summaryrefslogtreecommitdiffstats
diff options
context:
space:
mode:
authorDaniel Willmann <dwillmann@sysmocom.de>2018-08-21 14:18:00 +0200
committerDaniel Willmann <dwillmann@sysmocom.de>2018-08-21 14:18:00 +0200
commite8018967dec009469488818106e0526b03399724 (patch)
tree20a2d2e477ee4dcf1a3a35c4e19d4bc10f526d92
parent56101c935eed9dc14de27f39d4177fe8a628c044 (diff)
MSC_Tests: Reliably pass TC_cr_before_reset
I was able to reproduce the sporadic cr_before_reset failures by running both osmo-{stp,msc}-master with --cpus 0.1. In that case the first test run would fail for me because no BSSMAP RESET ACK was seen. Sleeping some seconds before sending the messages will make sure all components are connected. Note: TC_cr_before_reset is the first test executed in control so this seems to be why it sometimes fails in jenkins. Change-Id: Id745470231950e0a284f8c231246d3719f7617cc
-rw-r--r--msc/MSC_Tests.ttcn2
1 files changed, 2 insertions, 0 deletions
diff --git a/msc/MSC_Tests.ttcn b/msc/MSC_Tests.ttcn
index 42ad27c..4bdbc1d 100644
--- a/msc/MSC_Tests.ttcn
+++ b/msc/MSC_Tests.ttcn
@@ -1765,6 +1765,8 @@ testcase TC_cr_before_reset() runs on MTC_CT {
f_bssap_start(g_bssap[0]);
+ f_sleep(3.0);
+
/* Make a blind connection attemt, to trigger the deadlock condition */
BSSAP_DIRECT.send(ts_BSSAP_CONNECT_req(g_bssap[0].sccp_addr_peer, g_bssap[0].sccp_addr_own, 1, omit));