Retry failed resolvers #8571
Labels
kind/feature
Categorizes issue or PR as related to a new feature.
Comments
arewm
added
the
kind/feature
arewm
added a commit
to arewm/tektoncd-pipeline
that referenced
this issue
Feb 17, 2025
A bundle resolver is susceptible to flakes in network communication. In order to increase reliability, we can add backoff retries to mitigate transient issues. Partially addresses tektoncd#8571 Signed-off-by: arewm <[email protected]>
arewm
added a commit
to arewm/tektoncd-pipeline
that referenced
this issue
Feb 17, 2025
A bundle resolver is susceptible to flakes in network communication. In order to increase reliability, we can add backoff retries to mitigate transient issues. While it may be possible to add a generic retry functionality around resolvers, implementation will likely be simpler if the retry functionality is isolated to the individual resolvers. Future work is needed to add retries to git resolver requests. Partially addresses tektoncd#8571 Signed-off-by: arewm <[email protected]>
8 tasks
arewm
added a commit
to arewm/tektoncd-pipeline
that referenced
this issue
Mar 19, 2025
A bundle resolver is susceptible to flakes in network communication. In order to increase reliability, we can add backoff retries to mitigate transient issues. While it may be possible to add a generic retry functionality around resolvers, implementation will likely be simpler if the retry functionality is isolated to the individual resolvers. Future work is needed to add retries to git resolver requests. Partially addresses tektoncd#8571 Signed-off-by: arewm <[email protected]>
arewm
added a commit
to arewm/tektoncd-pipeline
that referenced
this issue
Mar 19, 2025
A bundle resolver is susceptible to flakes in network communication. In order to increase reliability, we can add backoff retries to mitigate transient issues. While it may be possible to add a generic retry functionality around resolvers, implementation will likely be simpler if the retry functionality is isolated to the individual resolvers. Future work is needed to add retries to git resolver requests. Partially addresses tektoncd#8571 Signed-off-by: arewm <[email protected]>
arewm
added a commit
to arewm/tektoncd-pipeline
that referenced
this issue
Mar 19, 2025
A bundle resolver is susceptible to flakes in network communication. In order to increase reliability, we can add backoff retries to mitigate transient issues. While it may be possible to add a generic retry functionality around resolvers, implementation will likely be simpler if the retry functionality is isolated to the individual resolvers. Future work is needed to add retries to git resolver requests. Partially addresses tektoncd#8571 Signed-off-by: arewm <[email protected]>
arewm
added a commit
to arewm/tektoncd-pipeline
that referenced
this issue
Mar 19, 2025
A bundle resolver is susceptible to flakes in network communication. In order to increase reliability, we can add backoff retries to mitigate transient issues. While it may be possible to add a generic retry functionality around resolvers, implementation will likely be simpler if the retry functionality is isolated to the individual resolvers. Future work is needed to add retries to git resolver requests. Partially addresses tektoncd#8571 Signed-off-by: arewm <[email protected]>
arewm
added a commit
to arewm/tektoncd-pipeline
that referenced
this issue
Mar 19, 2025
A bundle resolver is susceptible to flakes in network communication. In order to increase reliability, we can add backoff retries to mitigate transient issues. While it may be possible to add a generic retry functionality around resolvers, implementation will likely be simpler if the retry functionality is isolated to the individual resolvers. Future work is needed to add retries to git resolver requests. Partially addresses tektoncd#8571 Signed-off-by: arewm <[email protected]>
tekton-robot
pushed a commit
that referenced
this issue
Mar 28, 2025
A bundle resolver is susceptible to flakes in network communication. In order to increase reliability, we can add backoff retries to mitigate transient issues. While it may be possible to add a generic retry functionality around resolvers, implementation will likely be simpler if the retry functionality is isolated to the individual resolvers. Future work is needed to add retries to git resolver requests. Partially addresses #8571 Signed-off-by: arewm <[email protected]>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Feature request
While it is convenient to use resolvers to reference tasks and pipelines not present in the cluster, the resolution requests are very susceptible to reliability issues. Tekton should be able to automatically retry requests to resolve. Requests shouldn't be retried in perpetuity as this could present confusion as to why the pipeline isn't scheduling. It may be beneficial for users to configure the number of retries and/or the backoff strategy when deploying Tekton to the cluster.
Use case
As a user of Tekton bundles, I want to be able to resolve the pushed definitions even when the container registry is undergoing intermittent reliability/stability issues.
As a user of Tekton bundles, I want to be able to resolve the referenced definitions even when the git provider is undergoing intermittent reliability/stability issues.
The text was updated successfully, but these errors were encountered: