From 4b599ad1e810897494f818047a19e6dc67be4d44 Mon Sep 17 00:00:00 2001
From: Lucas Rodriguez <lucas.rodriguez9616@gmail.com>
Date: Sun, 14 Apr 2024 23:54:27 -0500
Subject: [PATCH] fix: filter on running pods when finding an image for
 injector pod (#2415)

## Description
filter on running pods when finding an image for injector pod


https://kubernetes.io/docs/concepts/workloads/pods/pod-lifecycle/#pod-phase

Description of the `Running` pod phase:

> The Pod has been bound to a node, and all of the containers have been
created. At least one container is still running, or is in the process
of starting or restarting.

## Related Issue
Fixes #2356
Fixes #2410

## Type of change

- [x] Bug fix (non-breaking change which fixes an issue)

## Checklist before merging

- [x] Test, docs, adr added or updated as needed
- [x] [Contributor Guide
Steps](https://github.com/defenseunicorns/zarf/blob/main/CONTRIBUTING.md#developer-workflow)
followed

---------

Co-authored-by: Austin Abro <37223396+AustinAbro321@users.noreply.github.com>
Co-authored-by: razzle <razzle@defenseunicorns.com>
---
 src/pkg/cluster/injector.go | 15 ++++++---------
 src/pkg/k8s/pods.go         |  7 +++----
 2 files changed, 9 insertions(+), 13 deletions(-)

diff --git a/src/pkg/cluster/injector.go b/src/pkg/cluster/injector.go
index 7403c567aa..716a260b96 100644
--- a/src/pkg/cluster/injector.go
+++ b/src/pkg/cluster/injector.go
@@ -23,6 +23,7 @@ import (
 	"github.com/mholt/archiver/v3"
 	corev1 "k8s.io/api/core/v1"
 	"k8s.io/apimachinery/pkg/api/resource"
+	metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
 	"k8s.io/apimachinery/pkg/util/intstr"
 )
 
@@ -445,24 +446,20 @@ func (c *Cluster) getImagesAndNodesForInjection(timeoutDuration time.Duration) (
 
 		// After delay, try running
 		default:
-			pods, err := c.GetPods(corev1.NamespaceAll)
+			pods, err := c.GetPods(corev1.NamespaceAll, metav1.ListOptions{
+				FieldSelector: fmt.Sprintf("status.phase=%s", corev1.PodRunning),
+			})
 			if err != nil {
-				return nil, fmt.Errorf("unable to get the list of pods in the cluster")
+				return nil, fmt.Errorf("unable to get the list of %q pods in the cluster: %w", corev1.PodRunning, err)
 			}
 
 		findImages:
 			for _, pod := range pods.Items {
 				nodeName := pod.Spec.NodeName
 
-				// If this pod doesn't have a node (i.e. is Pending), skip it
-				if nodeName == "" {
-					continue
-				}
-
 				nodeDetails, err := c.GetNode(nodeName)
-
 				if err != nil {
-					return nil, fmt.Errorf("unable to get the node %s", pod.Spec.NodeName)
+					return nil, fmt.Errorf("unable to get the node %q: %w", nodeName, err)
 				}
 
 				if nodeDetails.Status.Allocatable.Cpu().Cmp(injectorRequestedCPU) < 0 ||
diff --git a/src/pkg/k8s/pods.go b/src/pkg/k8s/pods.go
index 5e9b52d4ea..bf29291587 100644
--- a/src/pkg/k8s/pods.go
+++ b/src/pkg/k8s/pods.go
@@ -79,13 +79,12 @@ func (k *K8s) CreatePod(pod *corev1.Pod) (*corev1.Pod, error) {
 
 // GetAllPods returns a list of pods from the cluster for all namespaces.
 func (k *K8s) GetAllPods() (*corev1.PodList, error) {
-	return k.GetPods(corev1.NamespaceAll)
+	return k.GetPods(corev1.NamespaceAll, metav1.ListOptions{})
 }
 
 // GetPods returns a list of pods from the cluster by namespace.
-func (k *K8s) GetPods(namespace string) (*corev1.PodList, error) {
-	metaOptions := metav1.ListOptions{}
-	return k.Clientset.CoreV1().Pods(namespace).List(context.TODO(), metaOptions)
+func (k *K8s) GetPods(namespace string, listOpts metav1.ListOptions) (*corev1.PodList, error) {
+	return k.Clientset.CoreV1().Pods(namespace).List(context.TODO(), listOpts)
 }
 
 // WaitForPodsAndContainers attempts to find pods matching the given selector and optional inclusion filter