Enhance PSI example CPU stress pod; fix PSI Prometheus metrics grep regex

Add CPU resource requests and limits to the example CPU stress pod to produce more reliable CPU stress.
Fix PSI Prometheus metrics grep regex so that users can correctly query the metrics for the target containers.

Author: roycaihw

content/en/docs/reference/instrumentation/understand-psi-metrics.md

@@ -62,6 +62,11 @@ spec:
     - "stress"
     - "--cpus"
     - "1"
+    resources:
+      limits:
+        cpu: "500m"
+      requests:
+        cpu: "500m"
 ```
 
 Apply it to your cluster: `kubectl apply -f cpu-pressure-pod.yaml`
@@ -85,7 +90,7 @@ Query the `/metrics/cadvisor` endpoint to see the `container_pressure_cpu_waitin
 ```shell
 # Replace <node-name> with the name of the node where the pod is running
 kubectl get --raw "/api/v1/nodes/<node-name>/proxy/metrics/cadvisor" | \
-    grep 'container_pressure_cpu_waiting_seconds_total{container="cpu-stress",pod="cpu-pressure-pod"}'
+    grep 'container_pressure_cpu_waiting_seconds_total{container="cpu-stress"'
 ```
 The output should show an increasing value, indicating that the container is spending time stalled waiting for CPU resources.
 
@@ -139,7 +144,7 @@ Query the `/metrics/cadvisor` endpoint to see the `container_pressure_memory_wai
 ```shell
 # Replace <node-name> with the name of the node where the pod is running
 kubectl get --raw "/api/v1/nodes/<node-name>/proxy/metrics/cadvisor" | \
-    grep 'container_pressure_memory_waiting_seconds_total{container="memory-stress",pod="memory-pressure-pod"}'
+    grep 'container_pressure_memory_waiting_seconds_total{container="memory-stress"'
 ```
 In the output, you will observe an increasing value for the metric, indicating that the system is under significant memory pressure.
 
@@ -188,7 +193,7 @@ Query the `/metrics/cadvisor` endpoint to see the `container_pressure_io_waiting
 ```shell
 # Replace <node-name> with the name of the node where the pod is running
 kubectl get --raw "/api/v1/nodes/<node-name>/proxy/metrics/cadvisor" | \
-    grep 'container_pressure_io_waiting_seconds_total{container="io-stress",pod="io-pressure-pod"}'
+    grep 'container_pressure_io_waiting_seconds_total{container="io-stress"'
 ```
 You will see the metric's value increase as the Pod continuously writes to disk.