Fix stale volume list for NetApp 7-mode ISCSI driver

While contacting filer through DFM in order to create volume from
snapshot, the operation may fail with an error "No entry in LUN
table for volume". Although the LUN representing the volume was
created on the filer, the LUN list was not refreshed in time, which
 caused an error. This fix handles this situation for creating
volume from snapshots.

Note that this fix adds the requirement that the driver host and
DFM machine should be reasonably time-synchronized.

Fixes bug 1095633

Change-Id: I77fff4c36a3af72d28f2d01988a6067919093718

diff --git a/cinder/volume/drivers/netapp/iscsi.py b/cinder/volume/drivers/netapp/iscsi.py
index 701af63a0845f03d6838a0d12fc9ab09798eec04..d97c9e5c24154535b3a9e5823ad9b39d1cdbe5e2 100644 (file)
--- a/cinder/volume/drivers/netapp/iscsi.py
+++ b/cinder/volume/drivers/netapp/iscsi.py
@@ -910,15 +910,42 @@ class NetAppISCSIDriver(driver.ISCSIDriver):
     def _refresh_dfm_luns(self, host_id):
         """Refresh the LUN list for one filer in DFM."""
         server = self.client.service
-        server.DfmObjectRefresh(ObjectNameOrId=host_id, ChildType='lun_path')
+        refresh_started_at = time.time()
+        monitor_names = self.client.factory.create('ArrayOfMonitorName')
+        monitor_names.MonitorName = ['file_system', 'lun']
+        server.DfmObjectRefresh(ObjectNameOrId=host_id,
+                                MonitorNames=monitor_names)
+
+        max_wait = 10 * 60   # 10 minutes
+
         while True:
+            if time.time() - refresh_started_at > max_wait:
+                msg = _('Failed to get LUN list. Is the DFM host'
+                        ' time-synchronized with Cinder host?')
+                raise exception.VolumeBackendAPIException(msg)
+
+            LOG.info('Refreshing LUN list on DFM...')
             time.sleep(15)
             res = server.DfmMonitorTimestampList(HostNameOrId=host_id)
-            for timestamp in res.DfmMonitoringTimestamp:
-                if 'lun' != timestamp.MonitorName:
-                    continue
-                if timestamp.LastMonitoringTimestamp:
-                    return
+            timestamps = dict((t.MonitorName, t.LastMonitoringTimestamp)
+                              for t in res.DfmMonitoringTimestamp)
+            ts_fs = timestamps['file_system']
+            ts_lun = timestamps['lun']
+
+            if ts_fs > refresh_started_at and ts_lun > refresh_started_at:
+                return          # both monitor jobs finished
+            elif ts_fs == 0 or ts_lun == 0:
+                pass            # lun or file_system is still in progress, wait
+            else:
+                monitor_names.MonitorName = []
+                if ts_fs <= refresh_started_at:
+                    monitor_names.MonitorName.append('file_system')
+                if ts_lun <= refresh_started_at:
+                    monitor_names.MonitorName.append('lun')
+                LOG.debug('Rerunning refresh for monitors: ' +
+                          str(monitor_names.MonitorName))
+                server.DfmObjectRefresh(ObjectNameOrId=host_id,
+                                        MonitorNames=monitor_names)
 
     def _destroy_lun(self, host_id, lun_path):
         """Destroy a LUN on the filer."""