You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
When creating an oci-sif image, we call ociimage.FetchToLayout.
For an oci: layout, this will copy the image into the cache, or a tmpDir if the cache is disabled. This is un-necessary.
For an oci-archive: this will extract the tar file into a directory, and then treat it as an oci: layout above.
There isn't really any benefit of bringing the layers into the cache when we are working with a layout that can be directly read from disk. It takes up cache space, and I/O.
We seem to do the copies as a result of the ggcr based code incrementally following the containers/image code, and not being reviewed with sufficient depth on completion... my fault :-)