[LLDP] Fix for LLDP advertisements being sent with wrong information. (sonic-net#5493)

* Fix for LLDP advertisments being sent with wrong information.
Since lldpd is starting before lldpmgr, some advertisment packets might sent with default value, mac address as Port ID.
This fix hold the packets from being sent by the lldpd until all interfaces are well configured by the lldpmgrd.

Signed-off-by: Shlomi Bitton <[email protected]>

* Fix comments

* Fix unit-test output caused a failure during build

* Add 'run_cmd' function and use it

* Resume lldpd even if port init timeout reached

Author: shlomibitton

dockers/docker-lldp/lldpd.conf.j2

@@ -10,3 +10,5 @@ configure ports eth0 lldp portidsubtype local {{ mgmt_port_name }}
 configure system ip management pattern {{ ipv4 }}
 {% endif %}
 configure system hostname {{ DEVICE_METADATA['localhost']['hostname'] }}
+{# pause lldpd operations until all interfaces are well configured, resume command will run in lldpmgrd #}
+pause

dockers/docker-lldp/lldpmgrd

@@ -19,6 +19,9 @@ try:
     import subprocess
     import sys
 
+    import syslog
+    import os.path
+    import time
     from sonic_py_common import daemon_base
     from swsscommon import swsscommon
 except ImportError as err:
@@ -27,6 +30,7 @@ except ImportError as err:
 VERSION = "1.0"
 
 SYSLOG_IDENTIFIER = "lldpmgrd"
+PORT_INIT_TIMEOUT = 300
 
 
 class LldpManager(daemon_base.DaemonBase):
@@ -130,16 +134,14 @@ class LldpManager(daemon_base.DaemonBase):
         for (port_name, cmd) in self.pending_cmds.iteritems():
             self.log_debug("Running command: '{}'".format(cmd))
 
-            proc = subprocess.Popen(cmd, shell=True, stdout=subprocess.PIPE, stderr=subprocess.PIPE)
-
-            (stdout, stderr) = proc.communicate()
+            rc, stderr = run_cmd(self, cmd)
 
             # If the command succeeds, add the port name to our to_delete list.
             # We will delete this command from self.pending_cmds below.
             # If the command fails, log a message, but don't delete the command
             # from self.pending_cmds, so that the command will be retried the
             # next time this method is called.
-            if proc.returncode == 0:
+            if rc == 0:
                 to_delete.append(port_name)
             else:
                 self.log_warning("Command failed '{}': {}".format(cmd, stderr))
@@ -166,6 +168,13 @@ class LldpManager(daemon_base.DaemonBase):
         # Set select timeout to 10 seconds
         SELECT_TIMEOUT_MS = 1000 * 10
 
+        # Daemon is paused on the configuration file to avoid lldp packets with wrong information
+        # until all interfaces are well configured on lldpd
+        port_init_done = False
+        port_config_done = False
+        resume_lldp_sent = False
+        start_time = time.time()
+
         sel = swsscommon.Select()
 
         # Subscribe to PORT table notifications in the Config DB
@@ -204,9 +213,25 @@ class LldpManager(daemon_base.DaemonBase):
                             else:
                                 self.pending_cmds.pop(key, None)
 
+                elif key == "PortInitDone":
+                    port_init_done = True
+                elif key == "PortConfigDone":
+                    port_config_done = True
+
             # Process all pending commands
             self.process_pending_cmds()
 
+            # Resume the daemon since all interfaces data updated and configured to the lldpd so no miss leading packets will be sent
+            if not resume_lldp_sent:
+                if check_timeout(self, start_time):
+                    port_init_done = port_config_done = True
+            if port_init_done and port_config_done:
+                port_init_done = port_config_done = False
+                rc, stderr = run_cmd(self, "lldpcli resume")
+                if rc != 0:
+                    self.log_error("Failed to resume lldpd with command: 'lldpcli resume': {}".format(stderr))
+                    sys.exit(1)
+                resume_lldp_sent = True
 
 # ============================= Functions =============================
 
@@ -219,6 +244,16 @@ def main():
 
     lldpmgr.run()
 
+def run_cmd(self, cmd):
+    proc = subprocess.Popen(cmd, shell=True, stdout=subprocess.PIPE, stderr=subprocess.PIPE)
+    (stdout, stderr) = proc.communicate()
+    return proc.returncode, stderr
+
+def check_timeout(self, start_time):
+    if time.time() - start_time > PORT_INIT_TIMEOUT:
+        self.log_error("Port init timeout reached ({} seconds), resuming lldpd...".format(PORT_INIT_TIMEOUT))
+        return True
+    return False
 
 if __name__ == "__main__":
     main()

src/sonic-config-engine/tests/sample_output/py2/lldpd.conf

@@ -1,3 +1,4 @@
 configure ports eth0 lldp portidsubtype local eth0
 configure system ip management pattern 10.0.0.100
 configure system hostname switch-t0
+pause

src/sonic-config-engine/tests/sample_output/py3/lldpd.conf

@@ -1,3 +1,4 @@
 configure ports eth0 lldp portidsubtype local eth0
 configure system ip management pattern 10.0.0.100
 configure system hostname switch-t0
+pause