Hey @Marvin Using the web runner in 2.16 and a served, asyncronous subflow I am having the server process crash about 1300-1500 files into the flow regardless of the underlying flows. Can you help me with where I should look for logs or details on what is casuing the crash and are there any things off the top of your head that i should look for in my flows hat may be causing that?

I found some potentially useful information, but I'm afraid it’s not a perfect match for diagnosing your exact issue, as the web runner and server crash in asynchronous subflow scenarios seem to be a very specific case. However, for general troubleshooting, Prefect Error Summaries powered by Marvin AI might help. These summaries provide automated logs of failed flow runs, which could be insightful for you to begin triaging the problem. You can find more about error summaries in the following blog post Introducing Error Summaries by Marvin AI. In addition, block events can provide more context about particular issues. For instance, if your flow involves reading files from AWS S3 or using database credentials, block events can help you identify missing files or expired credentials. More about block events can be read in this blog post No Flow is an Island. Unfortunately, searching the GitHub issues for PrefectHQ/prefect didn’t return any open issues that match your description. If the information obtained from error summaries and block events isn’t sufficient, you may consider opening a new issue on the Prefect Github Issues page providing details about your crash. To mitigate the server process crash you are experiencing, look for patterns in your flows, such as a particular task that fails or resource consumption spikes. Consider implementing more granular logging in your flows and tasks, and perhaps limit concurrency or batch sizes if you suspect that resource exhaustion could be the root cause. Also, ensure you are not running into any limits of the system or services you use in the flows.