Fix check for nvidia-device-plugin-daemonset when deploying NVIDIA operator stack

[bdattoma]

As of now, when deploying the NVIDIA GPU operator stack, the nvidia-device-plugin-daemonset gets restarted after being in "init" status. Our script has already started waiting for the pod to be in Ready status but it waits for a pod which won't never be up and running. In addition, it has a 20 minutes timeout which makes CI spending more time than needed

Solution:

• instead of waiting 20 minutes after fetching the pod name, it waits only 10seconds and if the pod is not ready, it continues the loop (i.e., fetching the pod name again, wait 10 s, etc)
• wait for the GPU node to be ready before deploying the NVIDIA Stack, it reduces the risk of hitting the timeouts

PR validation:

1. provision smaller GPU on IBM cloud: rhods-ci-pr-test/3368 PASS - the job failure at the end is not related to this PR. It took 15 minutes from provisioning to stack deployment
2. provision bigger GPU on GCP: rhods-ci-pr-test/3371 - approx 15 min for e2e flow (provisioning + operator installtion)
3. test provisioning without this PR and compare timing: rhods-ci-pr-test/3372 PASS - it took 28 minutes for e2e flow (provisioning + operator installtion) - the job failure at the end is not related to this PR

[github-actions]

Robot Results

✅ Passed	❌ Failed	⏭️ Skipped	Total	Pass %
547	0	0	547	100

[sonarqubecloud]

Quality Gate passed

Issues
0 New issues
0 Accepted issues

Measures
0 Security Hotspots
0.0% Coverage on New Code
0.0% Duplication on New Code

See analysis details on SonarCloud