Uploaded image for project: 'OpenShift Bugs'
  1. OpenShift Bugs
  2. OCPBUGS-34021

[v2] users should be able to configure timeouts, retries

XMLWordPrintable

    • No
    • CLID Sprint 254
    • 1
    • False
    • Hide

      None

      Show
      None
    • Hide
      Previously, in oc-mirror v2 (Technology Preview), errors encountered on registries, such as timeouts, expired authentication tokens, HTTP 500 errors, and so on were not retried automatically. With this update, retries for these errors have been implemented. With this update, retries for these errors are implemented, and users can configure retry behavior with the following flags:

      * `--retry-times`: Specifies the number of retry attempts. Default is 2.
      * `--retry-delay`: Sets the delay between retries. Default is 1 second.
      * `--image-timeout`: Defines the timeout period for mirroring an image. Default is 10 minutes.
      * `--max-parallel-downloads`: Controls the maximum number of layers to pull simultaneously during a single copy operation. Default is 6.

      (link:https://issues.redhat.com/browse/OCPBUGS-34021[*OCPBUGS-34021*])
      Show
      Previously, in oc-mirror v2 (Technology Preview), errors encountered on registries, such as timeouts, expired authentication tokens, HTTP 500 errors, and so on were not retried automatically. With this update, retries for these errors have been implemented. With this update, retries for these errors are implemented, and users can configure retry behavior with the following flags: * `--retry-times`: Specifies the number of retry attempts. Default is 2. * `--retry-delay`: Sets the delay between retries. Default is 1 second. * `--image-timeout`: Defines the timeout period for mirroring an image. Default is 10 minutes. * `--max-parallel-downloads`: Controls the maximum number of layers to pull simultaneously during a single copy operation. Default is 6. (link: https://issues.redhat.com/browse/OCPBUGS-34021 [* OCPBUGS-34021 *])
    • Bug Fix
    • Done

      Description of problem:

      Errors on registries, such as timeouts, authentication token expiring, HTTP500, etc are not retried. These errors prevent the mirroring from completing and the user needs to restart mirroring from the start.
Ability to retry, set timeouts and throttling (by configuring how many parallel uploads can be done to the registry) can help prevent oc-mirror from exiting prematurely.

      Version-Release number of selected component (if applicable):4.16.0

      How reproducible: always

      Steps to Reproduce:
      This usually appears with long mirroring such as:
      apiVersion: mirror.openshift.io/v2alpha1
      kind: ImageSetConfiguration
      mirror:
      operators:

      • catalog: registry.redhat.io/redhat/redhat-operator-index:v4.16
        full: true 
            1.
    2.
    3.

        Actual results: A lot of failures

        Expected results: Fail less

        Additional info:

        https://redhat-internal.slack.com/archives/C050P27C71S/p1715847774869659?thread_ts=1715769635.682869&cid=C050P27C71S

              skhoury@redhat.com Sherine Khoury
              skhoury@redhat.com Sherine Khoury
              ying zhou ying zhou
              Subhashini T K Subhashini T K
              Votes:
              0 Vote for this issue
              Watchers:
              4 Start watching this issue

                Created:
                Updated:
                Resolved: