Fixes Kubernetes Service using expired credentials #50074
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
The Kubernetes service occasionally fails to forward requests to EKS clusters or retrieve the cluster schema due to AWS rejecting the request with an "expired token" error.
EKS access tokens are generated using STS presigned URLs, which include details such as the cluster, backend credentials, and assumed roles. By default, these tokens are valid for 15 minutes, and the Kubernetes service refreshes them every $(15 - 1) / 2 = 7\text{ }minutes$.
However, our cloud SDK caches the underlying `aws.Session`, particularly those with assumed roles, for 15 minutes.
This leads to a scenario where the token is refreshed a second time at approximately 14 minutes, close to the token's 15-minute validity. If the underlying credentials expire before the next token refresh, given that they were reused from the previous query and cached since then, it results in the Kubernetes Service considering the token valid (since it is a Base64-encoded presigned URL without knowledge about the credentials), but AWS EKS cluster rejects the request, treating the credentials as expired.
This PR adds an option to disable cache for EKS STS token signing which results in creating a session per EKS cluster sign process.
Bellow one can find the error message EKS returns.
```
2024-12-09T17:00:15Z ERRO [KUBERNETE] Failed to update cluster schema error:[
ERROR REPORT:
Original Error: *errors.StatusError the server has asked for the client to provide credentials
Stack Trace:
github.com/gravitational/teleport/lib/kube/proxy/scheme.go:140 github.com/gravitational/teleport/lib/kube/proxy.newClusterSchemaBuilder
github.com/gravitational/teleport/lib/kube/proxy/cluster_details.go:193 github.com/gravitational/teleport/lib/kube/proxy.newClusterDetails.func1
runtime/asm_amd64.s:1695 runtime.goexit
User Message: the server has asked for the client to provide credentials] pid:7.1 start_time:2024-12-09T17:00:15Z proxy/cluster_details.go:210
2024-12-09T17:00:24Z ERRO [KUBERNETE] Failed to update cluster schema error:[
ERROR REPORT:
Original Error: *errors.StatusError the server has asked for the client to provide credentials
Stack Trace:
github.com/gravitational/teleport/lib/kube/proxy/scheme.go:140 github.com/gravitational/teleport/lib/kube/proxy.newClusterSchemaBuilder
github.com/gravitational/teleport/lib/kube/proxy/cluster_details.go:193 github.com/gravitational/teleport/lib/kube/proxy.newClusterDetails.func1
runtime/asm_amd64.s:1695 runtime.goexit
User Message: the server has asked for the client to provide credentials] pid:7.1 start_time:2024-12-09T17:00:24Z proxy/cluster_details.go:210
```
Changelog: Fixes an intermittent EKS authentication failure when dealing with EKS auto-discovery.
Signed-off-by: Tiago Silva <[email protected]>
creack
approved these changes
Dec 12, 2024
rosstimothy
approved these changes
Dec 12, 2024
This was referenced Dec 13, 2024
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
The Kubernetes service occasionally fails to forward requests to EKS clusters or retrieve the cluster schema due to AWS rejecting the request with an "expired token" error.
EKS access tokens are generated using STS presigned URLs, which include details such as the cluster, backend credentials, and assumed roles. By default, these tokens are valid for 15 minutes, and the Kubernetes service refreshes them every$(15 - 1) / 2 = 7\text{ }minutes$ . However, our cloud SDK caches the underlying
aws.Session, particularly those with assumed roles, for 15 minutes.This leads to a scenario where the token is refreshed a second time at approximately 14 minutes, close to the token's 15-minute validity. If the underlying credentials expire before the next token refresh, given that they were reused from the previous query and cached since then, it results in the Kubernetes Service considering the token valid (since it is a Base64-encoded presigned URL without knowledge about the credentials), but AWS EKS cluster rejects the request, treating the credentials as expired.
This PR adds an option to disable cache for EKS STS token signing which results in creating a session per EKS cluster sign process.
Bellow one can find the error message EKS returns.
Changelog: Fixes an intermittent EKS authentication failure when dealing with EKS auto-discovery.
Fixes #50072