If yours was working with that directory. It feels like us as users may have lost read and write access to certain parts of the VM. Since this is cloud it might have been an update to their infra.
Regardless; if you follow the steps I outlined above, you can just bring your resources to a writable directory and things should work fine.
Also. Remember that llama_index JUST got a major release to v0.10.0 about a day or so. If you didn’t lock your requirements file to a specific version of llama_index then it would have updated and broken as well.
(Don’t think that’s your case but leaving here since stars align with the release and the solution to this thread)