diff --git a/libs/wlgen/wlgen/rta.py b/libs/wlgen/wlgen/rta.py
index 94acd60965a1d5b0923032d33c66a2d20326f280..1b792ffe92047d1b0471873c5a8959ce104e2036 100644
--- a/libs/wlgen/wlgen/rta.py
+++ b/libs/wlgen/wlgen/rta.py
@@ -119,6 +119,18 @@ class RTA(Workload):
         logging.info('%s',
                 "{" + ", ".join('"%r": %r' % (key, pload[key]) for key in pload) + "}")
 
+        # Sanity check calibration values for big.LITTLE systems
+        if 'bl' in target.modules:
+            bcpu = target.bl.bigs_online[0]
+            lcpu = target.bl.littles_online[0]
+            if pload[bcpu] > pload[lcpu]:
+                logging.warning("Calibration values reports big cores less "
+                                "capable than LITTLE cores")
+                raise RuntimeError("Calibration failed: try again or file a bug")
+            bigs_speedup = ((float(pload[lcpu]) / pload[bcpu]) - 1) * 100
+            logging.info("big cores are ~%.0f%% more capable than LITTLE cores",
+                        bigs_speedup)
+
         return pload
 
     def __postrun(self, params):
@@ -502,7 +514,7 @@ class RTA(Workload):
 
         self.rta_cmd  = self.target.executables_directory + '/rt-app'
         self.rta_conf = self.run_dir + '/' + self.json
-        self.command = '{0:s} {1:s}'.format(self.rta_cmd, self.rta_conf)
+        self.command = '{0:s} {1:s} 2>&1'.format(self.rta_cmd, self.rta_conf)
 
         # Set and return the test label
         self.test_label = '{0:s}_{1:02d}'.format(self.name, self.exc_id)