Bring your towel and join one of the fastest growing data communities. Welcome to our second-generation open source orchestration platform, a completely rethought approach to dataflow automation.

Prefect Community

<@ULVA73B9P> What could be causing this error when my kubernetes worker submits a flow run:
```Failed to submit flow run '545eeace-192d-4a4c-b577-9b1ecddde61c' to infrastructure.
Traceback (most recent call last):
  File "/usr/local/lib/python3.11/site-packages/urllib3/util/ssl_.py", line 440, in ssl_wrap_socket
    context.load_verify_locations(ca_certs, ca_cert_dir, ca_cert_data)
OSError: [Errno 24] Too many open files

urllib3.exceptions.MaxRetryError: HTTPSConnectionPool(host='10.100.0.1', port=443): Max retries exceeded with url: /apis/batch/v1/namespaces/default/jobs (Caused by SSLError(OSError(24, 'Too many open files')))```

The error message you're experiencing, `OSError: [Errno 24] Too many open files`, typically occurs when the application or process has reached the limit of file descriptors it can open simultaneously. This could be due to a misconfigured system or environment running out of resources.

Possible reasons and solutions for this error include:

1. *File Descriptor Limits on the Host Machine*:
   - The operating system has a limit on the maximum number of file descriptors available. You can check and raise this limit using the `ulimit` command in Unix-based systems.
     - Check the current limits with: `ulimit -n`
     - To increase the limit, use: `ulimit -n &lt;new_limit&gt;` (requires appropriate permissions)

2. *Excessive Connections*:
   - Review the code and system for potential resource leaks such as opened files or network connections that are not being closed properly.

3. *Kubernetes Configuration*:
   - Ensure that the Kubernetes configuration (such as limits and requests for CPU and memory resources) is appropriately set to handle the workload.

4. *Environment Issues*:
   - Check if there are any known issues specific to the Prefect version you're using or the interaction with Kubernetes.

To assist you further, I need to know which version of Prefect you are using. Can you provide that information, or should I search for potential issues and resolutions in both Prefect 2.x and 3.x?

<@ULVA73B9P> I'm getting this running on kubernetes. Restarting the coredns and worker deployments has helped resolve this issue, what could this mean?