msc: detect RTP to CN failure in call tests
When CN RTP is missing, the X2 timer will fire after all other call
signalling looks successful. So far we establish an MT call, wait three
seconds and directly disconnect, long before X2 or X2427 can fire.
Make X2 shorter. (By means of f_vty_config() from ttcn, so that we don't
need to edit various osmo-msc.cfg in various repositories. The short
timer is always critical for the test to be accurate.)
Add proper function to detect premature disconnect. Otherwise we just
get a cryptic message like "Couldn't find MnccExpect for incoming call"
because of MNCC messaging after the unexpected release event.
Change-Id: I3ccf541360cc8440e664f0e29494b9ce7b6f8943
diff --git a/msc/MSC_Tests.ttcn b/msc/MSC_Tests.ttcn
index 0490901..8fc212c 100644
--- a/msc/MSC_Tests.ttcn
+++ b/msc/MSC_Tests.ttcn
@@ -324,6 +324,10 @@
f_vty_config(MSCVTY, "msc", "osmux off");
}
+ /* Configure the MGCP timeout so that a failure to set up all RTP streams triggers within the time that we keep
+ * an otherwise established call open. */
+ f_vty_config(MSCVTY, "msc", "timer mgw X2 3");
+
for (var integer i := 0; i < num_bsc; i := i + 1) {
if (isbound(mp_bssap_cfg[i])) {
var RanOps ranops := BSC_RanOps;