spa)) {
+			mmp_thread_stop(spa);
+			spa->spa_multihost = B_FALSE;
+		}
+		mutex_exit(&spa->spa_props_lock);
+		spa_config_exit(spa, SCL_CONFIG, FTAG);
+		mutex_exit(&spa_namespace_lock);
+	}
+
+	mutex_enter(&spa_namespace_lock);
+	if (spa) {
+		spa_close(spa, FTAG);
+	}
+	mutex_exit(&spa_namespace_lock);
+}
+
 /* ARGSUSED */
 void
 ztest_spa_upgrade(ztest_ds_t *zd, uint64_t id)
diff --git a/include/sys/mmp.h b/include/sys/mmp.h
index 4da612d6a5a2..5b2fea1a66b1 100644
--- a/include/sys/mmp.h
+++ b/include/sys/mmp.h
@@ -50,6 +50,7 @@ extern void mmp_fini(struct spa *spa);
 extern void mmp_thread_start(struct spa *spa);
 extern void mmp_thread_stop(struct spa *spa);
 extern void mmp_update_uberblock(struct spa *spa, struct uberblock *ub);
+extern void mmp_signal_all_threads(void);
 
 /* Global tuning */
 extern ulong_t zfs_multihost_interval;
diff --git a/module/zfs/mmp.c b/module/zfs/mmp.c
index 35348f8b4dc3..7bddbbd97eee 100644
--- a/module/zfs/mmp.c
+++ b/module/zfs/mmp.c
@@ -458,13 +458,60 @@ mmp_thread(spa_t *spa)
 	mmp_thread_exit(mmp, &mmp->mmp_thread, &cpr);
 }
 
+/*
+ * Signal the MMP thread to wake it, when it is sleeping on
+ * its cv.  Used when some module parameter has changed and
+ * we want the thread to know about it.
+ * Only signal if the pool is active and mmp thread is
+ * running, otherwise there is no thread to wake.
+ */
+static void
+mmp_signal_thread(spa_t *spa)
+{
+	mmp_thread_t *mmp = &spa->spa_mmp;
+
+	mutex_enter(&mmp->mmp_thread_lock);
+	if (mmp->mmp_thread)
+		cv_broadcast(&mmp->mmp_thread_cv);
+	mutex_exit(&mmp->mmp_thread_lock);
+}
+
+void
+mmp_signal_all_threads(void)
+{
+	spa_t *spa = NULL;
+
+	mutex_enter(&spa_namespace_lock);
+	while ((spa = spa_next(spa)))
+		if (spa->spa_state == POOL_STATE_ACTIVE)
+			mmp_signal_thread(spa);
+	mutex_exit(&spa_namespace_lock);
+}
+
 #if defined(_KERNEL) && defined(HAVE_SPL)
+#include <linux/mod_compat.h>
+
+static int
+param_set_multihost_interval(const char *val, zfs_kernel_param_t *kp)
+{
+	int ret;
+
+	ret = param_set_ulong(val, kp);
+	if (ret < 0)
+		return (ret);
+
+	mmp_signal_all_threads();
+
+	return (ret);
+}
+
 /* BEGIN CSTYLED */
 module_param(zfs_multihost_fail_intervals, uint, 0644);
 MODULE_PARM_DESC(zfs_multihost_fail_intervals,
 	"Max allowed period without a successful mmp write");
 
-module_param(zfs_multihost_interval, ulong, 0644);
+module_param_call(zfs_multihost_interval, param_set_multihost_interval,
+    param_get_ulong, &zfs_multihost_interval, 0644);
 MODULE_PARM_DESC(zfs_multihost_interval,
 	"Milliseconds between mmp writes to each leaf");
 
diff --git a/tests/runfiles/linux.run b/tests/runfiles/linux.run
index e556f1e8cc1d..24b8f7584f37 100644
--- a/tests/runfiles/linux.run
+++ b/tests/runfiles/linux.run
@@ -412,7 +412,8 @@ tests = ['mmap_write_001_pos', 'mmap_read_001_pos']
 
 [tests/functional/mmp]
 tests = ['mmp_on_thread', 'mmp_on_uberblocks', 'mmp_on_off', 'mmp_interval',
-    'mmp_active_import', 'mmp_inactive_import', 'mmp_exported_import']
+    'mmp_active_import', 'mmp_inactive_import', 'mmp_exported_import',
+    'mmp_reset_interval']
 
 [tests/functional/mount]
 tests = ['umount_001', 'umountall_001']
diff --git a/tests/zfs-tests/tests/functional/mmp/Makefile.am b/tests/zfs-tests/tests/functional/mmp/Makefile.am
index 2399a6c377d2..564f4f71b34f 100644
--- a/tests/zfs-tests/tests/functional/mmp/Makefile.am
+++ b/tests/zfs-tests/tests/functional/mmp/Makefile.am
@@ -7,6 +7,7 @@ dist_pkgdata_SCRIPTS = \
 	mmp_active_import.ksh \
 	mmp_inactive_import.ksh \
 	mmp_exported_import.ksh \
+	mmp_reset_interval.ksh \
 	setup.ksh \
 	cleanup.ksh \
 	mmp.kshlib \
diff --git a/tests/zfs-tests/tests/functional/mmp/mmp.cfg b/tests/zfs-tests/tests/functional/mmp/mmp.cfg
index f17108a87630..29e030a4301a 100644
--- a/tests/zfs-tests/tests/functional/mmp/mmp.cfg
+++ b/tests/zfs-tests/tests/functional/mmp/mmp.cfg
@@ -31,7 +31,9 @@ export TXG_TIMEOUT_DEFAULT=5
 export MMP_POOL=mmppool
 export MMP_DIR=$TEST_BASE_DIR/mmp
 export MMP_HISTORY=100
+export MMP_HISTORY_OFF=0
 
+export MMP_INTERVAL_HOUR=$((60*60*1000))
 export MMP_INTERVAL_DEFAULT=1000
 export MMP_INTERVAL_MIN=100
 
diff --git a/tests/zfs-tests/tests/functional/mmp/mmp_reset_interval.ksh b/tests/zfs-tests/tests/functional/mmp/mmp_reset_interval.ksh
new file mode 100755
index 000000000000..0b58848651f7
--- /dev/null
+++ b/tests/zfs-tests/tests/functional/mmp/mmp_reset_interval.ksh
@@ -0,0 +1,67 @@
+#!/bin/ksh -p
+#
+# CDDL HEADER START
+#
+# This file and its contents are supplied under the terms of the
+# Common Development and Distribution License ("CDDL"), version 1.0.
+# You may only use this file in accordance with the terms of version
+# 1.0 of the CDDL.
+#
+# A full copy of the text of the CDDL should have accompanied this
+# source.  A copy of the CDDL is also available via the Internet at
+# http://www.illumos.org/license/CDDL.
+#
+# CDDL HEADER END
+#
+
+#
+# Copyright (c) 2017 by Lawrence Livermore National Security, LLC.
+#
+
+# DESCRIPTION:
+#	Ensure that the MMP thread is notified when zfs_multihost_interval is
+#	reduced.
+#
+# STRATEGY:
+#	1. Set zfs_multihost_interval to much longer than the test duration
+#	2. Create a zpool and enable multihost
+#       3. Verify no MMP writes occurred
+#	4. Set zfs_multihost_interval to 1 second
+#	5. Sleep briefly
+#	6. Verify MMP writes began
+#
+
+. $STF_SUITE/include/libtest.shlib
+. $STF_SUITE/tests/functional/mmp/mmp.cfg
+. $STF_SUITE/tests/functional/mmp/mmp.kshlib
+
+verify_runnable "both"
+
+function cleanup
+{
+	default_cleanup_noexit
+	log_must set_tunable64 zfs_multihost_interval $MMP_INTERVAL_DEFAULT
+	log_must mmp_clear_hostid
+}
+
+log_assert "mmp threads notified when zfs_multihost_interval reduced"
+log_onexit cleanup
+
+log_must set_tunable64 zfs_multihost_interval $MMP_INTERVAL_HOUR
+log_must mmp_set_hostid $HOSTID1
+
+default_setup_noexit $DISK
+log_must zpool set multihost=on $TESTPOOL
+
+prev_count=$(wc -l /proc/spl/kstat/zfs/$TESTPOOL/multihost)
+log_must set_tunable64 zfs_multihost_interval $MMP_INTERVAL_DEFAULT
+
+# slight delay to allow time for the mmp write to complete
+sleep 1
+curr_count=$(wc -l /proc/spl/kstat/zfs/$TESTPOOL/multihost)
+
+if [ $curr_count -eq $prev_count ]; then
+	log_fail "mmp writes did not start when zfs_multihost_interval reduced"
+fi
+
+log_pass "mmp threads notified when zfs_multihost_interval reduced"