| *** rmkdroid has joined #openstack-dev | 00:03 | |
| chmouel | notmyname: this would fix it btw: http://pastie.org/private/f0wj13be2qvstgwqdwqvza but not sure why it would have been set as invalidbucketname at first | 00:03 |
|---|---|---|
| *** eglynn has joined #openstack-dev | 00:13 | |
| *** novas0x2a|laptop has quit IRC | 00:21 | |
| devcamcar | looks like devstack integration tests are being intermittently flakey: https://jenkins.openstack.org/job/gate-integration-tests-devstack-vm/2075/ | 00:22 |
| jeblair | devcamcar: jesse thinks that may have been due to bug 945454, since fixed | 00:23 |
| uvirtbot` | Launchpad bug 945454 in devstack "images properties have bad values" [Medium,Fix committed] https://launchpad.net/bugs/945454 | 00:23 |
| devcamcar | cool, flipping a few more approve bits in horizon so we'll know soon | 00:23 |
| openstackgerrit | Verification of a change to openstack/horizon failed: set image attributes not modifiable as read-only https://review.openstack.org/4797 | 00:24 |
| openstackgerrit | Verification of a change to openstack/horizon failed: Return proper error message when adding security group with no params https://review.openstack.org/4804 | 00:25 |
| devcamcar | jeblair: https://jenkins.openstack.org/job/gate-integration-tests-devstack-vm/2095/console | 00:27 |
| devcamcar | sqlite issue | 00:27 |
| jeblair | devcamcar: whoops. that's my code. :( sorry. | 00:28 |
| jeblair | devcamcar: retrigger should be good, i'll look into improving that. | 00:28 |
| devcamcar | jeblair: will give it a shot | 00:29 |
| chmouel | 0:31:14 chmouel | notmyname: this would fix it btw: http://pastie.org/private/f0wj13be2qvstgwqdwqvza but not sure why │ akscram | 00:34 |
| chmouel | damn middle click on X11 | 00:34 |
| *** vizsla_p has quit IRC | 00:42 | |
| *** vizsla_p has joined #openstack-dev | 00:42 | |
| *** novas0x2a|laptop has joined #openstack-dev | 00:44 | |
| devcamcar | jeblair: worked that time, thanks | 00:49 |
| jeblair | devcamcar: great | 00:49 |
| *** anotherjesse_zz is now known as anotherjesse | 00:51 | |
| anotherjesse | jeblair: have a bit to talk about docs.openstack.org? | 00:55 |
| jeblair | anotherjesse: sure | 00:55 |
| anotherjesse | jeblair: first - you can kill that vm you saved for me if you haven't already | 00:55 |
| jeblair | anotherjesse: ok, thanks, will do | 00:56 |
| anotherjesse | jeblair: second - I pinged annegentle about making docs.openstack.org be a git repo and use jenkins to publish | 00:56 |
| anotherjesse | jeblair: is that something that you or monty have time for helping get setup? | 00:57 |
| jeblair | anotherjesse: that sounds swell, and something i'm sure we can make time for; do you know how it's published now? | 00:58 |
| anotherjesse | right now it is a static cloudsite page (I think) | 00:59 |
| anotherjesse | jeblair: I sent you a private msg with more info | 01:00 |
| *** littleidea has joined #openstack-dev | 01:03 | |
| *** rmkdroid has quit IRC | 01:19 | |
| *** anotherjesse is now known as anotherjesse_zz | 01:21 | |
| *** cjwilson has quit IRC | 02:12 | |
| *** novas0x2a|laptop has quit IRC | 02:43 | |
| *** pixelbeat has quit IRC | 03:00 | |
| *** eglynn has quit IRC | 03:11 | |
| *** anotherjesse_zz is now known as anotherjesse | 03:57 | |
| *** rods has quit IRC | 04:03 | |
| *** anotherjesse is now known as anotherjesse_zz | 04:24 | |
| openstackgerrit | Verification of a change to openstack/keystone failed: Make listen IP configurable https://review.openstack.org/4880 | 04:25 |
| *** anotherjesse_zz is now known as anotherjesse | 04:42 | |
| *** anotherjesse is now known as anotherjesse_zz | 04:57 | |
| *** andrewsmedina has quit IRC | 05:00 | |
| *** vizsla_p has quit IRC | 05:09 | |
| *** novas0x2a|laptop has joined #openstack-dev | 05:36 | |
| *** dtroyer has quit IRC | 05:40 | |
| *** littleidea has quit IRC | 05:58 | |
| *** x86brandon has quit IRC | 06:29 | |
| *** anotherjesse_zz is now known as anotherjesse | 06:34 | |
| openstackgerrit | Verification of a change to openstack/keystone failed: Make listen IP configurable https://review.openstack.org/4880 | 06:41 |
| *** Vek has quit IRC | 07:36 | |
| *** Vek has joined #openstack-dev | 07:38 | |
| *** Vek has quit IRC | 07:48 | |
| *** novas0x2a|laptop has quit IRC | 08:10 | |
| *** anotherjesse is now known as anotherjesse_zz | 08:16 | |
| *** anotherjesse_zz is now known as anotherjesse | 08:19 | |
| *** anotherjesse is now known as anotherjesse_zz | 08:21 | |
| *** adiantum has joined #openstack-dev | 08:22 | |
| *** Mkenneth has joined #openstack-dev | 08:40 | |
| *** adiantum has quit IRC | 08:46 | |
| *** Mkenneth has quit IRC | 08:49 | |
| *** Mkenneth has joined #openstack-dev | 09:07 | |
| *** Mkenneth has left #openstack-dev | 09:07 | |
| *** novas0x2a|laptop has joined #openstack-dev | 09:28 | |
| *** dalang has quit IRC | 10:04 | |
| *** dalang has joined #openstack-dev | 10:05 | |
| *** YorikSar has quit IRC | 10:51 | |
| *** YorikSar has joined #openstack-dev | 10:58 | |
| *** YorikSar has quit IRC | 11:04 | |
| *** vizsla_p has joined #openstack-dev | 11:05 | |
| *** vizsla_p has quit IRC | 11:09 | |
| *** berendt has joined #openstack-dev | 11:24 | |
| *** adiantum has joined #openstack-dev | 11:30 | |
| *** vizsla has joined #openstack-dev | 11:42 | |
| *** adiantum has quit IRC | 11:45 | |
| *** Mkenneth has joined #openstack-dev | 11:52 | |
| *** adiantum has joined #openstack-dev | 12:02 | |
| *** adiantum has quit IRC | 12:12 | |
| vizsla | bleh, I need a better jerb. | 12:22 |
| *** YorikSar has joined #openstack-dev | 12:53 | |
| *** adiantum has joined #openstack-dev | 13:10 | |
| *** adiantum has quit IRC | 13:18 | |
| *** bencherian has quit IRC | 13:20 | |
| *** danwent has joined #openstack-dev | 13:47 | |
| *** YorikSar has quit IRC | 13:58 | |
| *** littleidea has joined #openstack-dev | 14:09 | |
| *** YorikSar has joined #openstack-dev | 14:11 | |
| *** littleidea has quit IRC | 14:14 | |
| *** maploin has joined #openstack-dev | 14:19 | |
| *** maploin has quit IRC | 14:20 | |
| *** maploin has joined #openstack-dev | 14:20 | |
| *** Mkenneth has quit IRC | 14:29 | |
| *** stuntmachine has joined #openstack-dev | 14:32 | |
| *** stuntmachine has quit IRC | 14:35 | |
| *** adiantum has joined #openstack-dev | 14:44 | |
| *** Mkenneth has joined #openstack-dev | 14:46 | |
| *** littleidea has joined #openstack-dev | 14:53 | |
| *** danishman has joined #openstack-dev | 15:08 | |
| *** danwent has quit IRC | 15:14 | |
| *** stuntmachine has joined #openstack-dev | 15:25 | |
| *** pixelbeat has joined #openstack-dev | 15:31 | |
| chmouel | jeblair: how do I restore an abandoned change in gerrit (made by LaunchPad Sync) ? | 15:35 |
| *** zykes- has quit IRC | 15:39 | |
| *** dtroyer has joined #openstack-dev | 15:39 | |
| *** dtroyer has quit IRC | 15:40 | |
| *** crobinso has joined #openstack-dev | 15:41 | |
| *** ywu has quit IRC | 15:53 | |
| *** ywu has joined #openstack-dev | 15:54 | |
| jeblair | chmouel: there's a "restore" button on the change (if you're the owner of the change) | 16:02 |
| chmouel | jeblair: yes there is, thanks | 16:02 |
| *** Mkenneth has quit IRC | 16:35 | |
| *** danwent has joined #openstack-dev | 16:36 | |
| *** maploin has quit IRC | 16:38 | |
| *** berendt has quit IRC | 16:48 | |
| *** Mkenneth has joined #openstack-dev | 16:50 | |
| *** sdake has quit IRC | 16:53 | |
| *** nati has joined #openstack-dev | 16:57 | |
| *** Mkenneth has quit IRC | 17:06 | |
| *** hashar has joined #openstack-dev | 17:08 | |
| *** Vek has joined #openstack-dev | 17:13 | |
| *** x86brandon has joined #openstack-dev | 17:31 | |
| *** zul has quit IRC | 17:35 | |
| *** bencherian has joined #openstack-dev | 17:37 | |
| *** LinuxJedi has quit IRC | 17:50 | |
| *** LinuxJedi has joined #openstack-dev | 17:50 | |
| *** nati has quit IRC | 17:55 | |
| *** nati has joined #openstack-dev | 17:56 | |
| *** bencherian_ has joined #openstack-dev | 17:58 | |
| *** bencherian has quit IRC | 17:58 | |
| *** bencherian_ is now known as bencherian | 17:58 | |
| *** bencherian has quit IRC | 18:02 | |
| *** hashar has quit IRC | 18:13 | |
| notmyname | dweimer: ping | 18:24 |
| *** thingee has joined #openstack-dev | 18:26 | |
| dweimer | notmyname: I'm here. | 18:27 |
| notmyname | dweimer: plebian in twitter? | 18:27 |
| dweimer | notmyname: Yep, I'm really bad at choosing names. | 18:27 |
| notmyname | dweimer: I figured it would be easier to chat in here :-) | 18:27 |
| notmyname | heh. you're talking to "notmyname". If that's not bad, I'm not sure what is ;-) | 18:27 |
| notmyname | dweimer: I'm not sure why the mock/fake http_connect functions either. I'm playing with that now | 18:29 |
| dweimer | notmyname: Yeah, there's a few different status codes, but beyond that I can't tell. Replacing the mock_http_connect calls with fake will let the tests pass, but that might just be a side effect. | 18:30 |
| dweimer | notmyname: I have a side note to go through and document some of those tests. The string of status codes gets awfully confusing when trying to track down what requests hit each code. | 18:31 |
| notmyname | heh, indeed. especially as more are added for "simple" operations. (as is the case with object versioning) | 18:32 |
| notmyname | dweimer: my plan right now is to use fake_http_connect for everything and add the needed response codes to make the current tests pass | 18:32 |
| notmyname | and I hope that any real differences in the mock/fake calls can be resolved | 18:33 |
| dweimer | notmyname: I can submit a pull request for that if you want. I've gotten that far. | 18:33 |
| notmyname | cool. sure. one downside, though, is that once I submit this to gerrit the code has to be squashed into one commit. so the best I can do is to give you credit in the commit message | 18:34 |
| dweimer | notmyname: No big deal. I was mostly interested to try and understand the versioning code anyway. | 18:37 |
| notmyname | dweimer: ok. it shouldn't take me too long to get the tests cleaned up | 18:37 |
| notmyname | dweimer: does the object versioning make sense? | 18:38 |
| dweimer | notmyname: It does. I think it runs into the same >5GB file issue as post as copy though. | 18:39 |
| notmyname | dweimer: how so? the versioned object aren't >5GB | 18:40 |
| dweimer | notmyname: I'm also not sure about the extra latency due to the head costs. I'll need to run a few benchmarks. | 18:40 |
| notmyname | yes, we will have to do some tests too :-) | 18:40 |
| dweimer | notmyname: Couldn't they be though? What if I put a version header on an existing manifest? | 18:40 |
| notmyname | one of the things left to do with versioning is to decide how to make it work with large object manifests | 18:41 |
| notmyname | the simple answer is to not allow both headers to be set on an object | 18:41 |
| notmyname | but I'm not sure what it best yet | 18:41 |
| dweimer | We have some people using manifests for <5GB files as well, just for the upload threading. | 18:43 |
| dweimer | Man, adding all of the status codes was bad for the pep8 checks. | 18:43 |
| notmyname | dweimer: the good news is that HEADs are pretty cheap. yes there is latency since there are 3 of them (from x-newest). the question is if it's low enough to not impact a large cluster too much | 18:47 |
| *** danishman has quit IRC | 18:54 | |
| dweimer | notmyname: Have you guys looked into anything to reduce the per request latencies? Like spdy? | 18:55 |
| notmyname | no. we haven't generally found the per request latencies to be troublesome (and we are more concerned with lots of streams rather than single streams). so, in an effort to keep it simpler, we haven't looked at it. that being said, if you have some ideas, I'm sure I could get some people to look at it | 18:56 |
| *** nati has quit IRC | 19:00 | |
| *** bcwaldon is now known as bcwaldon_away | 19:05 | |
| dweimer | notmyname: Nothing concrete yet. I've just been thinkging about it since we have some users that would like to do things like recursively update object metdata or pull all of the checksums for verification. | 19:05 |
| dweimer | notmyname: Pull request submitted. | 19:05 |
| notmyname | looking at it | 19:05 |
| notmyname | did you mean to leave the print statements in? | 19:05 |
| dweimer | notmyname: oops not at all. | 19:05 |
| notmyname | no worries, I'll pull them out | 19:06 |
| notmyname | dweimer: print statements removed and commit pushed | 19:11 |
| *** danwent has quit IRC | 19:14 | |
| *** nati has joined #openstack-dev | 19:23 | |
| *** dalang has quit IRC | 19:26 | |
| *** RobertLaptop has joined #openstack-dev | 19:29 | |
| *** RobertLaptop has left #openstack-dev | 19:29 | |
| *** zykes- has joined #openstack-dev | 19:38 | |
| *** zykes- has quit IRC | 19:38 | |
| *** zykes- has joined #openstack-dev | 19:39 | |
| *** thingee_ has joined #openstack-dev | 19:48 | |
| *** hashar has joined #openstack-dev | 19:49 | |
| *** thingee has quit IRC | 19:50 | |
| *** thingee_ is now known as thingee | 19:50 | |
| dweimer | notmyname: hmm, there seems to be some sort of shared state in fake_http_connect. test_status_map results are affected by previous runs. | 19:50 |
| notmyname | that doesn't seem good | 19:51 |
| dweimer | notmyname: Nevermind. I just wasn't handling the -2 and -3 entries correctly so they weren't triggering excpetions. | 19:58 |
| notmyname | ah | 19:58 |
| notmyname | new pull? | 19:58 |
| *** nati has quit IRC | 19:59 | |
| *** nati has joined #openstack-dev | 20:00 | |
| dweimer | notmyname: Yep, just running a final test. | 20:01 |
| notmyname | cool, thanks | 20:01 |
| *** n0ano has joined #openstack-dev | 20:04 | |
| *** thingee has left #openstack-dev | 20:04 | |
| *** thingee has joined #openstack-dev | 20:06 | |
| *** jaypipes has quit IRC | 20:07 | |
| *** bencherian has joined #openstack-dev | 20:08 | |
| notmyname | dweimer: thanks. got it | 20:09 |
| dweimer | notmyname: Did you read the etherpad on versioning/dedup that the Nexenta dev posted a while back? | 20:13 |
| notmyname | dweimer: ya, but it's been a while | 20:13 |
| *** dtroyer has joined #openstack-dev | 20:18 | |
| dweimer | notmyname: It has a lot of ideas, but one that seems interesting is explicitly calling out the segments that make up a manifest rather than using a prefix. It could allow for testing some of the dedup ideas outside of the other changes. | 20:19 |
| *** nati has quit IRC | 20:20 | |
| *** dtroyer has quit IRC | 20:20 | |
| *** nati has joined #openstack-dev | 20:20 | |
| notmyname | dweimer: yes, but if I remember right, it required that GETs do a metadata lookup before it could actually get the object. that can introduce significant latency | 20:21 |
| notmyname | with mine, GETs remain metadata lookup free (if you aren't requesting a specific version, ie the "normal" case) | 20:21 |
| dweimer | That's a good point. For standard objects there's very little GET overhead. | 20:27 |
| *** sdake has joined #openstack-dev | 20:29 | |
| *** nati has quit IRC | 20:33 | |
| *** nati has joined #openstack-dev | 20:35 | |
| openstackgerrit | Verification of a change to openstack/nova failed: assertRaises(Exception, ...) considered harmful https://review.openstack.org/4875 | 20:36 |
| *** pixelbeat has quit IRC | 20:48 | |
| *** thingee has quit IRC | 20:49 | |
| *** eglynn has joined #openstack-dev | 20:55 | |
| openstackgerrit | Verification of a change to openstack/nova failed: Added docs on MySQL queries blocking main thread. https://review.openstack.org/4808 | 20:56 |
| *** eglynn has quit IRC | 21:01 | |
| *** hashar has quit IRC | 21:10 | |
| *** bencherian has quit IRC | 21:12 | |
| *** dtroyer has joined #openstack-dev | 21:14 | |
| *** anotherjesse_zz is now known as anotherjesse | 21:16 | |
| notmyname | dweimer: thanks for the help. I'll try to get some of my real tests pushed out soon. tests and docs (and anything that tests reveal), and this should be done | 21:23 |
| *** stuntmachine has quit IRC | 21:29 | |
| *** danwent has joined #openstack-dev | 21:32 | |
| *** dtroyer has quit IRC | 21:33 | |
| *** adjohn has joined #openstack-dev | 21:52 | |
| *** adjohn has quit IRC | 21:52 | |
| *** anotherjesse is now known as anotherjesse_zz | 22:05 | |
| *** novas0x2a|laptop has quit IRC | 22:16 | |
| *** littleidea has quit IRC | 22:18 | |
| *** littleidea has joined #openstack-dev | 22:21 | |
| *** x86brandon has quit IRC | 22:22 | |
| *** dtroyer has joined #openstack-dev | 22:30 | |
| *** x86brandon has joined #openstack-dev | 22:40 | |
| *** x86brandon has quit IRC | 22:44 | |
| *** Mandell has quit IRC | 22:44 | |
| *** danwent has quit IRC | 23:08 | |
| *** x86brandon has joined #openstack-dev | 23:09 | |
| *** anotherjesse_zz is now known as anotherjesse | 23:11 | |
| *** Mandell has joined #openstack-dev | 23:28 | |
| *** crobinso has quit IRC | 23:30 | |
| *** danwent has joined #openstack-dev | 23:34 | |
| *** jaypipes has joined #openstack-dev | 23:46 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!