*** eankutse has quit IRC | 00:06 | |
*** eankutse has joined #openstack-dns | 00:09 | |
*** eankutse has quit IRC | 00:09 | |
*** rmoe has quit IRC | 00:09 | |
*** matsuhashi has joined #openstack-dns | 00:21 | |
*** rmoe has joined #openstack-dns | 00:24 | |
*** vinod1 has quit IRC | 00:25 | |
*** vinod1 has joined #openstack-dns | 00:49 | |
*** vinod1 has quit IRC | 01:14 | |
*** crc32 has quit IRC | 01:23 | |
*** shakamunyi has joined #openstack-dns | 01:24 | |
*** shakamunyi has quit IRC | 01:24 | |
*** shakamunyi has joined #openstack-dns | 01:25 | |
*** vinod1 has joined #openstack-dns | 01:26 | |
*** nosnos has joined #openstack-dns | 01:44 | |
*** vinod1 has quit IRC | 02:04 | |
*** richm has left #openstack-dns | 02:07 | |
*** vinod1 has joined #openstack-dns | 02:10 | |
*** betsy has joined #openstack-dns | 02:38 | |
*** vinod1 has quit IRC | 02:53 | |
*** nosnos has quit IRC | 02:54 | |
*** nosnos has joined #openstack-dns | 02:54 | |
*** baffle has quit IRC | 02:55 | |
*** baffle has joined #openstack-dns | 02:56 | |
*** nosnos has quit IRC | 02:59 | |
*** nosnos has joined #openstack-dns | 03:05 | |
*** nkinder has joined #openstack-dns | 03:07 | |
*** matsuhashi has quit IRC | 03:18 | |
*** matsuhashi has joined #openstack-dns | 03:19 | |
*** matsuhashi has quit IRC | 03:24 | |
*** nosnos has quit IRC | 03:34 | |
*** nosnos has joined #openstack-dns | 03:35 | |
*** nosnos has quit IRC | 03:39 | |
*** CaptTofu_ has quit IRC | 03:48 | |
*** matsuhashi has joined #openstack-dns | 04:10 | |
*** msisk has quit IRC | 04:17 | |
*** nosnos has joined #openstack-dns | 04:19 | |
*** shakamunyi has quit IRC | 04:45 | |
*** CaptTofu_ has joined #openstack-dns | 05:01 | |
*** CaptTofu_ has quit IRC | 05:06 | |
*** ajc_ has joined #openstack-dns | 05:07 | |
*** betsy has quit IRC | 05:11 | |
*** betsy has joined #openstack-dns | 05:11 | |
*** shakamunyi has joined #openstack-dns | 05:33 | |
*** shakamunyi has quit IRC | 05:38 | |
*** shakamunyi has joined #openstack-dns | 06:34 | |
*** shakamunyi has quit IRC | 06:38 | |
*** CaptTofu_ has joined #openstack-dns | 06:49 | |
*** CaptTofu_ has quit IRC | 06:54 | |
*** openstackgerrit has quit IRC | 07:10 | |
*** shakamunyi has joined #openstack-dns | 07:34 | |
*** shakamunyi has quit IRC | 07:39 | |
*** shakamunyi has joined #openstack-dns | 08:35 | |
*** tylerdurden has joined #openstack-dns | 08:37 | |
*** shakamunyi has quit IRC | 08:37 | |
*** CaptTofu_ has joined #openstack-dns | 08:38 | |
*** tylerdurden has quit IRC | 08:42 | |
*** CaptTofu_ has quit IRC | 08:42 | |
*** shakamunyi has joined #openstack-dns | 09:38 | |
*** shakamunyi has quit IRC | 09:42 | |
*** openstackgerrit has joined #openstack-dns | 09:47 | |
*** ChanServ sets mode: +v openstackgerrit | 09:47 | |
*** amarao has joined #openstack-dns | 09:57 | |
*** nosnos has quit IRC | 10:19 | |
*** nosnos has joined #openstack-dns | 10:19 | |
*** matsuhashi has quit IRC | 10:23 | |
*** matsuhashi has joined #openstack-dns | 10:24 | |
*** nosnos has quit IRC | 10:24 | |
*** CaptTofu_ has joined #openstack-dns | 10:26 | |
*** matsuhashi has quit IRC | 10:28 | |
*** CaptTofu_ has quit IRC | 10:31 | |
*** shakamunyi has joined #openstack-dns | 10:38 | |
*** shakamunyi has quit IRC | 10:43 | |
openstackgerrit | Graham Hayes proposed a change to stackforge/designate: Ensure that 404's are returned as JSON https://review.openstack.org/101954 | 11:05 |
---|---|---|
*** shakamunyi has joined #openstack-dns | 11:39 | |
*** shakamunyi has quit IRC | 11:44 | |
*** CaptTofu_ has joined #openstack-dns | 11:59 | |
*** shakamunyi has joined #openstack-dns | 12:20 | |
*** tylerdurden has joined #openstack-dns | 12:22 | |
*** shakamunyi has quit IRC | 12:25 | |
*** tylerdurden has quit IRC | 12:26 | |
*** mwagner_lap has quit IRC | 12:45 | |
*** ajc_ has quit IRC | 12:50 | |
*** vinod1 has joined #openstack-dns | 12:52 | |
*** richm has joined #openstack-dns | 13:03 | |
*** nkinder has quit IRC | 13:10 | |
*** shakamunyi has joined #openstack-dns | 13:22 | |
*** shakamunyi has quit IRC | 13:27 | |
*** vinod1 has quit IRC | 13:33 | |
*** vinod1 has joined #openstack-dns | 13:37 | |
*** nkinder has joined #openstack-dns | 13:54 | |
*** vinod1 has quit IRC | 13:56 | |
*** shakamunyi has joined #openstack-dns | 13:59 | |
*** shakamunyi has quit IRC | 13:59 | |
*** eankutse has joined #openstack-dns | 14:02 | |
*** eankutse has quit IRC | 14:02 | |
*** eankutse has joined #openstack-dns | 14:02 | |
*** tsimmons has joined #openstack-dns | 14:06 | |
*** mwagner_lap has joined #openstack-dns | 14:12 | |
*** shakamunyi has joined #openstack-dns | 14:13 | |
openstackgerrit | Jordan Cazamias proposed a change to stackforge/designate: Change log string format to '%' for consistency https://review.openstack.org/101047 | 14:18 |
*** betsy has quit IRC | 14:22 | |
*** vinod1 has joined #openstack-dns | 14:24 | |
*** vinod1 has quit IRC | 14:25 | |
*** vinod1 has joined #openstack-dns | 14:25 | |
*** msisk has joined #openstack-dns | 14:27 | |
*** msisk has quit IRC | 14:27 | |
*** msisk has joined #openstack-dns | 14:28 | |
openstackgerrit | Kiall Mac Innes proposed a change to stackforge/designate: Objects should have a concrete list of their fields https://review.openstack.org/101632 | 14:32 |
*** vinod2 has joined #openstack-dns | 14:58 | |
*** vinod1 has quit IRC | 15:02 | |
*** bandarji has joined #openstack-dns | 15:07 | |
*** shakamunyi has quit IRC | 15:13 | |
*** betsy has joined #openstack-dns | 15:19 | |
*** zane1 has joined #openstack-dns | 15:20 | |
tsimmons | Found something a little weird while unit testing the quotas stuff: http://paste.openstack.org/show/84796/ Maybe I'm doing the Management API stuff wrong. But it may warrant changing? | 15:29 |
Kiall | tsimmons: that sounds like the second API call should have been rejected with a permission denied to me | 15:32 |
Kiall | if you add an X-Auth-Roles: admin header, does it get the right data? | 15:32 |
tsimmons | Ah, sorry meant to add that. No you don't get the right data. I'm using it with noauth so it defaults to admin. | 15:33 |
Kiall | Oh .. Humm | 15:34 |
Kiall | Sounds like a bug | 15:34 |
tsimmons | http://paste.openstack.org/show/84797/ <-- Added the header. | 15:35 |
tsimmons | Yeah I think it has to do with the context being passed in all the way through to storage, maybe it doesn't get other people's quotas just because you're an admin | 15:36 |
tsimmons | You could do a little override in the one of the find_quotas to impersonate the tenant id in question if you're an admin. | 15:37 |
Kiall | tsimmons: https://github.com/stackforge/designate/blob/master/designate/api/middleware.py#L142 | 15:37 |
Kiall | try adding a is_admin=True into that make_context call | 15:37 |
Kiall | I suspect we're still relying on that flag (somewhere..) | 15:37 |
*** jaycaz has joined #openstack-dns | 15:37 | |
tsimmons | No luck :/ | 15:39 |
*** rmoe has quit IRC | 15:50 | |
openstackgerrit | Jordan Cazamias proposed a change to stackforge/designate: Change log string format to '%' for consistency https://review.openstack.org/101047 | 15:50 |
*** vinod2 has quit IRC | 15:54 | |
tsimmons | Kiall: Without any extensive testing, L10-11 seem to fix the issue here. It's a huge hack but I think it isolates the issue at least http://paste.openstack.org/show/84800/ | 15:58 |
Kiall | Ooo.. Humm | 15:59 |
Kiall | Try this instead, context.all_tenants = True | 15:59 |
Kiall | I think that's a valid fix - as long as central is correctly applying the admin_or_owner policy. | 16:00 |
tsimmons | Yeah that works too. | 16:01 |
Kiall | In this case, we're explicitly breaking out of the current tenant, with policy enforcing who is allowed do that. So - By setting all_tenants = True, we let the storage layer know our intent - which causes it not to add a WHERE tenant_id=context.tenant_id and allows your criteria to applied as is | 16:02 |
tsimmons | Well crap. Now the owner thing doesn't seem to be working | 16:03 |
Kiall | Heh.. Typical -_- | 16:04 |
*** rmoe has joined #openstack-dns | 16:06 | |
tsimmons | OH. That's because I left "is_admin=True" in the NoAuth context. | 16:07 |
Kiall | tsimmons: do you still have the added is_admin | 16:07 |
Kiall | snap | 16:07 |
tsimmons | :P | 16:07 |
tsimmons | Of course the unit test that uncovered the issue is still failing.... | 16:11 |
* Kiall hides | 16:16 | |
Kiall | It's probably an existing bug or something -_- | 16:16 |
tsimmons | Oh my god now it's not working. | 16:19 |
tsimmons | Ok surely this is something I'm doing wrong. | 16:19 |
Kiall | I know exactly how you feel - starting at a stacktrace with no real clue what I'm looking at. | 16:22 |
Kiall | And to preempt mugsie ... Yes, this is representative http://weknowgifs.com/wp-content/uploads/2013/11/i-have-no-idea-what-im-doing-dog-gif.gif | 16:22 |
tsimmons | ^ Exactly how I feel | 16:22 |
tsimmons | I think it's just the way I'm testing it but it feels like it's toggling back and forth between working and failing. | 16:22 |
mugsie | tsimmons, I know the feeling | 16:27 |
* mugsie right now -> http://hba77098f93150e3dfce5dfe4e9c551fd.cdn.hpcloudsvc.com/WHATno.gif | 16:27 | |
*** vinod1 has joined #openstack-dns | 16:30 | |
tsimmons | http://weknowgifs.com/wp-content/uploads/2014/03/michael-scott-no-gif.gif | 16:31 |
Kiall | Well.. I've discovered my cause.. Fixing it is going to be painful though -_- | 16:37 |
Kiall | Which leads me to here... http://37.media.tumblr.com/tumblr_lmhg6eYEKd1qii6tmo1_250.gif | 16:37 |
tsimmons | I think I've got it figured out, just trying to make it clean. But really all I'm thinking about is http://media3.giphy.com/media/N9QbkLJWeL7wY/giphy.gif | 16:38 |
*** shakamunyi has joined #openstack-dns | 16:59 | |
*** shakamunyi has quit IRC | 16:59 | |
*** shakamunyi has joined #openstack-dns | 17:00 | |
*** mugsie has quit IRC | 17:06 | |
*** yidclare has joined #openstack-dns | 17:08 | |
*** jaycaz has quit IRC | 17:13 | |
*** amarao has quit IRC | 17:15 | |
rjrjr | rich: are you on? | 17:31 |
rjrjr | so, it has been determined that we will be using BIND9 for our DNS server. the good news, I'm not keenly interested in fixing the BIND9 driver so it works well. | 17:34 |
*** shakamunyi has quit IRC | 17:34 | |
rjrjr | can you please send me any deficiencies you know about with the driver (performance, bugs, etc.)? i'm going to go through the open bugs and blueprints and start working on fixes right away. | 17:35 |
*** openstackgerrit has quit IRC | 17:35 | |
rjrjr | i believe there was some thinking that the driver had performance issues, but i don't know if that has been captured in a bug or blueprint. | 17:36 |
*** openstackgerrit has joined #openstack-dns | 17:36 | |
*** ChanServ sets mode: +v openstackgerrit | 17:36 | |
rjrjr | ^I'm not^I'm now | 17:36 |
openstackgerrit | A change was merged to stackforge/designate: Objects should have a concrete list of their fields https://review.openstack.org/101632 | 17:41 |
*** vinod1 has quit IRC | 17:58 | |
*** jaycaz has joined #openstack-dns | 18:02 | |
*** shakamunyi has joined #openstack-dns | 18:09 | |
openstackgerrit | Jordan Cazamias proposed a change to stackforge/designate: Change log string format to '%' for consistency https://review.openstack.org/101047 | 18:11 |
*** mugsie has joined #openstack-dns | 18:12 | |
richm | rjrjr: I'm on | 18:13 |
openstackgerrit | Kiall Mac Innes proposed a change to stackforge/designate: WIP: Use RRData Objects everywhere https://review.openstack.org/102307 | 18:14 |
richm | rjrjr: I'm not aware of any issues with the bind9 backend other than the issues that I have filed bugs about | 18:15 |
Kiall | vinod: re the use of rdata objects.. I've already started it ^ :) | 18:15 |
Kiall | The pile of cleanup patches were mostly all about getting things ready to do it | 18:15 |
Kiall | rjrjr: I believe the biggest issue is that we re-render the entire file for every change, and re-render ALL files upon startup. | 18:17 |
*** mugsie has quit IRC | 18:17 | |
Kiall | Both could be made much cleaner + more performant by using dnspython to read the zonefile, make the necessary change, and save it again. | 18:17 |
rjrjr | richm: okay. | 18:18 |
rjrjr | kiall: i'll work on the bugs and using dnspython to help with the performance. i'll get a baseline on how it performs now and make sure these changes help. | 18:19 |
*** mugsie has joined #openstack-dns | 18:21 | |
*** mugsie has joined #openstack-dns | 18:21 | |
Kiall | rjrjr: oh, they will.. especially with large zones (and from memory, you have a small number of large zones rather than lots of small zones) | 18:21 |
*** vinod1 has joined #openstack-dns | 18:31 | |
Kiall | 19:15 <Kiall> vinod: re the use of rdata objects.. I've already started it https://review.openstack.org/102307 :) | 18:32 |
Kiall | 19:16 <Kiall> The pile of cleanup patches were mostly all about getting things ready to do it | 18:32 |
Kiall | vinod1: since you weren't actually on the first time I said it ;) | 18:33 |
tsimmons | Kiall: Should I submit a separate patchset/file a bug with the fix to that little quotas bug I found? Or should I just lump it in with the extension? | 18:33 |
Kiall | tsimmons: A bug would be nice, if it's a trivial fix, then I'd include it in the same review and reference the bug and the blueprint | 18:33 |
tsimmons | Ok will do. It's like ten lines. It's mostly working around the issue of having the correct is_admin in the noauth middleware. | 18:35 |
vinod1 | mugsie: Regarding https://review.openstack.org/#/c/101954/4/designate/api/middleware.py, is there a reason not to log 404's? | 18:38 |
mugsie | we do log 404's | 18:39 |
mugsie | that line just dumps a stacktrace in | 18:39 |
mugsie | (the API logs all requests) | 18:39 |
mugsie | and responce codes | 18:39 |
mugsie | (we never had this before, as pecan was doing the error handling, and it missed the middleware entirly) | 18:40 |
Kiall | Soo... I saw that too and thought .. "humm" | 18:40 |
mugsie | had this issue* | 18:40 |
mugsie | should we log a stack trace for 404s? | 18:40 |
mugsie | (IMHO I say no) | 18:40 |
Kiall | No - But I think we log stacktraces all all sorts of other stuff we shouldnt | 18:40 |
mugsie | true | 18:40 |
Kiall | and a generic mechanism would be nice :) | 18:41 |
mugsie | maybe if 500 -> 599 do a stack trace? | 18:41 |
mugsie | we should log things that cause a 5xx status code | 18:41 |
*** nkinder has quit IRC | 18:41 | |
Kiall | Well - I was thinking of things like the exceptions logged between the API<-->Central, where the code is meaningless | 18:42 |
Kiall | But.. I didn't give it any more thought than ^ | 18:42 |
mugsie | does the middleware log those? | 18:42 |
Kiall | The middleware isn't involved there ;) But the same exceptions are.. | 18:43 |
mugsie | ah | 18:43 |
Kiall | Maybe a "expected = True" attr on exceptions which shouldn't have a trace logger | 18:43 |
mugsie | right | 18:43 |
Kiall | Then the middleware are rpc code can be updated to handle it | 18:43 |
Kiall | But.. No clue if that's doable with oslo.rpc! | 18:43 |
Kiall | oslo.messaging* | 18:43 |
* mugsie has no idea | 18:43 | |
Kiall | As I said.. I saw it and though humm, but had to move on ;) | 18:43 |
mugsie | That would be a different patchset anyway ;) | 18:44 |
Kiall | oslo.rpc had a way to say "this is OK, don't log it" | 18:44 |
Kiall | I just can't remember what it was, or what the oslo.messaging gallivant is | 18:44 |
Kiall | equivalent* | 18:44 |
mugsie | k | 18:44 |
Kiall | damn spell checker -_- | 18:44 |
mugsie | but for the patchset ^, I think we can stay with what we have for the time being | 18:45 |
mugsie | or do you think differently Kiall ? | 18:46 |
Kiall | I thought "Humm" and moved on ;) | 18:46 |
Kiall | The oslo.messaging way is something like set an attr on the exception, then it makes sense to use that .. If it's not or more complicated, then what you have is fine IMO | 18:47 |
Kiall | If the* | 18:47 |
Kiall | https://github.com/openstack/oslo.messaging/blob/master/oslo/messaging/rpc/dispatcher.py#L135 | 18:48 |
Kiall | from memory, oslo.rpc saw "expected" exceptions, and wrapped then in that over the wire | 18:49 |
Kiall | the Q is where that happens in oslo.messaging -_- | 18:49 |
Kiall | aha | 18:50 |
Kiall | https://github.com/openstack/oslo.messaging/blob/master/oslo/messaging/rpc/server.py#L123 | 18:50 |
Kiall | If that decorator is used, any Exception classes passed to it are considered OK for that endpoint method | 18:50 |
Kiall | We can adapt that easily enough to support a "expected = True" property on the exception | 18:50 |
Kiall | that way we get the same mechanism available in the RPC / API / etc | 18:51 |
vinod1 | steering the discussion to a more mundane bug :-) | 18:54 |
vinod1 | (Unrelated to your change), mugsie, since you are dealing with pecan now - A trailing slash after a uuid causes a 404 to be returned. e.g. v2/zones/<zone_id>/ | 18:54 |
Kiall | Grr - We should just fix that once and for all and strip any trailing /'s before the requests hit pecan/flask | 18:56 |
openstackgerrit | Graham Hayes proposed a change to stackforge/designate: Ensure that 404's are returned as JSON https://review.openstack.org/101954 | 18:58 |
vinod1 | mugsie: That was fast | 18:59 |
Kiall | Fast'n'buggy :P | 19:05 |
Kiall | If anyone is intersted .. Infra team meeting in #openstack-meeting - I've put the designate repo rename onto the agenda | 19:06 |
openstackgerrit | Tim Simmons proposed a change to stackforge/designate: Add Extensions to the v2 API https://review.openstack.org/100361 | 19:06 |
openstackgerrit | Graham Hayes proposed a change to stackforge/designate: Ensure that 404's are returned as JSON https://review.openstack.org/101954 | 19:08 |
mugsie | vinod1, fixed | 19:11 |
Kiall | vinod1 / betsy / eankutse / rjrjr / richm "<jeblair> #agreed rename designate (and bash8) saturday 1500 UTC" | 19:19 |
openstackgerrit | A change was merged to stackforge/designate: Change log string format to '%' for consistency https://review.openstack.org/101047 | 19:19 |
tsimmons | Kiall: Cool, openstack/designate has a nice ring to it. | 19:19 |
Kiall | ;) | 19:19 |
vinod1 | great - so after that are all the open reviews marked as abandoned? | 19:21 |
Kiall | vinod1: no, things will be moved | 19:21 |
Kiall | There will be 10-15 minute Gerrit downtime while they rename it all | 19:22 |
Kiall | (for all of openstack ;)) | 19:22 |
Kiall | mugsie: https://review.openstack.org/#/c/94997/ | 19:29 |
betsy | kiall: good news. I’m surprised they’re able to do it so quickly | 19:30 |
mugsie | Kiall, https://review.openstack.org/#/c/102328/ | 19:42 |
*** yidclare has quit IRC | 19:43 | |
*** shakamunyi has quit IRC | 19:47 | |
*** rektide has quit IRC | 19:50 | |
*** shakamunyi has joined #openstack-dns | 19:53 | |
*** yidclare has joined #openstack-dns | 20:16 | |
openstackgerrit | Kiall Mac Innes proposed a change to stackforge/designate: Implement change tracking in DesignateObjects https://review.openstack.org/102337 | 20:21 |
Kiall | vinod1: next piece of "objects everywhere" - found while realizing how painful updates with a nested RData object were going to be | 20:22 |
*** mwagner_lap has quit IRC | 20:23 | |
*** nkinder has joined #openstack-dns | 20:24 | |
*** msisk has quit IRC | 20:35 | |
*** tsimmons has quit IRC | 20:37 | |
*** betsy has quit IRC | 20:39 | |
openstackgerrit | A change was merged to stackforge/designate: Added all fields to limits endpoint, and corrected casing https://review.openstack.org/101935 | 20:48 |
*** rektide has joined #openstack-dns | 20:52 | |
openstackgerrit | A change was merged to stackforge/designate: Fixes default content type on /v2/zones endpoint https://review.openstack.org/101929 | 20:53 |
*** eankutse has quit IRC | 21:13 | |
*** msisk has joined #openstack-dns | 21:14 | |
*** msisk has quit IRC | 21:15 | |
*** msisk has joined #openstack-dns | 21:16 | |
*** tsimmons has joined #openstack-dns | 21:24 | |
*** msisk has quit IRC | 21:28 | |
*** msisk has joined #openstack-dns | 21:31 | |
*** msisk has quit IRC | 21:35 | |
*** msisk has joined #openstack-dns | 21:35 | |
*** shakamunyi has quit IRC | 21:38 | |
*** shakamunyi has joined #openstack-dns | 21:39 | |
*** shakamunyi has quit IRC | 21:39 | |
*** shakamunyi has joined #openstack-dns | 21:40 | |
*** shakamunyi has quit IRC | 21:40 | |
*** shakamunyi has joined #openstack-dns | 21:43 | |
*** msisk has quit IRC | 21:44 | |
*** betsy has joined #openstack-dns | 21:46 | |
*** shakamunyi has quit IRC | 21:59 | |
*** shakamunyi has joined #openstack-dns | 22:00 | |
*** vinod1 has quit IRC | 22:00 | |
*** shakamunyi has quit IRC | 22:20 | |
*** shakamunyi has joined #openstack-dns | 22:29 | |
*** yidclare has quit IRC | 22:32 | |
openstackgerrit | A change was merged to stackforge/designate: Implement change tracking in DesignateObjects https://review.openstack.org/102337 | 22:38 |
*** yidclare has joined #openstack-dns | 22:59 | |
*** seventy3 has quit IRC | 23:03 | |
*** vinod1 has joined #openstack-dns | 23:04 | |
*** yidclare has quit IRC | 23:05 | |
*** seventy3 has joined #openstack-dns | 23:05 | |
*** vinod1 has quit IRC | 23:10 | |
*** seventy3 has quit IRC | 23:17 | |
*** seventy3 has joined #openstack-dns | 23:20 | |
*** mwagner_lap has joined #openstack-dns | 23:25 | |
*** tsimmons has quit IRC | 23:26 | |
*** jaycaz has left #openstack-dns | 23:34 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!