| === jfarschman is now known as MilesDenver | ||
| === jfarschman is now known as MilesDenver | ||
| === jfarschman is now known as MilesDenver | ||
| === jfarschman is now known as MilesDenver | ||
| === jfarschman is now known as MilesDenver | ||
| === vladk|offline is now known as vladk | ||
| === jfarschman is now known as MilesDenver | ||
| === CyberJacob|Away is now known as CyberJacob | ||
| 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 |
|---|---|---|
| ubot5 | Ubuntu bug 1338402 in MAAS "Static IP range can be changed with no checks that it would exclude currently allocated IPs" [Critical,Triaged] | 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 | 06:47 |
| 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... | 06:48 |
| 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 :) | 06:49 |
| 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 | 06:50 |
| bigjools | I have NFI why Django is screwing with that other filter thing | 06:52 |
| bigjools | that's totally hatstand | 06:52 |
| === vladk is now known as vladk|offline | ||
| === jfarschman is now known as MilesDenver | ||
| === vladk|offline is now known as vladk | ||
| bigjools | rvba: so Django is trying to be "helpful" again | 07:18 |
| 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 | 07:19 |
| 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 | 07:20 |
| rvba | heh | 07:21 |
| bigjools | because it gets used after you swear a lot at ORM bugs | 07:21 |
| 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? | 07:30 |
| rvba | bigjools: it's a bit ugly: | 07:31 |
| rvba | http://paste.ubuntu.com/7759034/ | 07:31 |
| bigjools | how did I guess that was coming | 07:32 |
| rvba | I just took it from the bug report. | 07:32 |
| === CyberJacob is now known as CyberJacob|Away | ||
| 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 | 07:32 |
| bigjools | o/ jtv | 07:37 |
| 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 :) | 07:45 |
| bigjools | thanks rvba | 07:54 |
| bigjools | eating now, bbiab | 07:54 |
| === jfarschman is now known as MilesDenver | ||
| allenap | BT have just arrived! | 08:29 |
| 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 | 08:51 |
| 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 | 09:00 |
| === jfarschman is now known as MilesDenver | ||
| rvba | allenap: here is a canonistack instance for you: 10.55.32.228 | 09:16 |
| rvba | (With a MAAS package on it) | 09:16 |
| jtv | Also a slightly nontrivial lint branch: https://code.launchpad.net/~jtv/maas/lint-2014-07-07/+merge/225794 | 09:34 |
| 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. | 09:35 |
| 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) | 09:48 |
| 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 | 09:49 |
| dimitern | rvba, jtv, that works for me too, I'll ask the others and send invites | 09:50 |
| dimitern | thanks! | 09:50 |
| jtv | ok | 09:54 |
| rvba | allenap: well, you won't probably need that instance now that you've got a proper connection :) | 09:57 |
| allenap | rvba: I need to stress test it :) | 09:58 |
| === jfarschman is now known as MilesDenver | ||
| 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." | 10:54 |
| jtv | Unlucky port number? | 10:55 |
| rvba | jtv: no, never seen that one | 10:57 |
| === jfarschman is now known as MilesDenver | ||
| jtv | It seems to be rare. | 11:06 |
| === vladk is now known as vladk|offline | ||
| === vladk|offline is now known as vladk | ||
| === jfarschman is now known as MilesDenver | ||
| === ming is now known as Guest65048 | ||
| === jfarschman is now known as MilesDenver | ||
| === Guest47111 is now known as jefferai | ||
| 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 | 13:43 |
| 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. | 13:44 |
| 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 | 13:45 |
| allenap | blake_r: Fwiw, I’m working on the RPC stuff for OperatingSystemRegistry. | 13:47 |
| blake_r | allenap: cool, you be better at it than me! | 13:48 |
| === jfarschman is now known as MilesDenver | ||
| blake_r | allenap: https://code.launchpad.net/~blake-rouse/maas/test-os-preseed-templates/+merge/225838 | 14:29 |
| 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? | 14:42 |
| gmb | rvba, allenap: ^^ | 14:44 |
| gmb | Looks like rsyslog hilarity but I haven’t seen that happen first-hand | 14:45 |
| 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… | 14:48 |
| jpds | rvba: It's not creating any directories under /var/log/maas/rsyslog | 14:49 |
| 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. | 14:55 |
| jpds | allenap: Owned by maas:maas. | 14:59 |
| allenap | jpds: Yeah, on my machine it’s owned by syslog:syslog. | 15:01 |
| rvba | jtv: reviewing your discover-ipv6-interfaces branch now. | 15:06 |
| jtv | Thanks. | 15:08 |
| 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 | 17:17 |
| 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 | 17:20 |
| 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. | 17:22 |
| 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 | 17:23 |
| blake_r | allenap: it shouldn't modify the cache, might have just been downloading all of the new hwe kernels | 17:24 |
| === vladk is now known as vladk|offline | ||
| === CyberJacob|Away is now known as CyberJacob | ||
| === vladk|offline is now known as vladk | ||
| === jfarschman is now known as MilesDenver | ||
| === jfarschman is now known as MilesDenver | ||
| === vladk is now known as vladk|offline | ||
| === vladk|offline is now known as vladk | ||
| === vladk is now known as vladk|offline | ||
| === mwhudson_ is now known as mwhudson | ||
| === CyberJacob is now known as CyberJacob|Away | ||
| === vladk|offline is now known as vladk | ||
| === vladk is now known as vladk|offline | ||
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!