rhel: mapping updater only attempts once per day

[RTann]

The repo-2cpe mapping file and the container-name mapping file only update once per day (I think), so it makes sense to only bother fetching it once per day. However, ClairCore only attempts this once per day. If the URL happens to be down for that tiny period of time while ClairCore attempts to fetch the data, then it fails and won't be tried again until the next day.

This is not ideal, but is passable as ClairCore can still fallback on previous data and most (if not all) RHEL-based images can still be indexed correctly. However, if the URL happens to be down for the very first attempt, and there was no initial mapping file embedded in the image running ClairCore, then no RHEL indexing will occur.

[hdonnay]

Can probably hack this into the current stuff; the real fix is a system-wide data store for this sort of garbage.