[ONOS-7594] Collect onos-diags when ping fails in SR test and other
minor fixes

Change-Id: I5ed5ab44264f7429bca746daad8c637e02094f41
diff --git a/TestON/tests/USECASE/SegmentRouting/SRBridging/dependencies/SRBridgingTest.py b/TestON/tests/USECASE/SegmentRouting/SRBridging/dependencies/SRBridgingTest.py
index d1a8ee3..9280ca8 100644
--- a/TestON/tests/USECASE/SegmentRouting/SRBridging/dependencies/SRBridgingTest.py
+++ b/TestON/tests/USECASE/SegmentRouting/SRBridging/dependencies/SRBridgingTest.py
@@ -73,8 +73,4 @@
             run.checkFlowsByDpid( main, dpid, self.topo[ topology ][ 4 ], sleep=5 )
         run.pingAll( main )
 
-        if hasattr( main, 'Mininet1' ):
-            run.cleanup( main )
-        else:
-            # TODO: disconnect TestON from the physical network
-            pass
+        run.cleanup( main )
diff --git a/TestON/tests/USECASE/SegmentRouting/SRClusterRestart/SRClusterRestart.params b/TestON/tests/USECASE/SegmentRouting/SRClusterRestart/SRClusterRestart.params
index 778370e..a991004 100755
--- a/TestON/tests/USECASE/SegmentRouting/SRClusterRestart/SRClusterRestart.params
+++ b/TestON/tests/USECASE/SegmentRouting/SRClusterRestart/SRClusterRestart.params
@@ -35,6 +35,7 @@
     <timers>
         <LinkDiscovery>12</LinkDiscovery>
         <SwitchDiscovery>12</SwitchDiscovery>
+        <OnosDiscovery>12</OnosDiscovery>
     </timers>
 
     <kill>
diff --git a/TestON/tests/USECASE/SegmentRouting/SRDhcprelay/dependencies/SRDhcprelayTest.py b/TestON/tests/USECASE/SegmentRouting/SRDhcprelay/dependencies/SRDhcprelayTest.py
index 1bbb0f9..9693add 100644
--- a/TestON/tests/USECASE/SegmentRouting/SRDhcprelay/dependencies/SRDhcprelayTest.py
+++ b/TestON/tests/USECASE/SegmentRouting/SRDhcprelay/dependencies/SRDhcprelayTest.py
@@ -72,8 +72,4 @@
         run.verifyOnosHostIp( main )
         run.verifyNetworkHostIp( main )
 
-        if hasattr( main, 'Mininet1' ):
-            run.cleanup( main )
-        else:
-            # TODO: disconnect TestON from the physical network
-            pass
+        run.cleanup( main )
diff --git a/TestON/tests/USECASE/SegmentRouting/SRDynamic/SRDynamic.params b/TestON/tests/USECASE/SegmentRouting/SRDynamic/SRDynamic.params
index 8beeccd..f15f207 100755
--- a/TestON/tests/USECASE/SegmentRouting/SRDynamic/SRDynamic.params
+++ b/TestON/tests/USECASE/SegmentRouting/SRDynamic/SRDynamic.params
@@ -35,6 +35,7 @@
     <timers>
         <LinkDiscovery>12</LinkDiscovery>
         <SwitchDiscovery>12</SwitchDiscovery>
+        <OnosDiscovery>12</OnosDiscovery>
     </timers>
 
     <kill>
diff --git a/TestON/tests/USECASE/SegmentRouting/SRDynamicConf/dependencies/SRDynamicConfTest.py b/TestON/tests/USECASE/SegmentRouting/SRDynamicConf/dependencies/SRDynamicConfTest.py
index 08ad86c..0341046 100644
--- a/TestON/tests/USECASE/SegmentRouting/SRDynamicConf/dependencies/SRDynamicConfTest.py
+++ b/TestON/tests/USECASE/SegmentRouting/SRDynamicConf/dependencies/SRDynamicConfTest.py
@@ -342,11 +342,7 @@
         run.checkFlows( main, minFlowCount=minFlowCountPerLeaf * topo[ topology ][ 1 ], sleep=5, dumpflows=False )
         run.pingAll( main, '%s_After' % TAG, retryAttempts=2 )
 
-        if hasattr( main, 'Mininet1' ):
-            run.cleanup( main )
-        else:
-            # TODO: disconnect TestON from the physical network
-            pass
+        run.cleanup( main )
 
     @staticmethod
     def updateIntfCfg( main, portNum, dualHomed, ips=[], untagged=0, tagged=[], native=0 ):
diff --git a/TestON/tests/USECASE/SegmentRouting/SRHighAvailability/SRHighAvailability.params b/TestON/tests/USECASE/SegmentRouting/SRHighAvailability/SRHighAvailability.params
index 0870eab..577edac 100644
--- a/TestON/tests/USECASE/SegmentRouting/SRHighAvailability/SRHighAvailability.params
+++ b/TestON/tests/USECASE/SegmentRouting/SRHighAvailability/SRHighAvailability.params
@@ -35,6 +35,7 @@
     <timers>
         <LinkDiscovery>12</LinkDiscovery>
         <SwitchDiscovery>12</SwitchDiscovery>
+        <OnosDiscovery>12</OnosDiscovery>
     </timers>
 
     <switches>
diff --git a/TestON/tests/USECASE/SegmentRouting/SROnosFailure/SROnosFailure.params b/TestON/tests/USECASE/SegmentRouting/SROnosFailure/SROnosFailure.params
index 778370e..a991004 100755
--- a/TestON/tests/USECASE/SegmentRouting/SROnosFailure/SROnosFailure.params
+++ b/TestON/tests/USECASE/SegmentRouting/SROnosFailure/SROnosFailure.params
@@ -35,6 +35,7 @@
     <timers>
         <LinkDiscovery>12</LinkDiscovery>
         <SwitchDiscovery>12</SwitchDiscovery>
+        <OnosDiscovery>12</OnosDiscovery>
     </timers>
 
     <kill>
diff --git a/TestON/tests/USECASE/SegmentRouting/SRRouting/SRRouting.params b/TestON/tests/USECASE/SegmentRouting/SRRouting/SRRouting.params
index 399152d..947666d 100644
--- a/TestON/tests/USECASE/SegmentRouting/SRRouting/SRRouting.params
+++ b/TestON/tests/USECASE/SegmentRouting/SRRouting/SRRouting.params
@@ -34,8 +34,13 @@
     </CTRL>
 
     <timers>
-        <LinkDiscovery>12</LinkDiscovery>
-        <SwitchDiscovery>12</SwitchDiscovery>
+        <LinkDiscovery>30</LinkDiscovery>
+        <SwitchDiscovery>30</SwitchDiscovery>
+        <OnosDiscovery>30</OnosDiscovery>
+        <loadNetcfgSleep>5</loadNetcfgSleep>
+        <startMininetSleep>25</startMininetSleep>
+        <dhcpSleep>60</dhcpSleep>
+        <balanceMasterSleep>10</balanceMasterSleep>
     </timers>
 
     <SLEEP>
diff --git a/TestON/tests/USECASE/SegmentRouting/SRRouting/dependencies/SRRoutingTest.py b/TestON/tests/USECASE/SegmentRouting/SRRouting/dependencies/SRRoutingTest.py
index df7c87d..0c65c1e 100644
--- a/TestON/tests/USECASE/SegmentRouting/SRRouting/dependencies/SRRoutingTest.py
+++ b/TestON/tests/USECASE/SegmentRouting/SRRouting/dependencies/SRRoutingTest.py
@@ -82,7 +82,7 @@
             run.loadLinkFailureChart( main )
 
         # wait some time
-        time.sleep( 5 )
+        time.sleep( float( main.params[ 'timers' ][ 'loadNetcfgSleep' ] ) )
 
         if hasattr( main, 'Mininet1' ):
             # Run the test with Mininet
@@ -94,9 +94,9 @@
             pass
 
         # wait some time for onos to install the rules!
-        time.sleep( 25 )
+        time.sleep( float( main.params[ 'timers' ][ 'startMininetSleep' ] ) )
         if ( dhcp ):
-            time.sleep( 60 )
+            time.sleep( float( main.params[ 'timers' ][ 'dhcpSleep' ] ) )
 
         SRRoutingTest.runChecks( main, test_idx, countFlowsGroups )
 
@@ -131,24 +131,18 @@
             for ctrl in xrange( numCtrls ):
                 # Kill node
                 run.killOnos( main, [ ctrl ], switches, links, ( numCtrls - 1 ) )
-                time.sleep( float( main.params[ 'timers' ][ 'SwitchDiscovery' ] ) )
                 main.Cluster.active(0).CLI.balanceMasters()
-                time.sleep( float( main.params[ 'timers' ][ 'SwitchDiscovery' ] ) )
+                time.sleep( float( main.params[ 'timers' ][ 'balanceMasterSleep' ] ) )
                 SRRoutingTest.runChecks( main, test_idx, countFlowsGroups )
 
                 # Recover node
                 run.recoverOnos( main, [ ctrl ], switches, links, numCtrls )
-                time.sleep( float( main.params[ 'timers' ][ 'SwitchDiscovery' ] ) )
                 main.Cluster.active(0).CLI.balanceMasters()
-                time.sleep( float( main.params[ 'timers' ][ 'SwitchDiscovery' ] ) )
+                time.sleep( float( main.params[ 'timers' ][ 'balanceMasterSleep' ] ) )
                 SRRoutingTest.runChecks( main, test_idx, countFlowsGroups )
 
         # Cleanup
-        if hasattr( main, 'Mininet1' ):
-            run.cleanup( main )
-        else:
-            # TODO: disconnect TestON from the physical network
-            pass
+        run.cleanup( main, copyKarafLog=False )
 
     @staticmethod
     def runChecks( main, test_idx, countFlowsGroups ):
@@ -159,4 +153,4 @@
         if countFlowsGroups:
             run.checkFlowsGroupsFromFile( main )
         # ping hosts
-        run.pingAll( main, 'CASE%03d' % test_idx, acceptableFailed=5, basedOnIp=True )
+        run.pingAll( main, 'CASE%03d' % test_idx, False, acceptableFailed=5, basedOnIp=True, skipOnFail=True )
diff --git a/TestON/tests/USECASE/SegmentRouting/dependencies/Testcaselib.py b/TestON/tests/USECASE/SegmentRouting/dependencies/Testcaselib.py
index 980db4d..62982c3 100644
--- a/TestON/tests/USECASE/SegmentRouting/dependencies/Testcaselib.py
+++ b/TestON/tests/USECASE/SegmentRouting/dependencies/Testcaselib.py
@@ -128,7 +128,7 @@
     @staticmethod
     def loadCount( main ):
         with open("%s/count/%s.count" % (main.configPath, main.cfgName)) as count:
-            main.count = json.load(count)
+                main.count = json.load(count)
 
     @staticmethod
     def loadJson( main ):
@@ -237,6 +237,17 @@
                                  onfail="Failed to assign switches to controllers" )
 
     @staticmethod
+    def saveOnosDiagnostics( main ):
+        """
+        Get onos-diags.tar.gz and save it to the log directory.
+        suffix: suffix string of the file name. E.g. onos-diags-case1.tar.gz
+        """
+        main.log.info( "Collecting onos-diags..." )
+        main.ONOSbench.onosDiagnostics( [ctrl.ipAddress for ctrl in main.Cluster.runningNodes],
+                                        main.logdir,
+                                        "-CASE%d" % main.CurrentTestCaseNumber )
+
+    @staticmethod
     def config( main, cfgName ):
         main.spines = []
 
@@ -408,7 +419,7 @@
         return
 
     @staticmethod
-    def pingAll( main, tag="", dumpflows=True, acceptableFailed=0, basedOnIp=False, sleep=10, retryAttempts=1 ):
+    def pingAll( main, tag="", dumpflows=True, acceptableFailed=0, basedOnIp=False, sleep=10, retryAttempts=1, skipOnFail=False ):
         '''
         Verify connectivity between hosts according to the ping chart
         acceptableFailed: max number of acceptable failed pings.
@@ -472,6 +483,10 @@
                     utilities.assert_equals( expect=expect, actual=pa,
                                              onpass="IP connectivity successfully tested",
                                              onfail="IP connectivity failed" )
+            if skipOnFail and pa != expect:
+                Testcaselib.saveOnosDiagnostics( main )
+                Testcaselib.cleanup( main )
+                main.skipCase()
 
         if dumpflows:
             main.ONOSbench.dumpONOSCmd( main.Cluster.active( 0 ).ipAddress,
@@ -650,7 +665,7 @@
                                  onfail="Failed to recover switch?" )
 
     @staticmethod
-    def cleanup( main, physical=False):
+    def cleanup( main, copyKarafLog=True ):
         """
         Stop Onos-cluster.
         Stops Mininet
@@ -666,10 +681,11 @@
         except ( NameError, AttributeError ):
             main.utils = Utils()
 
-        if not physical:
+        if hasattr( main, 'Mininet1' ):
             main.utils.mininetCleanup( main.Mininet1 )
 
-        main.utils.copyKarafLog( "CASE%d" % main.CurrentTestCaseNumber, before=True, includeCaseDesc=False )
+        if copyKarafLog:
+            main.utils.copyKarafLog( "CASE%d" % main.CurrentTestCaseNumber, before=True, includeCaseDesc=False )
 
         for ctrl in main.Cluster.active():
             main.ONOSbench.onosStop( ctrl.ipAddress )
@@ -682,6 +698,7 @@
         Completely Kill an ONOS instance and verify the ONOS cluster can see the proper change
         """
         main.step( "Killing ONOS instances with index(es): {}".format( nodes ) )
+        main.onosSleep = float( main.params[ 'timers' ][ 'OnosDiscovery' ] )
 
         for i in nodes:
             killResult = main.ONOSbench.onosDie( main.Cluster.runningNodes[ i ].ipAddress )
@@ -689,7 +706,7 @@
                                      onpass="ONOS instance Killed",
                                      onfail="Error killing ONOS instance" )
             main.Cluster.runningNodes[ i ].active = False
-        time.sleep( 12 )
+        time.sleep( main.onosSleep )
 
         if len( nodes ) < main.Cluster.numCtrls:
 
@@ -730,6 +747,7 @@
         """
         main.step( "Recovering ONOS instances with index(es): {}".format( nodes ) )
         [ main.ONOSbench.onosStart( main.Cluster.runningNodes[ i ].ipAddress ) for i in nodes ]
+        time.sleep( main.onosSleep )
         for i in nodes:
             isUp = main.ONOSbench.isup( main.Cluster.runningNodes[ i ].ipAddress )
             utilities.assert_equals( expect=main.TRUE, actual=isUp,