Upgrade reports no changes when containers-storage tag is stale

Aberta
#2,519 0 comentários 0 reações 0 responsáveis Ver no GitHub

Mantenedores costumam responder em até 2 dias

Ninguém assumiu esta issue ainda.

Avaliação

Dificuldade
4/5
Tempo estimado
3-5 dias
Facilidade para iniciantes
58/100
Tipo de issue
Bug
Clareza
Razoavelmente clara
Status de atividade
Ativa
Stack de tecnologia
rust

Direção de pesquisa

Start by reproducing the stale local tag scenario from the listed boot and upgrade steps. Trace the unified pull path from the registry to bootc storage and compare it with the direct registry path used by --check. Done means the local tag refreshes to the remote digest and the upgrade stages the new image instead of reporting no changes.

Escrita pelo modelo de indexação a partir do texto da issue.

Descrição

What happened

On bootc 1.16.13, with :testing at an older digest in local containers-storage and a newer digest published remotely, both plain update and update --apply printed a wall of Already exists lines and then concluded with No changes in => and No update available. Meanwhile status showed a cached update and upgrade --check correctly reported the newer version and digest as available.

Inspecting the local tag afterward showed it still pointed at the old digest: the pull step never refreshed it, so the import step compared stale-local against booted and found nothing to do.

What I expected

The pull step to refresh the local tag to the remote digest and the upgrade to stage, since --check proved the update existed.

Steps to reproduce

  1. Boot an image whose tag also exists in local containers-storage at the booted (older) digest.
  2. Publish a newer digest under the same remote tag.
  3. Run the update with apply. Observe Already exists spam followed by No changes pointing at the old digest.
  4. Confirm the local tag still resolves to the old digest.
  5. Run a plain container pull of the same tag, then retry the upgrade: it stages correctly (in my case, 18 changed layers, new ostree commit queued for next boot).

Additional context

System log lines during the failed runs show the unified pull path: pulling from transport registry to bootc storage, then importing from the containers-storage transport by tag name. The check path instead fetches from the registry directly, which is why --check saw the update while the real pull did not. Workaround is a manual container pull of the tag before upgrading. Seen twice in a row, five minutes apart, so unlikely to be a transient registry read.

Linguagem predominante
Rust
Estrelas
2.3k
Forks
230
Merge médio
2d 11h
PRs com merge (30d)
42

Preparar o ambiente

Primeiros passos

  1. Leia a issue inteira e depois o guia de contribuição do projeto.
  2. Comente na issue dizendo que vai assumir — evita que duas pessoas façam o mesmo trabalho.
  3. Faça um fork do repositório e trabalhe em uma branch.
  4. Abra um pull request que referencie o número da issue.

Mais de bootc-dev/bootc

Todas as issues de bootc-dev/bootc

Issues semelhantes

Mais issues de Rust

Receba novas issues na sua caixa de entrada

Um resumo curto de issues do GitHub para quem está começando.