Excerpts from Nick Coghlan's message of 2013-02-21 15:24:57 +1000:
On 02/21/2013 02:23 PM, Dan Callaghan wrote:
> That said, the harness itself *doesn't* need to know the server's
> URL
> for this, it's preconfigured in /etc/yum.repos.d. The env var BEAKER is
> set for tasks, but the harness doesn't use it. And since we don't want
> to encourage the harness to talk to the server directly, we should just
> drop it from the API altogether.
Don't we need to provide it anyway, so BEAKER can be set correctly for
RHTS-format tasks?
Maybe... except not.
So this is partly why I keep confusing myself. On the Beaker side, the
*only* thing we can set is system-wide environment variables. Which
means, by definition, they are set for the harness and for the tasks
executed by the harness.
(Actually this is a whole other kettle of fish... Are the tasks
expecting a clean environment? If they are, that's too bad, because even
if the harness clears the environment bash will still source
/etc/profile so the crap we have put in /etc/profile.d will get set
anyway... except maybe not under systemd!)
So anyway, I think the answer here is to drop the mention of it from the
*harness* API, since we aren't expecting the harness to rely on this
variable. Whether the environment variable happens to be set system-wide
by Beaker for *tasks* is a separate question.
--
Dan Callaghan <dcallagh(a)redhat.com>
Software Engineer, Infrastructure Engineering and Development
Red Hat, Inc.