prepare metadata for goa mirror

[kltm]

From geneontology/pipeline#305

[dustine32]

I think this is good to test. So, ontobio will start downloading from https://mirror.geneontology.io/ (for goa at least) since it always uses the source field and of course there should always be some file at that source location.

[kltm]

@dustine32 Correct: anything using the source field in this metadata file will start grabbing from the mirror.
The mirror is populated in a very basic way from the goa-copy-to-mirror branch, moving files to the go-mirror bucket, which then has Cloud Front producing mirror.geneontology.io.
That said, I think the new location should only be an improvement for most practical purposes. I'm not sure how to test here besides letting the pipelines at the new files.

[dustine32]

@kltm Oh sorry, I guess by "test" I meant let them run and in the unlikely event it doesn't work we'll change back (there's probably a better word).