Avoid looping over remote resource provisioned flag merge

As we purge the service status.json after provisioned flag merge,
the daemon retains the last known provisioned flag mtime and
unduly loops over remerge.

Also update the cached provisoned flag mtime, so the retained
value is up-to-date and we can skip the merge on the next loop
iterations.

lib/osvcd_mon.py

@@ -1637,6 +1637,7 @@ def merge_hb_data_provision(self):
                                           svc.svcname, resource.rid, str(remote.state), 
                                           nodename)
                             resource.write_is_provisioned_flag(remote.state, remote.mtime)
+                            shared.CLUSTER_DATA[rcEnv.nodename]["services"]["status"][svc.svcname]["resources"][resource.rid]["provisioned"]["mtime"] = remote.mtime
                             changed = True
                 if changed:
                     svc.purge_status_data_dump()