qa: wait for kernel client death

After sending the reboot command, we need to wait briefly for it to be
rebooted so that the kernel client doesn't voluntarily give up its Fb
cap.

Signed-off-by: Patrick Donnelly <pdonnell@redhat.com>
(cherry picked from commit 6b83f43ba0a99ad86850974d354b17c3e23877be)

diff --git a/qa/tasks/cephfs/test_client_recovery.py b/qa/tasks/cephfs/test_client_recovery.py
index 496b2faa1ec241386cd31b454ff92784b1426950..65760f76b9ffb80c9cb3a4da0c142d936868d3d4 100644 (file)
--- a/qa/tasks/cephfs/test_client_recovery.py
+++ b/qa/tasks/cephfs/test_client_recovery.py
@@ -299,6 +299,9 @@ class TestClientRecovery(CephFSTestCase):
         # Simulate client death
         self.mount_a.kill()
 
+        # wait for it to die so it doesn't voluntarily release buffer cap
+        time.sleep(5)
+
         try:
             # The waiter should get stuck waiting for the capability
             # held on the MDS by the now-dead client A