[02:22] <bigjools> jtv: with regarding to https://bugs.launchpad.net/maas/+bug/1365616 can you remember why we restricted API access (even read access) to cluster workers (and admins)?
[02:24] <jtv> bigjools: looking...
[02:24] <bigjools> jtv: just added a comment, refresh in 1 min
[02:24] <bigjools> adding*
[02:26] <jtv> bigjools: the only readily apparent reason I can see is that update privileges should be restricted to admins.
[02:27] <jtv> So I think the reason is probably just that we didn't have time to build both privileged and unprivileged handlers at a time when it wasn't yet clear that the latter would be needed.
[02:30] <bigjools> jtv: well I don't think it's a case of two handlers - the read code explicitly makes this check
[02:31] <jtv> The _read_ code makes a security check!?
[02:32] <bigjools> yes
[02:32] <bigjools> see src/maasserver/api/node_group_interfaces.py
[02:32] <bigjools> I think I know why
[02:32] <bigjools> it would enable attackers easier access
[02:34] <jtv> I think I see another reason.
[02:34] <jtv> The access checks look for two things:
[02:34] <jtv> 1. Admin.
[02:34] <jtv> 2. Cluster worker.
[02:35] <jtv> I think either we didn't realise that when we built the NGI API, or just didn't want the complication of also checking for different levels of access given time pressure.
[02:44] <bigjools> hmmm
[02:50] <bigjools> jtv: are you free for a pre-imp in about 10 minutes?
[03:21] <jtv> bigjools: sorry, didn't notice the IRC notification there.  On the bright side, your branch is now reviewed.  :)  Give me another few minutes.
[03:31] <bigjools> jtv: ok I need to remember what I wanted to talk about...
[03:31] <bigjools> but I have fresh coffee, so it'll come soon
[03:36] <bigjools> jtv: calling
[04:37] <jtv> Looks like we have that CI check now to ensure that a node's IP address is one from the static range.  But it's failing.
[04:37] <jtv> I thought we'd fixed MACAddress.cluster_interface?
[06:20] <jtv> Whooo!  My node is “Deployed”!
[06:20] <jtv> No longer just “Allocated.”
[06:21] <jtv> Thanks rvba.  :-)
[06:21] <rvba> \o/
[06:22] <rvba> jtv: why the new factory? (make_Network)
[06:22] <rvba> jtv: I don't see that pattern (i.e. make_*N*ode) used anywhere else…?
[06:22] <jtv> Look at the first branch first.  :)
[06:23] <rvba> ah
[06:49] <jtv> Sometimes I wonder: are we actually cleaning up the Celery-based code that we disable?
[07:02] <bigjools> seems not
[07:02] <bigjools> I plan on eviscerating some stuff later
[07:03] <bigjools> rvba: you might know this, there's a dupe of this but I can't work out which
[07:03] <bigjools> https://bugs.launchpad.net/maas/+bug/1365035
[07:03] <bigjools> rvba: it's where the power is slow to go off/on
[07:03] <bigjools> and the old machine gets re-used
[07:05] <rvba> bigjools: are you thinking about https://bugs.launchpad.net/maas/+bug/1325610 ?
[07:06] <bigjools> aha
[07:06] <bigjools> thanks
[07:06] <bigjools> oh you already did it
[07:07] <bigjools> ah no you didn't, confusing LP ui fail
[07:47] <bigjools> easy karma for someone, delete-only branch: https://code.launchpad.net/~julian-edwards/maas/remove-update-leases-api/+merge/233466
[14:17] <maastest> Can someone please help me with this question http://askubuntu.com/questions/520240/problem-deploying-node