Multiple calls of the process_died method occur when a node is launched with respawn set to true and a delay. #2385

Wiktor-99 · 2025-02-25T18:13:38Z

When a node is going to be respawned after some delay following code (inside process monitor run method )will be executed after each spin until configured delay is reached:

 for p in procs:
                try:
                    if not p.is_alive():
                        logger.debug("Process[%s] has died, respawn=%s, required=%s, exit_code=%s",
                                p.name,
                                "True(%f)" % p.respawn_delay if p.respawn else p.respawn,
                                p.required, p.exit_code)
                        exit_code_str = p.get_exit_description()
                        if p.required:
                            printerrlog('='*80+"REQUIRED process [%s] has died!\n%s\nInitiating shutdown!\n"%(p.name, exit_code_str)+'='*80)
                            self.is_shutdown = True
                        elif not p in respawn:
                            if p.exit_code:
                                printerrlog("[%s] %s"%(p.name, exit_code_str))
                            else:
                                printlog_bold("[%s] %s"%(p.name, exit_code_str))
                            dead.append(p)
                            
                        ## no need for lock as we require listeners be
                        ## added before process monitor is launched
                        for l in self.listeners:
                            l.process_died(p.name, p.exit_code)

As a result, the process_died method might be called multiple times. I know this is the end of life for ROS, so I am leaving this issue primarily for reference.

And potential fix:

 for p in procs:
                try:
                    if not p.is_alive():
                        logger.debug("Process[%s] has died, respawn=%s, required=%s, exit_code=%s",
                                p.name,
                                "True(%f)" % p.respawn_delay if p.respawn else p.respawn,
                                p.required, p.exit_code)
                        exit_code_str = p.get_exit_description()
                        if p.required:
                            printerrlog('='*80+"REQUIRED process [%s] has died!\n%s\nInitiating shutdown!\n"%(p.name, exit_code_str)+'='*80)
                            self.is_shutdown = True
                        elif not p in respawn:
                            if p.exit_code:
                                printerrlog("[%s] %s"%(p.name, exit_code_str))
                            else:
                                printlog_bold("[%s] %s"%(p.name, exit_code_str))
                            dead.append(p)
                            
                        ## no need for lock as we require listeners be
                        ## added before process monitor is launched
                       if not p in respawn:
                           for l in self.listeners:
                               l.process_died(p.name, p.exit_code)

Wiktor-99 mentioned this issue Apr 30, 2025

Prevent duplicate 'process died' events by checking if process is already scheduled for respawn #2387

Open

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Multiple calls of the process_died method occur when a node is launched with respawn set to true and a delay. #2385

Multiple calls of the process_died method occur when a node is launched with respawn set to true and a delay. #2385

Multiple calls of the process_died method occur when a node is launched with respawn set to true and a delay. #2385

Multiple calls of the process_died method occur when a node is launched with respawn set to true and a delay. #2385

Comments

Uh oh!