qa: unmount clients prior to marking fs down

Evicted RHEL7.5 clients may hang.

Fixes: http://tracker.ceph.com/issues/38677
Signed-off-by: Patrick Donnelly <pdonnell@redhat.com>
(cherry picked from commit 9aaf6118a4f259c9450af75d7b879ae2616bf77c)

Conflicts:
	qa/tasks/cephfs/test_failover.py
- mimic has a different set of tests: adapt

diff --git a/qa/tasks/cephfs/test_failover.py b/qa/tasks/cephfs/test_failover.py
index 48a4327058388b2c03dcbe1029150f395e404a7a..dd521167418b7e342ea9829fbd2097920db04f85 100644 (file)
--- a/qa/tasks/cephfs/test_failover.py
+++ b/qa/tasks/cephfs/test_failover.py
@@ -95,6 +95,8 @@ class TestClusterResize(CephFSTestCase):
         That marking a FS down twice does not wipe old_max_mds.
         """
 
+        self.mount_a.umount_wait()
+
         self.grow(2)
         self.fs.set_down()
         self.wait_for_health("MDS_ALL_DOWN", 30)
@@ -109,6 +111,8 @@ class TestClusterResize(CephFSTestCase):
         when actives come back online.
         """
 
+        self.mount_a.umount_wait()
+
         self.fs.set_down()
         self.wait_for_health("MDS_ALL_DOWN", 30)
         self.fs.set_down(False)