[muxorch] set mux state to init upon warm reboot (sonic-net#2834)

Warm reboot with manual mux mode was causing orchagent crash. Initializing mux state to INIT upon warm reboot, the state will then be updated to reflect previous APPDB state. Signed-off-by: Nikola Dancejic <[email protected]> Co-authored-by: Ying Xie <[email protected]>
vadymhlushko-mlnx · Aug 10, 2023 · 3e2974d · 3e2974d
1 parent 17c4d73
commit 3e2974d
Show file tree

Hide file tree

Showing 2 changed files with 50 additions and 3 deletions.
diff --git a/orchagent/muxorch.cpp b/orchagent/muxorch.cpp
@@ -24,6 +24,7 @@
 #include "routeorch.h"
 #include "fdborch.h"
 #include "qosorch.h"
+#include "warm_restart.h"
 
 /* Global variables */
 extern Directory<Orch*> gDirectory;
@@ -416,9 +417,18 @@ MuxCable::MuxCable(string name, IpPrefix& srv_ip4, IpPrefix& srv_ip6, IpAddress
     state_machine_handlers_.insert(handler_pair(MUX_STATE_INIT_STANDBY, &MuxCable::stateStandby));
     state_machine_handlers_.insert(handler_pair(MUX_STATE_ACTIVE_STANDBY, &MuxCable::stateStandby));
 
-    /* Set initial state to "standby" */
-    stateStandby();
-    state_ = MuxState::MUX_STATE_STANDBY;
+    if (WarmStart::isWarmStart()) {
+        /* Warmboot case, Set initial state to "init" 
+         * State will be updated to previous value upon APP DB sync
+         */
+        state_ = MuxState::MUX_STATE_INIT;
+    }
+    else
+    {
+        /* Set initial state to "standby" */
+        stateStandby();
+        state_ = MuxState::MUX_STATE_STANDBY;
+    }
 }
 
 bool MuxCable::stateInitActive()

diff --git a/tests/test_mux.py b/tests/test_mux.py
@@ -1576,6 +1576,43 @@ def test_soc_ip(self, dvs, dvs_route, setup_vlan, setup_mux_cable, testlog):
 
         self.create_and_test_soc(appdb, asicdb, dvs, dvs_route)
 
+    def test_warm_boot_mux_state(
+            self, dvs, dvs_route, setup_vlan, setup_mux_cable, setup_tunnel,
+            remove_peer_switch, neighbor_cleanup, testlog
+    ):
+        """
+        test mux initialization during warm boot.
+        """
+        appdb = swsscommon.DBConnector(swsscommon.APPL_DB, dvs.redis_sock, 0)
+        apdb = dvs.get_app_db()
+
+        self.set_mux_state(appdb, "Ethernet0", "active")
+        self.set_mux_state(appdb, "Ethernet4", "active")
+        self.set_mux_state(appdb, "Ethernet8", "standby")
+
+        # Execute the warm reboot
+        dvs.runcmd("config warm_restart enable swss")
+        dvs.stop_swss()
+        dvs.start_swss()
+
+        time.sleep(5)
+
+        fvs = apdb.get_entry(self.APP_MUX_CABLE, "Ethernet0")
+        for key in fvs:
+            if key == "state":
+                assert fvs[key] == "active", "Ethernet0 Mux state is not active after warm boot, state: {}".format(fvs[key])
+
+        fvs = apdb.get_entry(self.APP_MUX_CABLE, "Ethernet4")
+        for key in fvs:
+            if key == "state":
+                assert fvs[key] == "active", "Ethernet4 Mux state is not active after warm boot, state: {}".format(fvs[key])
+
+        fvs = apdb.get_entry(self.APP_MUX_CABLE, "Ethernet8")
+        for key in fvs:
+            if key == "state":
+                assert fvs[key] == "standby", "Ethernet8 Mux state is not standby after warm boot, state: {}".format(fvs[key])
+
+
 # Add Dummy always-pass test at end as workaroud
 # for issue when Flaky fail on final test it invokes module tear-down before retrying
 def test_nonflaky_dummy():