When executing opam-monorepo pull (after a lock), it looks up the tarball(s) in the local opam cache, but does not reach out to the archive-mirrors specified in the repo file of the repository(ies).
This leads, especially in low bandwidth scenarios with a network-local cache, to a lot of downloads that could be avoided.
As far as I can tell, at the moment, OpamRepository.pull_tree is used -- which does not respect the repository mirrors, but instead e.g. OpamUpdate.download_package_source could be used.
When executing
opam-monorepo pull
(after alock
), it looks up the tarball(s) in the local opam cache, but does not reach out to thearchive-mirrors
specified in therepo
file of the repository(ies).This leads, especially in low bandwidth scenarios with a network-local cache, to a lot of downloads that could be avoided.
As far as I can tell, at the moment,
OpamRepository.pull_tree
is used -- which does not respect the repository mirrors, but instead e.g.OpamUpdate.download_package_source
could be used.