for _, p := range procs {
		ns := getPidNamespace(p.PID)
		fd, err := unix.Open(ns, unix.O_RDONLY, 0)
		if err != nil {
			// Not uncommon, many processes are transient and we have a TOCTOU here.
			// No problem, must not be the one we are after.
			log.Debugf("failed to open pid %v: %v", p.PID, err)
			continue
		}
		id, err := netlink.GetNetNsIdByFd(fd)
		_ = unix.Close(fd)
		if err != nil {

				// and should never happen - however it has
				// been seen recently due to hardware issues
				// causes Reads() to go slower than Writes().
				//
				// Send such results anyways as this shall
				// expose a problem underneath.
				if totalPut > throughputHighestPut {
					throughputHighestResults = results
					throughputHighestPut = totalPut
					// let the client see lower value as well
					throughputHighestGet = totalGet

	// Verify if region is valid.
	sRegion := cred.scope.region
	// Region is set to be empty, we use whatever was sent by the
	// request and proceed further. This is a work-around to address
	// an important problem for ListBuckets() getting signed with
	// different regions.
	if region == "" {
		region = sRegion
	}
	// Should validate region, only if region is set.
	if !isValidRegion(sRegion, region) {

}

// Open opens the named file.
// When Open returns an error, it should be of type *PathError
// with the Op field set to "open", the Path field set to name,
// and the Err field describing the problem.
//
// Open should reject attempts to open names that do not satisfy
// ValidPath(name), returning a *PathError with Err set to
// ErrInvalid or ErrNotExist.

				// adding unexpected entries into the config.
				//
				// With newer config disallowing changing / turning off
				// notification targets without removing ARN in notification
				// configuration we won't see this problem anymore.
				if reflect.DeepEqual(queue.ARN, arnErr.ARN) && i < len(config.QueueList) {
					config.QueueList = append(config.QueueList[:i],
						config.QueueList[i+1:]...)
				}

	// because we are using `List()` in the handler, without this requeue,
	// the fake client will sometimes drop pod events leading to test flakes.
	//
	// WaitForCacheSync *helps*, but does not entirely fix this problem
	s.namespaces = kclient.New[*corev1.Namespace](kubeClient)
	s.namespaces.AddEventHandler(controllers.FromEventHandler(func(o controllers.Event) {
		s.queue.Add(o)
	}))

	return s
}

		buf := bytes.NewBuffer(nil)
		cmd.Stdout = buf
		cmd.Stderr = buf
		if err := cmd.Start(); err != nil {
			// We are starting so many processes that on
			// some systems (problem seen on Darwin,
			// Dragonfly, OpenBSD) the fork call will fail
			// with EAGAIN.
			if pe, ok := err.(*os.PathError); ok {
				err = pe.Err
			}