[06:47] <gmb> bigjools: I was going to pick up https://bugs.launchpad.net/maas/+bug/1338402 but then I saw your thread about IP address filtering and Django weirdness… are you working on this bug already/?
[06:47] <bigjools> gmb: I'm trying to fix Andreas' bug, not that one
[06:47] <bigjools> I was thinking about that one on the side
[06:47] <gmb> bigjools: Okay, cool.
[06:47] <bigjools> needs a cursor query I think
[06:48] <gmb> Sounds like it. Django’s magic is a distinct brand of weird sometimes.
[06:48] <bigjools> select TRUE FROM blah HAVING MIN(min_range) AND MAX(max_range)   etc...
[06:49] <gmb> Right.
[06:49] <bigjools> so should get True/False depending on whether the min from existing addresses is outside the entered range
[06:49] <bigjools> min/max
[06:49] <bigjools> well that query as it stands is crap but you get the idea :)
[06:50] <gmb> bigjools: Cool, thanks. I’d got as far as “cursor query” in my head and then was thinking about having another cup of tea before starting coding :).
[06:50] <bigjools> heh
[06:50] <bigjools> needs an ORDER in there too
[06:50] <bigjools> possibly... not sure if PG DTRT here
[06:52] <bigjools> I have NFI why Django is screwing with that other filter thing
[06:52] <bigjools> that's totally hatstand
[07:18] <bigjools> rvba: so Django is trying to be "helpful" again
[07:19] <rvba> bigjools: yeah, I just found where it's done…
[07:19] <bigjools> rvba: there's loads of bugs filed in the django tracker about host()
[07:19] <rvba> bigjools: in django/db/backends/postgresql_psycopg2/operations.py:DatabaseOperations
[07:20] <rvba>     def field_cast_sql(self, db_type):
[07:20] <bigjools> right
[07:20] <rvba>         if db_type == 'inet':
[07:20] <rvba>             return 'HOST(%s)'
[07:20] <rvba>         return '%s'
[07:20] <rvba> And GenericIPField = inet
[07:20] <bigjools> cursor.... sigh
[07:20] <bigjools> should be called a curser
[07:21] <rvba> heh
[07:21] <bigjools> because it gets used after you swear a lot at ORM bugs
[07:30] <rvba> bigjools: there is a workaround
[07:30] <rvba> n [2]: print StaticIPAddress.objects.filter(ip__lte='10.0.0.96').query
[07:30] <rvba> SELECT "maasserver_staticipaddress"."id", "maasserver_staticipaddress"."created", "maasserver_staticipaddress"."updated", "maasserver_staticipaddress"."ip", "maasserver_staticipaddress"."alloc_type", "maasserver_staticipaddress"."user_id" FROM "maasserver_staticipaddress" WHERE "maasserver_staticipaddress"."ip" <= 10.0.0.96
[07:30] <bigjools> what's the workaround?
[07:31] <rvba> bigjools: it's a bit ugly:
[07:31] <rvba> http://paste.ubuntu.com/7759034/
[07:32] <bigjools> how did I guess that was coming
[07:32] <rvba> I just took it from the bug report.
[07:32] <bigjools> I'm going to use a cursor
[07:32] <rvba> bigjools: HOST is used only if get_internal_type return GenericIPField
[07:32] <bigjools> it's ugly, but not as ugly as that
[07:32] <bigjools> right - but what else would break?
[07:32] <rvba> The good thing with this workaround is that it's a one-off
[07:37] <bigjools> o/ jtv
[07:45] <bigjools> rvba: https://code.launchpad.net/~julian-edwards/maas/dup-static-ip-bug-1338452/+merge/225779
[07:45] <rvba> bigjools: on it
[07:45] <bigjools> TIA :)
[07:54] <bigjools> thanks rvba
[07:54] <bigjools> eating now, bbiab
[08:29] <allenap> BT have just arrived!
[08:51] <jtv> Who's willing to review the "identify NGI by name instead of network interface" API branch?  It's a large diff but only a few dozen actual lines changed: https://code.launchpad.net/~jtv/maas/api-ngi-name/+merge/225650
[09:00] <bigjools> gmb: one more thing that bug should have said - clearing those fields while an IP is allocated is just as bad
[09:00] <gmb> bigjools: I figured that based on what Raphers said, but yes.
[09:00] <bigjools> gmb: tiptop
[09:16] <rvba> allenap: here is a canonistack instance for you: 10.55.32.228
[09:16] <rvba> (With a MAAS package on it)
[09:34] <jtv> Also a slightly nontrivial lint branch: https://code.launchpad.net/~jtv/maas/lint-2014-07-07/+merge/225794
[09:35] <rvba> jtv: I'll take it
[09:35] <jtv> Thanks.
[09:35] <jtv> Annoying lint.
[09:35] <jtv> Just didn't have the energy to address it Friday.
[09:48] <dimitern> bigjools, jtv, gmb, rvba, allenap, hey guys, we need to have a quick chat some time this week about networking in maas and juju, when will be a good time? (a couple of you for >1h g+ would be great)
[09:49] <jtv> Maybe tomorrow 09:00 UTC?
[09:49] <rvba> WFM
[09:49] <rvba> allenap: Do not forget to create the symlink to /mnt before you import the images btw
[09:50] <dimitern> rvba, jtv, that works for me too, I'll ask the others and send invites
[09:50] <dimitern> thanks!
[09:54] <jtv> ok
[09:57] <rvba> allenap: well, you won't probably need that instance now that you've got a proper connection :)
[09:58] <allenap> rvba: I need to stress test it :)
[10:54] <jtv> rvba, have you seen a spurious test failure in maasserver.tests.test_dns.TestDNSConfigModifications.test_add_node_updates_zone?
[10:54] <jtv> rndc gets a "connection refused."
[10:55] <jtv> Unlucky port number?
[10:57] <rvba> jtv: no, never seen that one
[11:06] <jtv> It seems to be rare.
[13:43] <blake_r> rvba: cleanup branch from the license key stuff
[13:43] <blake_r> rvba: https://code.launchpad.net/~blake-rouse/maas/utils-osystems-cleanup/+merge/225828
[13:44] <rvba> blake_r: Thanks… I'll have a look in a bit… I think allenap is working on this too, he might want to have a look as well.
[13:45] <allenap> blake_r, rvba: Yeah, I’ll review that.
[13:45] <blake_r> rvba: alot will change when rpc is done but wanted to get it in there
[13:47] <allenap> blake_r: Fwiw, I’m working on the RPC stuff for OperatingSystemRegistry.
[13:48] <blake_r> allenap: cool, you be better at it than me!
[14:29] <blake_r> allenap: https://code.launchpad.net/~blake-rouse/maas/test-os-preseed-templates/+merge/225838
[14:42] <jpds> rsyslogd-3000: Could not open dynamic file '/var/log/maas/rsyslog/10-1-56-17.cloud/2014/07/07/messages' [state -3000] - discarding message
[14:42] <jpds> Anyone know why that would be happening?
[14:44] <gmb> rvba, allenap: ^^
[14:45] <gmb> Looks like rsyslog hilarity but I haven’t seen that happen first-hand
[14:48] <rvba> jpds: never seen that one… like gmb said, it seems something is wrong with rsyslog;  if this happens consistently, I'd suggest changing rsyslog's log level to see if you can get more information about the error…
[14:49] <jpds> rvba: It's not creating any directories under /var/log/maas/rsyslog
[14:55] <allenap> jpds: Check ownership of /var/log/maas/rsyslog and the user that rsyslog runs as. It might be that /var/log/maas/rsyslog has the wrong owner.
[14:59] <jpds> allenap: Owned by maas:maas.
[15:01] <allenap> jpds: Yeah, on my machine it’s owned by syslog:syslog.
[15:06] <rvba> jtv: reviewing your discover-ipv6-interfaces branch now.
[15:08] <jtv> Thanks.
[17:17] <blake_r> allenap: jtv: rvba: gmb: can I get a review of this for 1.6 upgrade https://code.launchpad.net/~blake-rouse/maas/cluster-upgrade-boot-resources/+merge/225868
[17:20] <allenap> blake_r: I have dinner now, but I’ll look at it in ~2 hours. Is that okay?
[17:20] <blake_r> allenap: no rush, just wanted to inform you all
[17:22] <allenap> blake_r: I’d like to review it, and I’m really happy you’ve worked on it. I was puzzled earlier trying to figure out why loads of new things were being downloaded to the cache so I want to see what I missed.
[17:23] <blake_r> allenap: i remember you all mentioning it on irc or email, that it was an issue, couldnt find a bug for it, so i reported a new one
[17:24] <blake_r> allenap: it shouldn't modify the cache, might have just been downloading all of the new hwe kernels