New GKE nodes are unable to pull images from Container Registry on Google Cloud Platform
I have a GKE cluster on GCP with 2 existing node pools. I'm trying to create a new node pool:
Any new node pools I create are unable to pull containers from Container Registry with the following error:
The settings and service account I'm using for the new node pool are exactly the same as the existing node pools (other than machine type):
Everything is working as expected on the existing node pools, but each time I create a new node pool I'm having the same problem:
The logs show the following error on image pull:
E0128 06:44:03.656599 1763 remote_image.go:242] "PullImage from image service failed" err="rpc error: code = NotFound desc = failed to pull and unpack image "eu.gcr.io/{PROJECT}/{IMAGE}@sha256:": failed to copy: httpReadSeeker: failed open: could not fetch content descriptor sha256: (application/vnd.docker.container.image.v1+json) from remote: not found" image="eu.gcr.io/{PROJECT}/{IMAGE}@sha256:*****"
As per the container registry official doc, Container Registry is deprecated. After May 15, 2024. Artifact Registry will host images for the gcr.io domain in projects without previous Container Registry usage Starting January 8, 2024, if your organization has not previously used Container Registry, new gcr.io repositories will be hosted on Artifact Registry by default. For more information on this change, see gcr.io hosted on Artifact Registry.
As per the log from pull image error failed to copy: httpReadSeeker: failed open: could not fetch content descriptor sha256: seems that this might be due to image tag typo error and also Skaffold is not supporting the correct tag. As per the Skaffold documentation, it supports multiple taggers or tag policies for tagging images. By following this Skaffold documentation add the necessary configuration in skaffold.yaml and give it a try.
Note: Cross-check the image tags for mistakes which might be the reason for error.
- python 3: Using pisa for creating pdf with multiple image urls in html content, rendering images incorrectly
- schedule autoscaler in terraform for GCP?
- Multiple STRING_AGG in one query
- Why aren't nacked messages ending up in my dead letter topic?
- Handle 503 error when making jobs.insert call
- Use http Google Cloud Functions "Require Authentication" with Firebase
- Google App Engine slow cold boot time (Flask app)
- gcloud SSH in same terminal window
- Cannot enable API on GCP (User role=Owner && Billing account is linked)
- Select All Columns Except Some in Google BigQuery?
- Why does my flex env google app engine instance have a minimum of 2 instances running, even when there is no traffic?
- How Does GCP Global Application Load Balancer Select the Closest Backend?
- How to use Google Search Grounding with Dynamic Retrieval Configuration
- Google Stackdriver Logging add comments to advanced filter
- Exporting from Cloud SQL to CSV with column headers
- BigQuery Date-Partitioned Views
- Check TPU workload/utilization
- Google Forms API raises google.auth.exceptions.RefreshError: 'No access token in response.'
- How can I see request count (not rate) for a Google Cloud Run application?
- I can write to my Firebase Firestore database, but am unable to read from it (Android Studio using Java)
- How to get task count in GCP cloud task
- Programmatically get current Service Account on GCP
- How to Access GKE Private Master from VPN in Hub VPC with Peering
- My gcp service account with "artifact registry admin" account was not able to push the docker image to repo
- GCP API Gateway: Cannot use path params
- Flask api on gcloud's URL keeps on loading
- GCP CloudBuild substitution fails in pipeline
- Is it secure to use an .env file to store API keys in Firebase Cloud Functions instead of Google Secret Manager?
- How do I keep a long-running background task from stopping on Google Cloud Run?
- GCP Cloud SQL Proxy times out: connectex