neels has submitted this change. (
https://gerrit.osmocom.org/c/osmo-ttcn3-hacks/+/35098?usp=email )
Change subject: bsc: tests recovery from BORKEN after release failed
......................................................................
bsc: tests recovery from BORKEN after release failed
Related: osmo-bsc Ic4728b3efe843ea63e2a0b54b1ea8a925347484a
Related: SYS#6655
Change-Id: I9b4ddfc4a337808d9d5ec538c25fd390b1b2530f
---
M bsc/BSC_Tests.ttcn
1 file changed, 139 insertions(+), 3 deletions(-)
Approvals:
daniel: Looks good to me, but someone else must approve
laforge: Looks good to me, approved
Jenkins Builder: Verified
diff --git a/bsc/BSC_Tests.ttcn b/bsc/BSC_Tests.ttcn
index 57b44e0..1fc42ba 100644
--- a/bsc/BSC_Tests.ttcn
+++ b/bsc/BSC_Tests.ttcn
@@ -8529,7 +8529,8 @@
}
friend function f_perform_clear(RSL_DCHAN_PT rsl_pt := RSL, RSLEM_PROC_PT
rsl_proc_pt := RSL_PROC,
- template PDU_ML3_NW_MS exp_rr_rel_tmpl :=
tr_RRM_RR_RELEASE)
+ template PDU_ML3_NW_MS exp_rr_rel_tmpl :=
tr_RRM_RR_RELEASE,
+ boolean send_rel_ack := true)
runs on MSC_ConnHdlr {
var default ack_dlcx := activate(as_mgcp_ack_all_dlcx());
var default ack_rel_req := activate(as_rsl_ack_all_rel_req());
@@ -8549,8 +8550,10 @@
}
[] rsl_pt.receive(tr_RSL_RF_CHAN_REL(g_chan_nr)) {
f_logp(BSCVTY, "Got RSL RF Chan Rel, sending Rel Ack");
- rsl_pt.send(ts_RSL_RF_CHAN_REL_ACK(g_chan_nr));
- f_rslem_unregister(0, g_chan_nr, PT := rsl_proc_pt);
+ if (send_rel_ack) {
+ rsl_pt.send(ts_RSL_RF_CHAN_REL_ACK(g_chan_nr));
+ f_rslem_unregister(0, g_chan_nr, PT :=
rsl_proc_pt);
+ }
}
}
deactivate(ack_dlcx);
@@ -12439,6 +12442,125 @@
f_shutdown_helper(ho := true);
}
+private function f_vty_expect_borken(TELNETasp_PT pt, boolean expect_borken)
+{
+ var charstring lchan_summary;
+ lchan_summary := f_vty_transceive_ret(pt, "show lchan summary");
+ var boolean found_borken := (f_strstr(lchan_summary, "State BORKEN") >=
0);
+ if (found_borken != expect_borken) {
+ if (expect_borken) {
+ setverdict(fail, "Expected 'BORKEN' state in ",
lchan_summary);
+ } else {
+ setverdict(fail, "Expected no 'BORKEN' state in ",
lchan_summary);
+ }
+ mtc.stop;
+ }
+}
+
+private function f_tc_unbreak_lchan_after_missing_rel_ack(integer
test_variant) runs on MSC_ConnHdlr {
+
+ /* Establish a channel, so we can mess with the release of it. */
+ var template PDU_BSSAP exp_compl := f_gen_exp_compl();
+ var PDU_BSSAP ass_cmd := f_gen_ass_req();
+
+ ass_cmd.pdu.bssmap.assignmentRequest.channelType :=
valueof(ts_BSSMAP_IE_ChannelType);
+ ass_cmd.pdu.bssmap.assignmentRequest.codecList :=
valueof(ts_BSSMAP_IE_CodecList({ts_CodecFR}));
+
+ f_establish_fully(ass_cmd, exp_compl);
+
+ /* shorten X28, the time after which a BORKEN lchan recovers, from 30s
to something short */
+ f_vty_enter_cfg_network(BSCVTY);
+ f_vty_transceive(BSCVTY, "timer X28 3");
+ f_vty_transceive(BSCVTY, "end");
+
+ /* Allow us to include handling of Channel Activation in this
component, to test recovery from BORKEN */
+ f_rslem_set_auto_chan_act_ack(RSL_PROC, false);
+
+ /* give a moment before releasing */
+ f_sleep(2.0);
+
+ /* Perform a channel release, but omit the RF CHAN REL ACK. The aim is
to get the lchan to go BORKEN. */
+ f_perform_clear(send_rel_ack := false);
+
+ f_vty_expect_borken(BSCVTY, false);
+
+ /* After a timeout of X6, the lchan ends up in LCHAN_ST_BORKEN */
+ f_sleep(6.0);
+
+ f_vty_expect_borken(BSCVTY, true);
+
+ /* Now within two seconds, we expect OsmoBSC to attempt recovering from
the broken state.
+ *
+ * test_variant == 1:
+ * It will first attempt a Channel Activation. If the BTS accepts that,
it means the state is back in sync.
+ * Then OsmoBSC releases the lchan again and makes it available for use.
+ *
+ * test_variant == 2:
+ * It will first attempt a Channel Activation and the BTS NACKs that.
+ * It will then attempt a release, which we ACK, and the channel will
be available again.
+ */
+
+ /* Receive the Channel Activation */
+ var ASP_RSL_Unitdata rx_rsl_ud;
+ RSL.receive(tr_ASP_RSL_UD(tr_RSL_MsgTypeD(RSL_MT_CHAN_ACTIV), sid :=
?)) -> value rx_rsl_ud;
+
+ var RslChannelNr chan_nr;
+ var RSL_IE_Body ie;
+ if (f_rsl_find_ie(rx_rsl_ud.rsl, RSL_IE_CHAN_NR, ie) == true) {
+ chan_nr := ie.chan_nr;
+ } else {
+ setverdict(fail, "Unable to find RSL_IE_CHAN_NR in ",
rx_rsl_ud);
+ return;
+ }
+
+ if (test_variant == 1) {
+ RSL.send(ts_ASP_RSL_UD(ts_RSL_CHAN_ACT_ACK(chan_nr, 23),
rx_rsl_ud.streamId));
+
+ /* Back in sync, receive the Channel Release */
+ RSL.receive(tr_RSL_RF_CHAN_REL(chan_nr));
+ RSL.send(ts_RSL_RF_CHAN_REL_ACK(chan_nr));
+ } else if (test_variant == 2) {
+ RSL.send(ts_ASP_RSL_UD(ts_RSL_CHAN_ACT_NACK(chan_nr,
RSL_ERR_EQUIPMENT_FAIL), rx_rsl_ud.streamId));
+
+ /* Instead tries a release */
+ RSL.receive(tr_RSL_RF_CHAN_REL(chan_nr));
+ RSL.send(ts_RSL_RF_CHAN_REL_ACK(chan_nr));
+ /* That worked, channel usable again */
+ } else {
+ setverdict(fail, "undefined test_variant ", test_variant);
+ }
+
+ f_vty_expect_borken(BSCVTY, false);
+ setverdict(pass);
+}
+
+/* Test recovery path from a BORKEN lchan */
+private function f_TC_unbreak_lchan_after_missing_rel_ack_1(charstring id)
runs on MSC_ConnHdlr {
+ f_tc_unbreak_lchan_after_missing_rel_ack(test_variant := 1);
+}
+testcase TC_unbreak_lchan_after_missing_rel_ack_1() runs on test_CT {
+ var TestHdlrParams pars := f_gen_test_hdlr_pars();
+ var MSC_ConnHdlr vc_conn;
+ f_init(1, true);
+ f_sleep(1.0);
+ vc_conn :=
f_start_handler(refers(f_TC_unbreak_lchan_after_missing_rel_ack_1), pars);
+ vc_conn.done;
+ f_shutdown_helper();
+}
+
+private function f_TC_unbreak_lchan_after_missing_rel_ack_2(charstring id)
runs on MSC_ConnHdlr {
+ f_tc_unbreak_lchan_after_missing_rel_ack(test_variant := 2);
+}
+testcase TC_unbreak_lchan_after_missing_rel_ack_2() runs on test_CT {
+ var TestHdlrParams pars := f_gen_test_hdlr_pars();
+ var MSC_ConnHdlr vc_conn;
+ f_init(1, true);
+ f_sleep(1.0);
+ vc_conn :=
f_start_handler(refers(f_TC_unbreak_lchan_after_missing_rel_ack_2), pars);
+ vc_conn.done;
+ f_shutdown_helper();
+}
+
control {
/* CTRL interface testing */
execute( TC_ctrl_msc_connection_status() );
@@ -12792,6 +12914,9 @@
/* Run TC_ho_out_of_this_bsc last, because it may trigger a segfault
before osmo-bsc's patch
* with change-id I5a3345ab0005a73597f5c27207480912a2f5aae6 */
execute( TC_ho_out_of_this_bsc() );
+
+ execute( TC_unbreak_lchan_after_missing_rel_ack_1() );
+ execute( TC_unbreak_lchan_after_missing_rel_ack_2() );
}
}
--
To view, visit https://gerrit.osmocom.org/c/osmo-ttcn3-hacks/+/35098?usp=email
To unsubscribe, or for help writing mail filters, visit
https://gerrit.osmocom.org/settings
Gerrit-Project: osmo-ttcn3-hacks
Gerrit-Branch: master
Gerrit-Change-Id: I9b4ddfc4a337808d9d5ec538c25fd390b1b2530f
Gerrit-Change-Number: 35098
Gerrit-PatchSet: 2
Gerrit-Owner: neels <nhofm...@sysmocom.de>
Gerrit-Reviewer: Jenkins Builder
Gerrit-Reviewer: daniel <dwillm...@sysmocom.de>
Gerrit-Reviewer: laforge <lafo...@osmocom.org>
Gerrit-Reviewer: neels <nhofm...@sysmocom.de>
Gerrit-MessageType: merged
