Project

General

Profile

Actions

Bug #19982

closed

Docker repository not immediately available on local registry after sync

Added by Thomas McKay almost 7 years ago. Updated almost 6 years ago.

Status:
Closed
Priority:
Normal
Assignee:
Category:
-
Target version:
Difficulty:
Triaged:
Fixed in Releases:
Found in Releases:

Description

Cloned from https://bugzilla.redhat.com/show_bug.cgi?id=1452149

Description of problem:
After syncing a docker repository using hammer:
`hammer repository synchronize --id="666"`
the `repository info` subcommand reports sync status as success and reports
'published at: sat6.server.com:5000/foo-bar-baz,

however, if you try to pull the repo using docker: `docker pull sat6.server.com:5000/foo-bar-baz`, you'll get:

Using default tag: latest
Trying to pull repository sat6.server.com:5000/foo-bar-baz ...
unknown: Not Found

Foreman takes some time before the repo is really available.
I believe, the sync task should finish AFTER the repo is really available.

How reproducible:
always on pulling the repo just after syncing it.

Steps to Reproduce:
1. hammer repository create --product-id="1" --docker-upstream-name="busybox" --name="foo" --publish-via-http="true" --url="https://registry-1.docker.io" --content-type="docker"
2. hammer repository synchronize --id=1
3. (on a docker host with katello consumer ca installed): docker pull sat6-server.com:5000/foo-bar-baz

Actual results:
unknown: Not Found

Expected results:
repo available and pull finishes successfully

Additional info:
workaround: give foreman some time to publish the repo.


Related issues 1 (1 open0 closed)

Blocks Katello - Tracker #19682: container image and registry trackerNewThomas McKay05/26/2017

Actions
Actions #1

Updated by Thomas McKay almost 7 years ago

Actions #2

Updated by Thomas McKay almost 7 years ago

  • Subject changed from Docker repository not immediately available on local registry after sync to Docker repository not immediately available on local registry after sync

https://github.com/pulp/crane/blob/master/docs/index.rst#configuration

It is likely that this is just a case of crane not having yet polled for the new data. Perhaps the data_dir_polling_interval should be decreased from the default of 60 seconds.

Actions #3

Updated by Eric Helms almost 7 years ago

  • translation missing: en.field_release set to 250
Actions #4

Updated by Justin Sherrill over 6 years ago

  • translation missing: en.field_release changed from 250 to 114
Actions #5

Updated by Thomas McKay almost 6 years ago

  • Status changed from New to Closed
  • translation missing: en.field_release changed from 114 to 166
Actions

Also available in: Atom PDF