Kyle Tarplee
Kyle Tarplee
Ideally a flag would not be needed but I do not see how to do that yet.
The slack conversation is [here](https://cloud-native.slack.com/archives/CJ1KHJM5Z/p1704224368331799) for reference.
`oras attach` also stopped working for the Gitlab container registry. I assumed it was a gitlab regression but it is not been resolved in a few updates of Gitlab so...
> > `oras attach` also stopped working for the Gitlab container registry. I assumed it was a gitlab regression but it is not been resolved in a few updates of...
It looks like the latest version of Gitlab fixed their regression with `oras attach`. I issue we where seeing before is not longer an issue. It works again so I...
BTW, `oras attach` now works again on Gitlab (after the latest updates).
@Wwwsylvia I'm not sure what you mean. It is not an issue with the distribution spec. The spec gives the client more flexibility in that they do not have to...
Someone could theoretically change the S3 bucket at any time to make it out of sync with the local DB. Why only "re-sync" at startup. Gremlins (silent data corruption or...
Why not include the version of zot in the database. When it different from the zot binary version (forward or backward) start the sync process. If it is the same...
The real issue is that the startup time for ZOT is a function of the amount of data in S3. That is a bad property. Even databases do not have...