Azure: Check if interface is up after sleep when trying to bring it up (#972)

When bringing interface up by unbinding and then binding hv_netvsc
driver, it might take a short delay after binding for the link to be
up. So before trying unbind/bind again after sleep, check if the link
is up. This is a corner case when a preprovisioned VM is reused and
the NICs are hot-attached.

2 files changed, 29 insertions, 0 deletions

diff --git a/cloudinit/sources/DataSourceAzure.py b/cloudinit/sources/DataSourceAzure.py
index 6df9934b..ba23139b 100755
--- a/cloudinit/sources/DataSourceAzure.py
+++ b/cloudinit/sources/DataSourceAzure.py
@@ -923,6 +923,16 @@ class DataSourceAzure(sources.DataSource):
 
             sleep(sleep_duration)
 
+            # Since we just did a unbind and bind, check again after sleep
+            # but before doing unbind and bind again to avoid races where the
+            # link might take a slight delay after bind to be up.
+            if self.distro.networking.is_up(ifname):
+                msg = ("Link is up after checking after sleeping for %d secs"
+                       " after %d attempts" %
+                       (sleep_duration, attempts))
+                report_diagnostic_event(msg, logger_func=LOG.info)
+                return
+
     @azure_ds_telemetry_reporter
     def _create_report_ready_marker(self):
         path = REPORTED_READY_MARKER_FILE
diff --git a/tests/unittests/test_datasource/test_azure.py b/tests/unittests/test_datasource/test_azure.py
index 63eaf384..03609c3d 100644
--- a/tests/unittests/test_datasource/test_azure.py
+++ b/tests/unittests/test_datasource/test_azure.py
@@ -2907,6 +2907,25 @@ class TestPreprovisioningHotAttachNics(CiTestCase):
         dsa.wait_for_link_up("eth0")
         self.assertEqual(1, m_is_link_up.call_count)
 
+    @mock.patch(MOCKPATH + 'net.is_up', autospec=True)
+    @mock.patch(MOCKPATH + 'util.write_file')
+    @mock.patch('cloudinit.net.read_sys_net')
+    @mock.patch('cloudinit.distros.networking.LinuxNetworking.try_set_link_up')
+    def test_wait_for_link_up_checks_link_after_sleep(
+            self, m_is_link_up, m_read_sys_net, m_writefile, m_is_up):
+        """Waiting for link to be up should return immediately if the link is
+           already up."""
+
+        distro_cls = distros.fetch('ubuntu')
+        distro = distro_cls('ubuntu', {}, self.paths)
+        dsa = dsaz.DataSourceAzure({}, distro=distro, paths=self.paths)
+        m_is_link_up.return_value = False
+        m_is_up.return_value = True
+
+        dsa.wait_for_link_up("eth0")
+        self.assertEqual(2, m_is_link_up.call_count)
+        self.assertEqual(1, m_is_up.call_count)
+
     @mock.patch(MOCKPATH + 'util.write_file')
     @mock.patch('cloudinit.net.read_sys_net')
     @mock.patch('cloudinit.distros.networking.LinuxNetworking.try_set_link_up')