launchpad-dev team mailing list archive

[John Arbash Meinel <john@xxxxxxxxxxxxxxxxx>]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Summary:
  When cleaning up data from the librarian, is the money we save by
  removing the data worth the cost if we remove the wrong stuff?

As part of rt #53574, we're looking to clean out 'unused' data from
the Librarian. We have some idea of what we can do, but there still
seems to be a policy decision that is missing.

Specifically Colin Watson signed off on deleting:

1) Binary packages that are part of an EOL series (at this point,
karmic and maverick)
2) That have been superseded before the release date
3) That have not existed in any pocket other than release
4) (except maybe in the future -proposed)

I'm trying to update the query to add in (2) and (3):
https://wiki.canonical.com/InformationInfrastructure/OSA/LPHowTo/RemoveSupersededBinaries

By my numbers, we can free up about 300GB per release (600 total).

- From what I can tell, each release adds something like 500GB total,
and ~300GB of that would be something that we can clean up (if my
encoding of the above rules is correct.)

So that is about 600GB per year (once things EOL) in disk space that
we save. Probably 1TB is a reasonable upper bound on it.

Is that worth doing?

John
=:->
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/

iEYEARECAAYFAk/pcGMACgkQJdeBCYSNAAOQUACfejDis0501fW0iAqiH37RlmjM
hEIAoKWh/cLtQ5ts48TlbIRc4mJbBCZG
=vMoF
-----END PGP SIGNATURE-----