Stop using RepositoryCache

[jpopelka]

It's been used for SRPM builds which we now (after #1822) do only in propose-downstream job and @lachmanfrantisek says that we don't need the repository cache.

• remove the usage from BaseJobHelper & LocalProjectMixin
• revert Add persistent volume for repository cache deployment#210
• revert Repository cache packit#1214

@lachmanfrantisek correct me if I'm wrong

[lachmanfrantisek]

👍

Maybe to add, the original goal of this was to make it possible to clone kernel (or any other huge repo with a really long history), but they don't use Packit. And we actually don't cache any project if I am checking that correctly.

If we really really want to use that, we might want to check the possibility of read-many volumes to not need to have so many (same) volumes.

[jpopelka]

After I resolved #1858 I was checking the logs for some time and saw that the workers spent quite a lot of time cloning systemd repo so the cache could still be useful I think.

I guess we can still leave the code in place, close this issue and revisit it in case we'd move to another cluster and decide whether to investigate read-many volumes or disable the cache.

Meanwhile, I reduced (packit/deployment#436) the volumes so that we now request much less storage (all together 72GiB -> 24 GiB if I count correctly).

[jpopelka]

packit/deployment#476 will improve the situation even further, closing