| jtran | mtaylor, hey have u had a chance to look at that diablo stable issue? | 00:00 |
|---|---|---|
| mtaylor | jtran: no. someone else was looking at it, but I forget who | 00:01 |
| jtran | was it me? ;) | 00:01 |
| mtaylor | jtran: hehe | 00:01 |
| mtaylor | nope | 00:01 |
| jtran | good thing! | 00:02 |
| *** epim has quit IRC | 00:04 | |
| jtran | mtaylor, just ping / pm / email me if u figure out who it is. graci! | 00:04 |
| *** apevec has quit IRC | 00:05 | |
| mtaylor | jtran: I will! | 00:05 |
| *** jakedahn_zz is now known as jakedahn | 00:06 | |
| *** nati_ueno has quit IRC | 00:06 | |
| *** nati_ueno has joined #openstack-dev | 00:08 | |
| *** yuanz has joined #openstack-dev | 00:12 | |
| *** Mandell has quit IRC | 00:13 | |
| *** mokas has quit IRC | 00:13 | |
| *** rnirmal has quit IRC | 00:13 | |
| *** edygarcia has joined #openstack-dev | 00:18 | |
| *** s0mik has quit IRC | 00:18 | |
| *** dolphm has joined #openstack-dev | 00:19 | |
| *** s0mik has joined #openstack-dev | 00:21 | |
| *** dolphm has quit IRC | 00:23 | |
| *** s0mik has quit IRC | 00:23 | |
| *** dolphm has joined #openstack-dev | 00:24 | |
| *** Shrews has joined #openstack-dev | 00:25 | |
| *** jakedahn is now known as jakedahn_zz | 00:25 | |
| *** yuanz has left #openstack-dev | 00:28 | |
| *** danwent has quit IRC | 00:30 | |
| *** jtran has quit IRC | 00:33 | |
| *** markmcclain has joined #openstack-dev | 00:33 | |
| *** issackelly has quit IRC | 00:38 | |
| *** nati_ueno has left #openstack-dev | 00:46 | |
| *** littleidea has quit IRC | 00:46 | |
| *** mfer has joined #openstack-dev | 00:47 | |
| *** winston-d has joined #openstack-dev | 00:47 | |
| winston-d | jgriffith, hi | 00:47 |
| *** koolhead17 has joined #openstack-dev | 00:48 | |
| *** johnpur has left #openstack-dev | 00:53 | |
| *** spiffxp has quit IRC | 00:53 | |
| *** thingee has quit IRC | 00:56 | |
| *** mfer has quit IRC | 00:58 | |
| *** adjohn has quit IRC | 01:01 | |
| *** steveb_ has quit IRC | 01:01 | |
| *** adjohn has joined #openstack-dev | 01:05 | |
| *** novas0x2a|laptop has quit IRC | 01:05 | |
| *** jdurgin has quit IRC | 01:06 | |
| *** roge has quit IRC | 01:06 | |
| *** matwood has quit IRC | 01:08 | |
| *** markmcclain has quit IRC | 01:10 | |
| *** mfer has joined #openstack-dev | 01:12 | |
| *** littleidea has joined #openstack-dev | 01:14 | |
| *** Ryan_Lane has quit IRC | 01:17 | |
| *** steveb_ has joined #openstack-dev | 01:18 | |
| *** koolhead17 has quit IRC | 01:20 | |
| *** dolphm has quit IRC | 01:29 | |
| *** koolhead17 has joined #openstack-dev | 01:33 | |
| *** edygarcia has quit IRC | 01:33 | |
| *** sunxin has joined #openstack-dev | 01:35 | |
| *** dolphm has joined #openstack-dev | 01:37 | |
| *** mfer has quit IRC | 01:41 | |
| *** edygarcia has joined #openstack-dev | 01:47 | |
| *** adjohn has quit IRC | 01:50 | |
| *** Aaton is now known as Aaton_off | 01:51 | |
| *** blamar has quit IRC | 01:51 | |
| *** pixelbeat has quit IRC | 01:52 | |
| *** maurosr has joined #openstack-dev | 02:00 | |
| *** anniec has quit IRC | 02:01 | |
| *** maurosr has quit IRC | 02:05 | |
| *** matwood has joined #openstack-dev | 02:05 | |
| *** roge has joined #openstack-dev | 02:08 | |
| *** SuperSonicSound has joined #openstack-dev | 02:08 | |
| *** maurosr has joined #openstack-dev | 02:08 | |
| *** andrewbogott has quit IRC | 02:16 | |
| *** maoy has quit IRC | 02:16 | |
| *** thingee has joined #openstack-dev | 02:16 | |
| *** mokas has joined #openstack-dev | 02:18 | |
| *** PotHix has quit IRC | 02:22 | |
| *** dolphm has quit IRC | 02:27 | |
| *** jakedahn_zz is now known as jakedahn | 02:27 | |
| *** gongys has joined #openstack-dev | 02:28 | |
| *** gyee has quit IRC | 02:31 | |
| *** koolhead17 has quit IRC | 02:34 | |
| *** littleidea has quit IRC | 02:37 | |
| jgriffith | winston-d: ping | 02:48 |
| *** thingee has quit IRC | 02:53 | |
| *** matiu has joined #openstack-dev | 02:57 | |
| *** matiu has quit IRC | 02:57 | |
| *** matiu has joined #openstack-dev | 02:57 | |
| *** rnirmal has joined #openstack-dev | 03:02 | |
| *** matiu has quit IRC | 03:03 | |
| *** matiu has joined #openstack-dev | 03:04 | |
| *** chrisfer has quit IRC | 03:04 | |
| *** matiu has quit IRC | 03:07 | |
| *** matiu has joined #openstack-dev | 03:07 | |
| *** matiu has quit IRC | 03:07 | |
| *** matiu has joined #openstack-dev | 03:07 | |
| jgriffith | zul: ping | 03:08 |
| zul | jgriffith: whats up | 03:08 |
| jgriffith | zul: tgtadm: can't send the request to the tgt daemon, Transport endpoint is not connected | 03:09 |
| jgriffith | zul: This is on nova-vol | 03:09 |
| zul | jgriffith: with the devstack changes? | 03:09 |
| zul | jgriffith: log file? | 03:10 |
| jgriffith | zul: nova-vol shouldn't need the devstack changes | 03:10 |
| zul | jgriffith: log file please :) | 03:10 |
| jgriffith | zul: jdg-debug: issue tgadmin command: tgt-admin --conf /opt/stack/nova/volumes/volume-e85be7d8-a28b-4159-a506-c37ab1a70480 -e --update volume-e85be7d8-a28b-4159-a506-c37ab1a70480 | 03:10 |
| jgriffith | zul: I did a pdb here to catch the command | 03:11 |
| jgriffith | zul: Then ran tgtadm show | 03:11 |
| zul | jgriffith: can you do me a favor? can you do sudo : tgt-admin -v --conf /opt/stack/nova/volumes/volume-e85be7d8-a28b-4159-a506-c37ab1a70480 -e --update volume-e85be7d8-a28b-4159-a506-c37ab1a70480 and pastebin it | 03:13 |
| jgriffith | zul: Was wondering, can we use something like: tgtadm --lld iscsi --op new --mode target --tid 1 -T | 03:15 |
| jgriffith | zul: Or am I missing a key point here :) | 03:15 |
| zul | jgriffith: no we cant..then it wont be written to the targets.conf | 03:15 |
| *** jog0 has quit IRC | 03:16 | |
| *** matiu has quit IRC | 03:16 | |
| *** Mandell has joined #openstack-dev | 03:16 | |
| *** matiu has joined #openstack-dev | 03:18 | |
| *** matiu has quit IRC | 03:18 | |
| *** matiu has joined #openstack-dev | 03:18 | |
| *** dolphm has joined #openstack-dev | 03:19 | |
| *** edygarcia has quit IRC | 03:20 | |
| jgriffith | zul: What about tgtadm --dump >> tgt.conf ? | 03:21 |
| jgriffith | zul: BTW, I'm just trying to understand/learn | 03:21 |
| jgriffith | zul: Not criticizing etc | 03:21 |
| zul | jgriffith: no worries...isnt that a bit racy, becaue you are basically re-reading all of the iscsitargets | 03:22 |
| jgriffith | zul: Well, there is that! | 03:22 |
| jgriffith | :) | 03:22 |
| jgriffith | zul: good point | 03:22 |
| *** dubsquared has joined #openstack-dev | 03:22 | |
| *** blamar has joined #openstack-dev | 03:28 | |
| *** dachary has quit IRC | 03:41 | |
| *** dachary has joined #openstack-dev | 03:41 | |
| notmyname | zaitcev: ping | 03:47 |
| zaitcev | notmyname: pong | 03:47 |
| notmyname | you should lurk in #openstack-swift too ;-) | 03:47 |
| zaitcev | notmyname: I updated https://review.openstack.org/10582 with a postfactum comment so we have the information. | 03:48 |
| notmyname | ya, just read it | 03:48 |
| notmyname | so the traceback happens with newer version of webob if format=foobar? | 03:49 |
| notmyname | for some value of foobar | 03:49 |
| zaitcev | Yes, format=はは does it | 03:49 |
| zaitcev | However, I should note, the proxy server continues to operate. It only returns 500 for the request. | 03:50 |
| notmyname | ya | 03:50 |
| *** spiffxp has joined #openstack-dev | 03:52 | |
| *** winston-d has quit IRC | 03:54 | |
| cloudfly | so out of curiousity how many of you keep an openstack cloud in your closet? | 04:00 |
| *** edygarcia has joined #openstack-dev | 04:01 | |
| *** amotoki has joined #openstack-dev | 04:07 | |
| *** littleidea has joined #openstack-dev | 04:08 | |
| *** dolphm has quit IRC | 04:20 | |
| *** shang has quit IRC | 04:25 | |
| *** shang has joined #openstack-dev | 04:39 | |
| *** SuperSonicSound has quit IRC | 04:39 | |
| *** sacharya has quit IRC | 04:41 | |
| *** dubsquared has quit IRC | 04:44 | |
| *** anniec has joined #openstack-dev | 04:49 | |
| *** anniec_ has joined #openstack-dev | 04:51 | |
| *** roge has quit IRC | 04:53 | |
| *** amotoki has quit IRC | 04:54 | |
| *** anniec has quit IRC | 04:55 | |
| *** anniec_ is now known as anniec | 04:55 | |
| *** maoy has joined #openstack-dev | 04:59 | |
| *** zaitcev has quit IRC | 05:05 | |
| *** dolphm_ has joined #openstack-dev | 05:05 | |
| *** cp16net is now known as cp16net|away | 05:08 | |
| *** cp16net|away is now known as cp16net | 05:08 | |
| *** johnpostlethwait has joined #openstack-dev | 05:10 | |
| bcwaldon | jeblair: fun for tomorrow https://jenkins.openstack.org/job/gate-tempest-devstack-vm/5537/consoleFull | 05:15 |
| *** gongys has quit IRC | 05:18 | |
| *** blamar has quit IRC | 05:21 | |
| *** mcaway is now known as markmc | 05:25 | |
| *** edygarcia has quit IRC | 05:36 | |
| *** Drakiz has joined #openstack-dev | 05:38 | |
| *** steveb_ has quit IRC | 05:38 | |
| *** Ryan_Lane has joined #openstack-dev | 05:41 | |
| *** dolphm_ has quit IRC | 05:44 | |
| *** sunxin has quit IRC | 05:54 | |
| *** mikal has quit IRC | 05:57 | |
| *** mikal has joined #openstack-dev | 05:59 | |
| *** erikzaadi has joined #openstack-dev | 05:59 | |
| *** hattwick has quit IRC | 06:02 | |
| *** rnirmal has quit IRC | 06:03 | |
| *** littleidea has quit IRC | 06:07 | |
| *** GheRivero has joined #openstack-dev | 06:10 | |
| *** Ryan_Lane has quit IRC | 06:11 | |
| *** amotoki has joined #openstack-dev | 06:14 | |
| *** spiffxp has quit IRC | 06:17 | |
| *** vuntz has quit IRC | 06:22 | |
| *** danwent has joined #openstack-dev | 06:32 | |
| *** steveb_ has joined #openstack-dev | 06:32 | |
| *** lzyeval has joined #openstack-dev | 06:44 | |
| *** xtrusia has joined #openstack-dev | 06:50 | |
| *** johnpostlethwait has quit IRC | 07:09 | |
| *** matwood has quit IRC | 07:13 | |
| *** milner has quit IRC | 07:13 | |
| *** xtrusia_ has joined #openstack-dev | 07:16 | |
| *** Ryan_Lane has joined #openstack-dev | 07:19 | |
| *** vuntz has joined #openstack-dev | 07:25 | |
| *** blamar has joined #openstack-dev | 07:27 | |
| *** milner has joined #openstack-dev | 07:27 | |
| *** zhuadl has joined #openstack-dev | 07:29 | |
| *** EmilienM has joined #openstack-dev | 07:35 | |
| *** matiu has quit IRC | 07:38 | |
| *** sulochan has joined #openstack-dev | 07:40 | |
| *** dachary has quit IRC | 07:41 | |
| *** sulochan has left #openstack-dev | 07:41 | |
| *** eglynn has joined #openstack-dev | 07:46 | |
| *** Mandell has quit IRC | 07:52 | |
| *** Ryan_Lane has quit IRC | 07:54 | |
| *** dachary has joined #openstack-dev | 07:58 | |
| *** blamar has quit IRC | 07:58 | |
| *** zhuadl has quit IRC | 08:07 | |
| *** pixelbeat has joined #openstack-dev | 08:16 | |
| *** derekh has joined #openstack-dev | 08:17 | |
| *** lzyeval has quit IRC | 08:22 | |
| *** alex88 has joined #openstack-dev | 08:29 | |
| *** alex88 has joined #openstack-dev | 08:29 | |
| *** xtrusia has quit IRC | 08:33 | |
| *** apevec has joined #openstack-dev | 08:33 | |
| *** apevec has joined #openstack-dev | 08:33 | |
| *** xtrusia_ has left #openstack-dev | 08:34 | |
| *** xtrusia_ has joined #openstack-dev | 08:34 | |
| *** hattwick has joined #openstack-dev | 08:43 | |
| *** erikzaadi has quit IRC | 08:44 | |
| *** steveb_ has quit IRC | 08:47 | |
| *** danpb has joined #openstack-dev | 08:49 | |
| *** armaan has joined #openstack-dev | 09:04 | |
| *** armaan has left #openstack-dev | 09:04 | |
| eglynn | soren: hey, a while back you were talking on IRC about failing to boot an instance on devstack where the host's own hostname mysteriously couldn't be resolved | 09:06 |
| eglynn | soren: log message like ... 2012-06-12 20:11:12 ERROR nova.rpc.common [-] Returning exception Failed to add interface: sudo: unable to resolve host devstack-1339526523 | 09:06 |
| eglynn | soren: just wondering if you got to the bottom of it ... | 09:06 |
| *** rods has joined #openstack-dev | 09:14 | |
| *** dachary has quit IRC | 09:29 | |
| *** zhuadl has joined #openstack-dev | 09:30 | |
| soren | eglynn: hmm... | 09:34 |
| soren | eglynn: Are yo usure it was with devstack | 09:34 |
| soren | ? | 09:34 |
| eglynn | soren: just assumed fromt he hostname (devstack-1339526523) | 09:35 |
| soren | eglynn: Oh, that was from my logfile? | 09:35 |
| eglynn | soren: from IRC log, one sec ... | 09:35 |
| soren | eglynn: Yeah, then you're probably right. | 09:35 |
| eglynn | soren: http://eavesdrop.openstack.org/irclogs/%23openstack-dev/%23openstack-dev.2012-06-13.log | 09:35 |
| eglynn | soren: no worries if its lost in the mists of time at this stage ... | 09:36 |
| ttx | markmc: when did you plan to release 2012.1.2 ? Thursday ? | 09:37 |
| markmc | ttx, yes | 09:37 |
| markmc | ttx, will ask for testing at tonight's meeting - i.e. name the RC tarballs | 09:37 |
| ttx | markmc: ok, great | 09:37 |
| soren | eglynn: Ah, these were not my own instances. I was just trying to help in debugging this issue: https://bugs.launchpad.net/openstack-ci/+bug/1012595 | 09:39 |
| uvirtbot | Launchpad bug 1012595 in openstack-ci "Intermittent failure running new instances" [Critical,Fix released] | 09:39 |
| soren | eglynn: Wow. I was actually (for once!) smart enough to write down the cause. | 09:39 |
| eglynn | soren: a-ha, cool, thanks! | 09:39 |
| soren | eglynn: No problem. | 09:40 |
| * soren pats himself on the back for adding that comment on lp | 09:40 | |
| soren | There's *no* way I'd have remembered otherwise. | 09:40 |
| eglynn | soren++ | 09:40 |
| apevec | markmc, ttx - keystone has few stable/essex patches in the queue https://review.openstack.org/#/q/status:open+project:openstack/keystone+branch:stable/essex,n,z | 09:42 |
| *** timello has quit IRC | 09:42 | |
| *** Vakaris has joined #openstack-dev | 09:44 | |
| apevec | hmm, looks like some where not proposed to master first https://review.openstack.org/#/c/10365/ | 09:44 |
| *** fc__ has quit IRC | 09:47 | |
| *** fc__ has joined #openstack-dev | 09:47 | |
| *** shang has quit IRC | 09:48 | |
| ttx | markmc: I specifically added the topic to the meeting tonight, start of meeting | 09:49 |
| markmc | ttx, cool | 09:49 |
| markmc | apevec, yeah, hoping to get to them | 09:50 |
| markmc | apevec, wasn't clear at a glance how important they are | 09:50 |
| apevec | ayoung said ldap ones are important | 09:50 |
| *** dachary has joined #openstack-dev | 09:51 | |
| *** sandywalsh has joined #openstack-dev | 09:52 | |
| *** sandywalsh has quit IRC | 09:52 | |
| ttx | markmc: reading cfg code... is there any mechanism to mark an option deprecated ? Or is it just about writing special code issuing a deprecated exception if this option is still used ? | 09:57 |
| ttx | (where that option is used, rather than when that option is first defined) | 09:58 |
| markmc | ttx, yeah, all we have so far is 'deprecated_name' which is basically just for renaming options | 10:00 |
| *** shang has joined #openstack-dev | 10:01 | |
| ttx | markmc: I'm trying to find a way to detect root_helper was used to complain about that. But given how the default works, looks like I should rather be detecting when the new replacement options are /not/ used | 10:03 |
| ttx | markmc: i.e. define the new rootwrap_path and rootwrap_conf options with "None" as default value -- and complain if they are None. Does that make sense ? | 10:05 |
| *** maurosr has quit IRC | 10:07 | |
| asalkeld | any openstack-common reviewer's about? | 10:16 |
| asalkeld | https://review.openstack.org/#/c/10514/ | 10:16 |
| asalkeld | https://review.openstack.org/#/c/10398/ | 10:16 |
| asalkeld | https://review.openstack.org/#/c/10594/ | 10:17 |
| *** armaan has joined #openstack-dev | 10:26 | |
| *** armaan has left #openstack-dev | 10:26 | |
| *** kbringard has joined #openstack-dev | 10:28 | |
| markmc | ttx, hmm, not totally sure I follow, but if there is a default for an option and yet we're setting default=None, we're doing something wrong | 10:29 |
| markmc | asalkeld, apologies, I'm way behind on those | 10:29 |
| asalkeld | k no worries | 10:30 |
| markmc | ttx, why not set default=None for root_helper and warn if it's not None ? | 10:30 |
| *** darraghb has joined #openstack-dev | 10:30 | |
| asalkeld | markmc, the nova side looks like this: https://github.com/asalkeld/nova/commits/master | 10:31 |
| markmc | asalkeld, cool stuff; kudos | 10:31 |
| asalkeld | helps to see both sides | 10:31 |
| ttx | markmc: because that would break configurations relying on root_helper default value ? | 10:32 |
| ttx | markmc: you could imagine there are some configs out there just using the default value | 10:32 |
| markmc | ttx, oh, you're not removing root_helper=sudo support yet? | 10:32 |
| ttx | markmc: no, the idea is to mark it deprecated but still support it in Folsom. | 10:33 |
| markmc | ttx, ok, so warn if root_helper!=rootwrap ? | 10:33 |
| markmc | ttx, then just remove root_helper in Grizzly | 10:34 |
| markmc | ttx, if root_helper=rootwrap is still in people's Grizzly config, it'll just be ignored | 10:34 |
| ttx | markmc: that would work... do you think there is a sane default for rootwrap_config though ? | 10:35 |
| markmc | ttx, can't recall what it is? | 10:36 |
| ttx | markmc: my idea is to replace root_helper='sudo nova-rootwrap /etc/nova/rootwrap.conf' with rootwrap_config=/etc/nova/rootwrap.conf | 10:36 |
| markmc | ttx, make the default rootwrap.conf and resolve it with ConfigOpts.find_file() ? | 10:37 |
| ttx | markmc: will look into that. was concerned about the security , but this value is actually not used in the trusted path, so it should be ok. | 10:38 |
| ttx | (the value specified in sudoers is the one that is root-trusted) | 10:39 |
| ttx | markmc: thx for your help! | 10:39 |
| markmc | ttx, np | 10:39 |
| *** zhuadl has quit IRC | 10:43 | |
| apevec | ttx, swift 1.6.0 is coming today? | 10:49 |
| *** mikal has quit IRC | 11:01 | |
| *** mikal has joined #openstack-dev | 11:03 | |
| *** Vakaris has quit IRC | 11:09 | |
| *** Vakaris has joined #openstack-dev | 11:14 | |
| *** almaisan-away has quit IRC | 11:18 | |
| *** alex88 has quit IRC | 11:20 | |
| *** maoy has quit IRC | 11:22 | |
| *** dolphm has joined #openstack-dev | 11:30 | |
| *** Vakaris has quit IRC | 11:31 | |
| *** Vakaris has joined #openstack-dev | 11:40 | |
| ttx | apevec: unless it's QA-delayed again, yes | 11:42 |
| *** maurosr has joined #openstack-dev | 11:43 | |
| *** Vakaris has quit IRC | 11:58 | |
| *** chrisfer has joined #openstack-dev | 12:06 | |
| *** rkukura has left #openstack-dev | 12:08 | |
| *** Vakaris has joined #openstack-dev | 12:13 | |
| *** cbj has joined #openstack-dev | 12:22 | |
| *** utlemming has joined #openstack-dev | 12:23 | |
| *** markmcclain has joined #openstack-dev | 12:26 | |
| *** armaan has joined #openstack-dev | 12:28 | |
| *** armaan has left #openstack-dev | 12:28 | |
| *** Vakaris has quit IRC | 12:29 | |
| *** markmcclain has quit IRC | 12:30 | |
| *** lts has joined #openstack-dev | 12:35 | |
| *** darraghb has quit IRC | 12:36 | |
| *** darraghb has joined #openstack-dev | 12:39 | |
| *** sacharya has joined #openstack-dev | 12:39 | |
| *** dprince has joined #openstack-dev | 12:41 | |
| *** wiliam has joined #openstack-dev | 12:46 | |
| *** flaviamissi has joined #openstack-dev | 12:48 | |
| *** roge has joined #openstack-dev | 12:50 | |
| *** dprince has quit IRC | 12:55 | |
| *** mancdaz has joined #openstack-dev | 12:57 | |
| *** sacharya has quit IRC | 12:58 | |
| *** kzhen has joined #openstack-dev | 12:58 | |
| *** andrewsmedina has quit IRC | 13:00 | |
| *** salgado has joined #openstack-dev | 13:03 | |
| *** salgado has joined #openstack-dev | 13:03 | |
| *** kbringard has joined #openstack-dev | 13:04 | |
| *** zhuadl has joined #openstack-dev | 13:11 | |
| *** avishay has joined #openstack-dev | 13:14 | |
| *** salgado has quit IRC | 13:16 | |
| *** salgado has joined #openstack-dev | 13:17 | |
| *** Slower has quit IRC | 13:18 | |
| *** Slower has joined #openstack-dev | 13:18 | |
| *** salgado has quit IRC | 13:20 | |
| *** salgado has joined #openstack-dev | 13:23 | |
| *** salgado has joined #openstack-dev | 13:23 | |
| *** littleidea has joined #openstack-dev | 13:25 | |
| *** Dr_Who has joined #openstack-dev | 13:28 | |
| *** andrewsmedina has joined #openstack-dev | 13:28 | |
| *** markmcclain has joined #openstack-dev | 13:30 | |
| *** dprince has joined #openstack-dev | 13:35 | |
| dprince | ayoung: around? | 13:36 |
| ayoung | dprince, yes I am | 13:36 |
| dprince | ayoung: morning! | 13:36 |
| dprince | ayoung: couple more questions about PKI if you have time. | 13:37 |
| ayoung | dprince, are thes 'you are going to want to go and get another cup of cofffee" type questions? | 13:37 |
| *** sacharya has joined #openstack-dev | 13:37 | |
| dprince | ayoung: Well... I haven't had a cup of good tea yet so they won't be that good ;) | 13:38 |
| dprince | ayoung: the first issue is with running 'keystone-manage pki_setup' | 13:38 |
| ayoung | dprince, fire at Will. Will has been a total slacker lately and could use the excitement. | 13:38 |
| dprince | ayoung: Typically I think people would run that as root... or sudo. | 13:38 |
| dprince | ayoung: Which causes the resulting certs to be owned by root:root in /etc/keystone/ssl/* | 13:39 |
| dprince | ayoung: This causes problems when running keystone... from say a packaged RPM which runs the daemon as a nologin 'keystone' user. | 13:40 |
| ayoung | dprince, let me thinkg...the keystone user needs to be able to read the pkey, but no one else should. | 13:40 |
| dprince | ayoung: So I guess the question here is: Is it it expected that users will just know this and chown the files themselves? | 13:40 |
| ayoung | rest of the files should be world readable | 13:40 |
| dprince | ayoung: Correct. The private key needs to be owned by the user. | 13:40 |
| dprince | ayoung: What I'm getting at is using keystone-manage for this might end up being a bit clunky. | 13:41 |
| ayoung | But since this is a web app, the basic rule is: the user running the process should not be able to rewrite its own files... | 13:41 |
| dprince | ayoung: The issue here is the user (keystone) doesn't own the files to begin with. | 13:41 |
| ayoung | so having pkey owned by root, readable by keystone, and no readable by anyone else might be the right policy | 13:42 |
| dprince | ayoung: Yeah. Something like that. | 13:42 |
| *** littleidea has quit IRC | 13:42 | |
| dprince | ayoung: Is that something we should perhaps do in keystone-manage? Or just leave it up to the packagers to take care of. | 13:43 |
| *** littleidea has joined #openstack-dev | 13:43 | |
| ayoung | dprince, ticket it. We'll discuss during the Keystone meeting today | 13:43 |
| dprince | ayoung: okay. sounds good. | 13:43 |
| dprince | ayoung: second question is regarding the disable_pki=False flag. I really dislike double negatives to *enable* this feature. | 13:44 |
| dprince | ayoung: would it be reasonable to switch to enable_pki instead? | 13:44 |
| dprince | ayoung: Did I perhaps miss a conversation on this already? | 13:44 |
| ayoung | dprince, this will not be a double negative long term | 13:45 |
| ayoung | I chose it deliberately so as to not have to change the key when it defaults to on | 13:45 |
| ayoung | disable_pki=False will be the default, and to turn it off in the config file you will do disable_pki=True | 13:45 |
| ayoung | dprince, make sense? | 13:47 |
| dprince | ayoung: well. Can't you do the same with enable_pki? I mean we have control over the default value of boolean configs right? | 13:47 |
| danpb | IMHO having either 'enable' or 'disable' in the names of config options is bad practice | 13:47 |
| danpb | better to have a name that doesn't suggest an enablement state | 13:47 |
| danpb | eg just pki_security=True|False | 13:48 |
| dprince | danpb: I'm not opposed to this... however when using the feature I would expect to set it to 'True'. | 13:48 |
| dprince | danpb: That is my main gripe here I guess. | 13:48 |
| danpb | oh i agree that disable_pki=False is particularly bad due to the double negative | 13:49 |
| danpb | since we expect the defaults to change, IMHO, it is better to use a naming scheme as i suggested | 13:50 |
| dprince | ayoung: I'll throw up a ticket for this as well. Perhaps you could mention it at the meeting today as well? | 13:50 |
| dprince | I like danpb's suggestion's here. | 13:50 |
| dprince | ayoung: next question. The 'id' column. | 13:52 |
| dprince | ayoung: https://bugs.launchpad.net/keystone/+bug/1031191 | 13:53 |
| uvirtbot | Launchpad bug 1031191 in keystone "PKI key_size=2048 causes truncation errors on 'id' column" [High,In progress] | 13:53 |
| *** blamar has joined #openstack-dev | 13:53 | |
| *** dolphm has quit IRC | 13:53 | |
| dprince | ayoung: I'm hitting that when using the default key_size of 2048. What is the long term plan as for storing this data? Should it actually be in the database? If so do we need to make the column larger to accomidate a more secure key size. | 13:54 |
| dprince | ayoung: My quick fix for this is just to set the default key_size to 1024. | 13:54 |
| *** edygarcia has joined #openstack-dev | 13:55 | |
| *** davidha has joined #openstack-dev | 13:55 | |
| ayoung | Sounds like the ID column was not migrated in that case | 13:57 |
| ayoung | With PKI key_size of 2048, how big is the data? | 13:58 |
| *** salgado` has joined #openstack-dev | 13:58 | |
| dprince | ayoung: It was. Otherwise my work around wouldn't have worked (setting it to 1024) | 13:58 |
| davidha | hey, SWIFT RINGS: I encountered a strange issue - After building the SAIO rings based on http://docs.openstack.org/developer/swift/development_saio.html, I noticed that unpickling the rings takes forever (few seconds each) - I have ring files that I created back in the 1.4 version using the same SAIO instructions and there it unpickles in milliseconds.... | 13:58 |
| *** edygarcia has quit IRC | 13:59 | |
| dprince | ayoung: With a key_size of 2048 the data is roughly twice that size. | 13:59 |
| davidha | unziped = GzipFile(_path, 'rb') is fast but than ring_data = pickle.load(unziped) takes seconds for a small and simple SAIO ring file | 14:00 |
| davidha | I tested on two machines. | 14:00 |
| ayoung | Ah...right...so if we continue to store the token, we needto make sure that the column size stays ahead of the key size. And I tought I was overestimating with 2048... | 14:01 |
| BLZbubba | anyone here use nfs-rdma or any other infiniband technology? | 14:01 |
| *** edygarcia has joined #openstack-dev | 14:01 | |
| BLZbubba | declaring war on latency here | 14:01 |
| dprince | ayoung: I proposed we do this for now: https://review.openstack.org/#/c/10584/ | 14:01 |
| *** Mandell has joined #openstack-dev | 14:01 | |
| *** amotoki has quit IRC | 14:02 | |
| ayoung | dprince, it is a viable option. | 14:02 |
| ayoung | dprince, here's the course forward that I think makes the most sense. We use the keyring change to locally store the token, so that people don't end up generating a slew of them, and reuse them instead. We also drop storing the certificate in the database. | 14:04 |
| *** salgado has quit IRC | 14:04 | |
| ayoung | I am almost of the opinion that we should not be storing the tokens at all | 14:04 |
| *** salgado has joined #openstack-dev | 14:04 | |
| dprince | ayoung: Okay That sounds better. This seems like an awful lot of token data to keep in the DB as is. | 14:05 |
| ayoung | the only reason I stored them in the PKI patch was because we had done so in the past. It provides a reasonable way to check for revocation, but it really is not the smartest approach | 14:05 |
| davidha | has there been any change in the ring structure/code or in the code creating rings since 1.4.3 ? | 14:05 |
| *** blamar has quit IRC | 14:05 | |
| sdague | jaypipes: did the white box testing land over the weekend? | 14:05 |
| dprince | davidha: I've noticed creating Swift rings is slow as well... I'd be interested to try your suggestions here and see if it speeds things up. | 14:06 |
| dprince | davidha: I'm dong the same thing... SAIO swift node for dev/testing, etc. | 14:06 |
| dprince | doing | 14:06 |
| *** salgado has joined #openstack-dev | 14:07 | |
| *** salgado has joined #openstack-dev | 14:07 | |
| dprince | ayoung: That is all my questions sir. Thanks for the help. | 14:07 |
| ayoung | dprince, good stuff | 14:07 |
| dprince | ayoung: Oh yeah. When I set the key_size to 1024... I was able to run Torpedo/Tempest against a PKI enabled keystone. | 14:08 |
| dprince | ayoung: So it works! | 14:08 |
| dprince | ayoung: I have some more things to try but initially it looks to be functioning. | 14:09 |
| ayoung | dprince, is there some way we can measure network traffic both with and without PKI enabled? | 14:09 |
| dprince | ayoung: I suppose we could look into it. Is that your main concern here? | 14:10 |
| ayoung | danpb, dprince as far as enabling/disabling PKI, I would prefer to use something more documenting, like: token_strategy=PKI vs token_strategy=uuid | 14:10 |
| ayoung | the disable/enable hack is just that. | 14:10 |
| *** spiffxp has joined #openstack-dev | 14:12 | |
| *** mdomsch has quit IRC | 14:13 | |
| *** salgado has quit IRC | 14:14 | |
| *** salgado has joined #openstack-dev | 14:15 | |
| *** salgado has joined #openstack-dev | 14:15 | |
| *** openstackjenkins has quit IRC | 14:15 | |
| *** timello has joined #openstack-dev | 14:16 | |
| *** openstackjenkins has joined #openstack-dev | 14:16 | |
| jaypipes | sdague: sorry, no it did not. hope to have it done today | 14:18 |
| dansmith | jaypipes: just doing some housekeeping.. we're still waiting on something to unblock this, right? https://review.openstack.org/#/c/10169/ | 14:19 |
| danpb | ayoung: yes, that sounds even better | 14:19 |
| *** armaan has joined #openstack-dev | 14:19 | |
| *** armaan has left #openstack-dev | 14:19 | |
| *** Mandell has quit IRC | 14:19 | |
| jaypipes | dansmith: yeah, not sure. I just rekicked jenkins to see if that was a spurious failure on the first run. | 14:20 |
| *** zhuadl has quit IRC | 14:20 | |
| dansmith | jaypipes: okay, I just recheck'd a few minutes ago and it came back with the same failure from a few days ago | 14:22 |
| jaypipes | dansmith: oh... ok. any idea why the failure is occurring? (it seems to be in the test setUp method?) | 14:23 |
| dansmith | it doesn't fail here, and if you see the patch, it's just removing a skip ... on an unrelated test to the failure | 14:23 |
| *** dubsquared has joined #openstack-dev | 14:23 | |
| jaypipes | right... that's why I'm curious about it :) | 14:24 |
| jaypipes | dansmith: back in a sec... gotta fire up a VPN. :( | 14:24 |
| *** jaypipes has quit IRC | 14:24 | |
| *** andrewsmedina has left #openstack-dev | 14:27 | |
| *** jaypipes has joined #openstack-dev | 14:27 | |
| *** creiht has quit IRC | 14:29 | |
| *** creiht has joined #openstack-dev | 14:29 | |
| *** ChanServ sets mode: +v creiht | 14:29 | |
| *** markmcclain has quit IRC | 14:30 | |
| *** sc68cal has quit IRC | 14:37 | |
| *** danwent has quit IRC | 14:37 | |
| *** sc68cal has joined #openstack-dev | 14:38 | |
| *** blamar has joined #openstack-dev | 14:43 | |
| *** armaan has joined #openstack-dev | 14:48 | |
| *** armaan has left #openstack-dev | 14:48 | |
| *** dolphm has joined #openstack-dev | 14:51 | |
| markmc | ttx, btw, see nova.common.deprecated | 14:53 |
| *** jakedahn is now known as jakedahn_zz | 14:53 | |
| ttx | markmc: yeah, saw that and will use | 14:55 |
| *** sc68cal has quit IRC | 14:55 | |
| davidha | dprince: I am not complaining about the ring creation being slow - instead I noticed that the loading of a pre-created ring file by the swift server became much slower than before | 14:55 |
| dprince | davidha: Ah. I see. Thanks for the clarification then. | 14:56 |
| *** rkukura has joined #openstack-dev | 14:58 | |
| *** jakedahn_zz is now known as jakedahn | 14:59 | |
| *** e1mer has quit IRC | 14:59 | |
| *** jakedahn is now known as jakedahn_zz | 15:00 | |
| *** s0mik has joined #openstack-dev | 15:02 | |
| *** matiu has joined #openstack-dev | 15:03 | |
| *** bcwaldon has quit IRC | 15:03 | |
| *** danwent has joined #openstack-dev | 15:04 | |
| *** renier has quit IRC | 15:05 | |
| *** sc68cal has joined #openstack-dev | 15:05 | |
| *** sc68cal has joined #openstack-dev | 15:05 | |
| *** renier has joined #openstack-dev | 15:05 | |
| *** blamar has quit IRC | 15:06 | |
| *** blamar has joined #openstack-dev | 15:06 | |
| *** s0mik has quit IRC | 15:09 | |
| *** sc68cal has quit IRC | 15:11 | |
| *** sulochan has joined #openstack-dev | 15:12 | |
| *** rnirmal has joined #openstack-dev | 15:12 | |
| *** blamar has quit IRC | 15:12 | |
| jgriffith | ttx: ping | 15:13 |
| *** blamar has joined #openstack-dev | 15:13 | |
| ttx | jgriffith: pong | 15:13 |
| jgriffith | ttx: 1031121 isn't a duplicate, it's related but it's a *new* problem | 15:14 |
| jgriffith | ttx: regardless, should be fixed here shortly anyway. Just FYI | 15:15 |
| ttx | jgriffith: ok then maybe open a cinder task on the other bug to cover it in cinder as well | 15:17 |
| *** timello has quit IRC | 15:17 | |
| ttx | (if they are separate bugs) | 15:17 |
| jgriffith | ttx: Well... | 15:17 |
| *** sc68cal has joined #openstack-dev | 15:19 | |
| *** sc68cal has joined #openstack-dev | 15:19 | |
| *** ewindisch has joined #openstack-dev | 15:25 | |
| *** halfss has joined #openstack-dev | 15:26 | |
| *** sc68cal has quit IRC | 15:26 | |
| *** dtroyer has quit IRC | 15:28 | |
| sdague | markmc: process question on launch & openstack-common. given that openstack-common doesn't have releases when do you move items from fix-committed to fix-released? | 15:29 |
| *** ayoung has quit IRC | 15:30 | |
| *** dachary has quit IRC | 15:30 | |
| *** s0mik has joined #openstack-dev | 15:31 | |
| *** halfss has quit IRC | 15:32 | |
| *** gyee has joined #openstack-dev | 15:32 | |
| markmc | sdague, I think all I've done so far is mark all the ones which I reckon were fixed in essex - i.e. the code for it was (probably) copied into the various projects | 15:33 |
| markmc | sdague, I should do the same for folsom milestones | 15:33 |
| jaypipes | vishy: is there an equivalent to cpu_allocation_ratio in Diablo? | 15:33 |
| jaypipes | vishy: in other words, is there a way to force overutilization? | 15:33 |
| *** heckj has joined #openstack-dev | 15:33 | |
| *** timello has joined #openstack-dev | 15:34 | |
| *** ayoung has joined #openstack-dev | 15:36 | |
| *** s0mik has quit IRC | 15:39 | |
| dansmith | jaypipes: hmm, I guess jenkins does "if ($user != "jaypipes") fail_on(rand() > 0.2);" | 15:41 |
| jaypipes | dansmith: lol. | 15:42 |
| *** anniec has quit IRC | 15:43 | |
| vishy | jaypipes: afaik there is no ram check in diablo | 15:47 |
| vishy | jaypipes: at least with simple scheduler | 15:47 |
| vishy | jaypipes: but you can change a flag to overallocate cores | 15:48 |
| *** rohitk has joined #openstack-dev | 15:48 | |
| *** dachary has joined #openstack-dev | 15:48 | |
| *** cp16net is now known as cp16net|away | 15:48 | |
| *** s34n has joined #openstack-dev | 15:49 | |
| vishy | jaypipes: the flag is max_cores | 15:49 |
| *** maoy has joined #openstack-dev | 15:49 | |
| vishy | jaypipes: assuming you're using simple scheduler | 15:49 |
| *** salgado is now known as salgado-lunch | 15:51 | |
| davidkranz | eglynn: Ping. | 15:51 |
| *** cloudvirt has joined #openstack-dev | 15:51 | |
| eglynn | davidkranz: hey | 15:51 |
| davidkranz | eglynn: Can you answer Jay's question in https://review.openstack.org/#/c/10552/ | 15:52 |
| *** nunosantos has joined #openstack-dev | 15:52 | |
| * eglynn looking ... | 15:52 | |
| eglynn | davidkranz: on the tracker bug, or? | 15:53 |
| davidkranz | eglynn: Yes, is there a bug ticket that your patch is going to fix? | 15:53 |
| eglynn | davidkranz: yep, referencing that in gerrit now (bug #1022961) | 15:54 |
| uvirtbot | Launchpad bug 1022961 in nova "[novaclient] root-password fails and does not return error" [Undecided,In progress] https://launchpad.net/bugs/1022961 | 15:54 |
| davidkranz | eglynn: Sorry, I should have followed the link and seen that. Thanks. | 15:54 |
| *** dachary1 has joined #openstack-dev | 15:55 | |
| eglynn | davidkranz: np! (thanks BTW for pushing ahead with the tempest change) | 15:55 |
| *** cp16net|away is now known as cp16net | 15:57 | |
| *** dachary has quit IRC | 15:57 | |
| *** dolphm has quit IRC | 16:00 | |
| *** dolphm has joined #openstack-dev | 16:01 | |
| *** rohitk_1 has joined #openstack-dev | 16:03 | |
| dansmith | markmc: presumably you've seen (or will see) that eglynn had a question on the api lock checking patch, right? | 16:04 |
| *** rohitk has quit IRC | 16:04 | |
| markmc | dansmith, yeah, I'm reviewing your latest patch now | 16:04 |
| *** spiffxp has quit IRC | 16:05 | |
| dansmith | markmc: cool | 16:05 |
| *** rohitk_1 has quit IRC | 16:05 | |
| *** GheRivero_ has joined #openstack-dev | 16:05 | |
| *** datsun180b has joined #openstack-dev | 16:05 | |
| markmc | eglynn, bit confused by your question tbh? | 16:05 |
| *** dolphm has quit IRC | 16:05 | |
| markmc | eglynn, the server has to assume the request comes from an old client and that lock_checked=False | 16:06 |
| eglynn | markmc, dansmith: just to clarify, I didn't mean the lock_checked param in the RPC message | 16:06 |
| markmc | eglynn, so the client would always need to set lock_checked=True to avoid the server-side lock checking | 16:06 |
| *** rohitk has joined #openstack-dev | 16:06 | |
| markmc | eglynn, ah, I was thinking | 16:06 |
| eglynn | markmc, dansmith: I just meant the param in the rpcapi method | 16:06 |
| markmc | eglynn, have some smarts in rpcapi.py itself to add it to the RPC message | 16:06 |
| markmc | eglynn, ok, I'll just reply in gerrit - I've a more evil idea | 16:07 |
| eglynn | markmc: exactly, seeing as we'd always want it in that version of rpcapi, right? | 16:07 |
| eglynn | markmc: cool | 16:07 |
| markmc | eglynn, only where the decorator is on the calling method | 16:07 |
| eglynn | markmc: yep, but always for those individual methods (usually 1:1 correspondence between api methods and rpcapi methods called) | 16:08 |
| *** jtran has joined #openstack-dev | 16:08 | |
| markmc | eglynn, only where we actually remember to add the decorator to the caller :) | 16:08 |
| ayoung | danpb, are there any F17 Jeos images available? | 16:09 |
| markmc | eglynn, forgetting it would be a bug, granted | 16:09 |
| *** amotoki has joined #openstack-dev | 16:09 | |
| eglynn | markmc: the lock_checked param seems to add lots of noise to the rpcapi methods, when we'd never want to *deliberately* set it false AFAICS | 16:10 |
| markmc | eglynn, agree | 16:10 |
| dansmith | markmc: uhm, if not in the rpc message, then where were you asking for it | 16:11 |
| dansmith | ? | 16:11 |
| *** hemna has joined #openstack-dev | 16:11 | |
| markmc | dansmith, I am asking for it in the RPC message (on the wire) | 16:11 |
| markmc | dansmith, it's a question of how the client code gets it there | 16:11 |
| *** sc68cal has joined #openstack-dev | 16:12 | |
| *** sc68cal has joined #openstack-dev | 16:12 | |
| markmc | dansmith, I'll have a comment in gerrit in a minute | 16:12 |
| *** sniperd has quit IRC | 16:12 | |
| * dansmith gets confused with api.py being called the "client" | 16:12 | |
| *** sniperd has joined #openstack-dev | 16:12 | |
| hemna | I'm having an issue with nova volume. I can't seem to get the LogginVolumeDriver to log out when I try and create a volume | 16:14 |
| hemna | I'm using the LoggingVolumeDriver because my driver isn't being called when I try and create a volume either. | 16:15 |
| hemna | I figured the problem was with my driver | 16:15 |
| hemna | but it doesn't seem like it, as the LoggingVolumeDriver's create_volume never gets called either. | 16:15 |
| *** bencherian has joined #openstack-dev | 16:17 | |
| markmc | dansmith, funny, calling it the client has helped me a tonne | 16:17 |
| *** gongys has joined #openstack-dev | 16:17 | |
| markmc | dansmith, the nova-api service as a client of the nova-compute service | 16:17 |
| markmc | dansmith, helps especially when considering the RPC API versioning | 16:18 |
| markmc | dansmith, anyway, commented in gerrit now | 16:18 |
| dansmith | markmc: I understand why you're calling it that, it just isn't how I think of it :) | 16:18 |
| markmc | russellb, could you take a gander at https://review.openstack.org/10355 ? | 16:18 |
| russellb | NO! | 16:18 |
| markmc | russellb, it actually helps no-db-compute a bunch | 16:18 |
| * russellb kids | 16:19 | |
| markmc | russellb, and I have an evil idea that I hope you can tell me is going to break something | 16:19 |
| markmc | dansmith, how do you think of it? | 16:19 |
| * markmc curious | 16:19 | |
| dansmith | markmc: for me, "client" is the client (python-novaclient, or other), api.py is the "external interface" and manager.py is the "implementation" | 16:20 |
| dansmith | not that it's right or anything... | 16:20 |
| russellb | rpcapi.py is the direct interface for manager.py | 16:20 |
| markmc | dansmith, yeah, that works ... except when you need to take into account the RPC versioning | 16:21 |
| russellb | api.py uses rpcapi.py when completing things requires rpc | 16:21 |
| russellb | other code outside of api.py uses rpcapi.py by the way, so you need to be very careful | 16:21 |
| russellb | and search for other rpcapi uses | 16:21 |
| russellb | to make sure nothing is bypassing these checks now | 16:21 |
| dansmith | russellb: but all users of that are trusted? | 16:22 |
| russellb | well yeah, but they used to have the instance lock checked for them, and now would not | 16:22 |
| russellb | if they were using rpcapi directly | 16:22 |
| dansmith | russellb: it seems strange to me to trust anything that could be on another box over an untrusted network | 16:22 |
| * russellb has only read the commit message so far btw | 16:22 | |
| russellb | i saw a much earlier version of this but i think it has mostly been purged, heh | 16:23 |
| russellb | no-db-messaging rotting the brain | 16:23 |
| dansmith | russellb: no, that's still on the back burner, this is slightly different | 16:23 |
| russellb | k ... let me read the code and shut up | 16:23 |
| russellb | so that my comments are actually informed | 16:23 |
| markmc | dansmith, nice, russellb telling you to shut up | 16:24 |
| markmc | thems fighting words | 16:24 |
| * russellb was telling himself to shut up | 16:24 | |
| russellb | to clarify. | 16:24 |
| markmc | heh :) | 16:24 |
| dansmith | jaypipes: can I borrow that giraffe tendon? | 16:24 |
| russellb | O.O | 16:24 |
| dansmith | markmc: I'm not sure what DRY means in this context | 16:24 |
| russellb | Don't Repeat Yourself | 16:24 |
| russellb | it's hipster coding lingo | 16:24 |
| dansmith | heh | 16:24 |
| * markmc trying to pretend he's hip | 16:24 | |
| markmc | in fact, I just read some of a Ruby book once | 16:25 |
| markmc | actually, no | 16:25 |
| *** gongys has quit IRC | 16:25 | |
| markmc | The Pragmatic Programmer is all over DRY | 16:25 |
| markmc | that's old curmudgeoney long-beard territory | 16:25 |
| dansmith | markmc: could we slap the lock_checked into the context? I think the eglynn suggested that in the first place | 16:25 |
| eglynn | markmc, dansmith: how about just adding lock_checked=True to the kwargs passed by the checks_instance_lock decorator? | 16:26 |
| dansmith | I did a little thinking on it early, can't remember if it was a problmem, but then you suggested the argument | 16:26 |
| markmc | dansmith, context does not sound crazy at all | 16:26 |
| dansmith | the argument was so much more work, and I really wanted to write all that useless code, so... | 16:26 |
| dansmith | markmc: I feel like I'm missing some design definition about what things should be in the context, what components are trusted, etc | 16:27 |
| dansmith | granted, I've only been faking the role of "nova developer" for a handful of weeks yet.. | 16:27 |
| *** markvoelker has joined #openstack-dev | 16:27 | |
| markmc | dansmith, we do all sorts of whacky stuff with context - see e.g. read_deleted | 16:27 |
| dansmith | yeah | 16:28 |
| eglynn | dansmith, markmc: yep, context seems a neat place to put it without disrupting loads of method signatures | 16:28 |
| russellb | :-/ | 16:28 |
| russellb | feels like turning context into a dumping ground of random state info | 16:28 |
| markmc | request state info | 16:28 |
| markmc | not instance state | 16:28 |
| hemna | isn't a nova volume driver's create_volume method supposed to be called when you run "nova volume-create --display_name foo 10" ? | 16:29 |
| *** spiffxp has joined #openstack-dev | 16:29 | |
| dansmith | markmc, russellb: by that definition, "lock_checked" would be valid, "is_locked" would not.. | 16:29 |
| russellb | markmc: yeah, i suppose this is request state data ... | 16:29 |
| dansmith | (for context) | 16:29 |
| * russellb nods | 16:29 | |
| eglynn | yep | 16:29 |
| russellb | markmc: "The server side check actually looks up the instance from the DB again." ... where? | 16:29 |
| markmc | russellb, you wrote it, you should know | 16:30 |
| dansmith | if I do this enough times, I'll have every possible implementation coded and the core folks can just pick the one they like and merge it :D | 16:30 |
| russellb | ah, that get_lock shenanigans | 16:30 |
| markmc | russellb, self._get_lock() | 16:30 |
| *** mokas has quit IRC | 16:30 | |
| markmc | yeah | 16:30 |
| russellb | got it | 16:30 |
| *** Gordonz has joined #openstack-dev | 16:32 | |
| *** eglynn is now known as hungry-eglynn | 16:32 | |
| *** _0x44 has quit IRC | 16:32 | |
| *** dachary1 has quit IRC | 16:33 | |
| *** _0x44 has joined #openstack-dev | 16:33 | |
| *** issackelly has joined #openstack-dev | 16:33 | |
| *** issackelly has quit IRC | 16:34 | |
| *** issackelly has joined #openstack-dev | 16:36 | |
| dansmith | markmc: okay, so, "try it in context" is a go? | 16:37 |
| russellb | one sec, reading some code to make sure it's ok | 16:37 |
| russellb | yeah, it gets a +1 from me | 16:38 |
| russellb | left it as a comment on the review too | 16:38 |
| * russellb will start watching this review now | 16:38 | |
| dansmith | russellb: okay, thanks | 16:39 |
| markmc | dansmith, yep, I think so | 16:39 |
| dansmith | I've got a lot of merging to do since the current set had things I need to keep, a ton of crap to remove, and russellb has significantly changed master underneath :/ | 16:39 |
| russellb | you might want to start over ... | 16:40 |
| *** armaan has joined #openstack-dev | 16:40 | |
| russellb | and just copy in the parts you need | 16:40 |
| *** armaan has left #openstack-dev | 16:40 | |
| russellb | yeah, i'm causing a lot of movement in nova-compute right now ... | 16:40 |
| russellb | it's why i haven't been watching reviews as much as usual | 16:40 |
| jaypipes | dansmith: lol. | 16:40 |
| dansmith | well, I think I can start from patch 2 as the path of least resistance, but I'll see | 16:40 |
| dansmith | heh, things you love to hear from ~nova-core: "you might want to just start over, loser" | 16:41 |
| russellb | lol | 16:41 |
| russellb | I didn't say "loser" ... the code is good, i just know it's going to be a pain to update to the context approach | 16:41 |
| dansmith | yeah, yeah :D | 16:42 |
| * russellb keeps digging his hole | 16:42 | |
| *** Mandell has joined #openstack-dev | 16:42 | |
| dansmith | when I'm done with these two infernal patches, I'm never touching the locking or task_state stuff again.. evar. | 16:43 |
| russellb | heh | 16:43 |
| jaypipes | vishy: thx | 16:43 |
| russellb | dansmith: once you have an update, ping me on IRC, i'll try to help get it in asap | 16:43 |
| *** jakedahn_zz is now known as jakedahn | 16:44 | |
| dansmith | russellb: woohoo! the guilt trip worked! :D | 16:44 |
| ayoung | pixelbeat, where are the F17 Folsom RPMS kept? | 16:44 |
| russellb | dansmith: correct | 16:44 |
| dansmith | heh | 16:44 |
| *** markmcclain has joined #openstack-dev | 16:45 | |
| *** hungry-eglynn is now known as eglynn | 16:46 | |
| *** danpb has quit IRC | 16:47 | |
| ayoung | russellb, eglynn markmc do we have a public Yum repo for Folsom/F17 RPMs? I thought we did, but my link was to the F16 repo | 16:47 |
| apevec | ayoung, folsom RPMs are rawhide only, if there's popular demand, I could start Folsom Preview for f17 | 16:47 |
| ayoung | apevec, Ah, thanks | 16:47 |
| russellb | ayoung: not that i knwo of. | 16:47 |
| russellb | apevec to the rescue | 16:47 |
| apevec | so is there popular demand?? | 16:48 |
| *** Mandell has quit IRC | 16:48 | |
| russellb | there is ayoung demand | 16:48 |
| russellb | and he is pretty popular. | 16:48 |
| *** Mandell has joined #openstack-dev | 16:48 | |
| apevec | ok, and ayoung==popular | 16:48 |
| russellb | ha | 16:48 |
| ayoung | actually, I am acting as a go-between for the Hyper-V devs, believe it or not | 16:49 |
| *** rohitk has quit IRC | 16:49 | |
| *** johnpur has joined #openstack-dev | 16:49 | |
| *** ChanServ sets mode: +v johnpur | 16:49 | |
| *** alexpilotti has joined #openstack-dev | 16:49 | |
| apevec | ayoung, oh, then we can trade w/ them? Maybe few free MSDN subscriptions? :) | 16:50 |
| *** andrewbogott has joined #openstack-dev | 16:51 | |
| ayoung | apevec, I think I am going to stay out of that particular horse trade, thanks | 16:51 |
| apevec | joking | 16:51 |
| apevec | so I'll re-use the same repo, just that f16 will stay Essex for now and update f17 to folsom snapshot | 16:52 |
| s34n | devstack's stack.sh dumps what looks like the contents of a shell script with every line prefixed with '+' onto my screen, but doesn't seem to actually execute those commands. | 16:53 |
| *** markmc is now known as mcaway | 16:53 | |
| *** Mandell_ has joined #openstack-dev | 16:54 | |
| ayoung | apevec, +1 | 16:55 |
| russellb | s34n: that's the result of "set +o xtrace" | 16:55 |
| ayoung | alexpilotti, meet apevec, | 16:55 |
| russellb | it's showing the commands as they are executed | 16:55 |
| * ayoung work here is done | 16:55 | |
| s34n | russellb: but they don't execute | 16:56 |
| russellb | well i'm pretty sure devstack doesn't have any magic "pretend but don't actually do it" mode :-) | 16:57 |
| *** derekh has quit IRC | 16:57 | |
| russellb | but hey, i could be wrong. | 16:57 |
| apevec | ayoung, alexpilotti - I'll start koji scratch rebuilds now, you can expect repo populated by tomorrow http://repos.fedorapeople.org/repos/apevec/openstack-preview/fedora-17/ | 16:57 |
| *** mokas has joined #openstack-dev | 16:57 | |
| *** Mandell has quit IRC | 16:58 | |
| *** mnaser has joined #openstack-dev | 16:58 | |
| *** mnaser has joined #openstack-dev | 16:58 | |
| alexpilotti | apevec, ayoung: tx, you are referring to the Folsom RPMs? | 16:58 |
| apevec | yes, I'll rebuild f2 RPMs for f17 | 16:59 |
| ayoung | alexpilotti, yeah, apevec had an F16 repo for Essex | 16:59 |
| ayoung | apevec, but those are going to be hand generated, right? You aren't going to automitacially update thenm periodically | 16:59 |
| alexpilotti | apevec, ayoung: great, I'll check the repo tomorrow then! | 16:59 |
| hemna | anyone familiar with nova volume drivers? | 17:00 |
| pixelbeat | ayoung, dprince builds folsom rpms for his testing that you could try. They'll be more up to date | 17:00 |
| apevec | yes, periodically | 17:00 |
| apevec | ah right, there are those trunk-chase rpms, they're rebuilt on each commit | 17:00 |
| apevec | koji builds are updated per snapshot | 17:01 |
| apevec | pixelbeat, were are they published? | 17:01 |
| apevec | that's smokestack right? | 17:02 |
| alexpilotti | pixelbeat: do you have a link to dprince's repo? tx | 17:02 |
| s34n | is anybody here part of devstack development? | 17:02 |
| hemna | s34n, I presume this entire channel...but I can't get help bleh | 17:03 |
| ayoung | s34n, devstack makes heavy use of screen. perhaps that is what is messing you up. WHat are you seeing? | 17:08 |
| pixelbeat | alexpilotti, Sorry I don't know if there is a publically available repo. The package sources are here at least https://github.com/fedora-openstack/openstack-nova | 17:09 |
| *** cloudvirt has quit IRC | 17:09 | |
| alexpilotti | pixelbeat: ok tx! | 17:09 |
| *** andrewbogott has quit IRC | 17:10 | |
| ayoung | alexpilotti, I'd say use devstack for F17, and we'll get a Rawhide (F18) process moving shortly. We'll also close the loop on F17/Folsom | 17:10 |
| *** andrewbogott has joined #openstack-dev | 17:11 | |
| *** jdurgin has joined #openstack-dev | 17:11 | |
| *** mnaser has quit IRC | 17:12 | |
| *** bcwaldon has joined #openstack-dev | 17:13 | |
| *** anniec has joined #openstack-dev | 17:13 | |
| *** sstent has quit IRC | 17:16 | |
| s34n | ayoung: I'm trying to run stack.sh on F17. I've run into a few snags. | 17:17 |
| s34n | ayoung: for instance, selinux prevents it from creating the stack user | 17:17 |
| ayoung | s34n, I've been through it a couple times. What are you seeing | 17:17 |
| *** sstent has joined #openstack-dev | 17:17 | |
| ayoung | sudo setenforce Permissive | 17:17 |
| s34n | ayoung: sudo can't run without a tty | 17:17 |
| ayoung | one sec... that is a suders change | 17:18 |
| s34n | ayoung: I know how to set permissive. I did that to overcome the first snag | 17:18 |
| ayoung | s34n, yeah, there is a no_tty falg or something to sudo, too | 17:18 |
| s34n | ayoung: I also changed sudo temporarily | 17:18 |
| ayoung | I ssh in to the machiens, so I didn't see that problem | 17:18 |
| ayoung | machine | 17:19 |
| s34n | I'm just thinking that those things could be build into the script | 17:19 |
| s34n | ayoung: I am also ssh'ed in | 17:19 |
| *** cloudvirt has joined #openstack-dev | 17:19 | |
| ayoung | s34n, well, turning off SELinux is evil | 17:19 |
| *** cloudvirt has quit IRC | 17:20 | |
| jaypipes | mikal: ah, I understand now.... (re: the logging and instance_uuid thing...) | 17:21 |
| *** cloudvirt has joined #openstack-dev | 17:21 | |
| s34n | I don't like turning off selinux, but with openstack, you don't have much choice | 17:21 |
| ayoung | s34n, you might be able to use the -t option to ssh to force a pseudo-tty to be used. I'm guessing that Fedora is doing something like that under the covers, which is why I wasn't seeing the problem | 17:22 |
| ayoung | s34n, so no show stoppers yet...what else | 17:22 |
| *** s0mik has joined #openstack-dev | 17:22 | |
| *** jog0 has joined #openstack-dev | 17:22 | |
| ayoung | actually, I am running ssh -X, which I think does it for me | 17:23 |
| *** amotoki_ has joined #openstack-dev | 17:26 | |
| s34n | ayoung: after those two bumps, it looks like the script completes. | 17:26 |
| *** troytoman-away is now known as troytoman | 17:26 | |
| s34n | ayoung: I presume I have to open up some ports for it, though. | 17:27 |
| s34n | apparently more than just port 80 | 17:27 |
| ayoung | s34n, or just shut down iptables, but yes | 17:27 |
| ayoung | let me see what I have 1 sec | 17:27 |
| *** s0mik has quit IRC | 17:27 | |
| ayoung | heh. pretty much everything. I've been on too many projects | 17:28 |
| *** amotoki has quit IRC | 17:28 | |
| ayoung | 35357 and 5000 for Keystone....I'd have to dig for the rest | 17:28 |
| jaypipes | anyone know Jim Fehlig's IRC nick? | 17:29 |
| *** salgado-lunch is now known as salgado | 17:30 | |
| *** amotoki_ is now known as amotoki | 17:30 | |
| s34n | hmm. shutting down iptables wasn't enough. I get a forbidden in my browser. | 17:31 |
| *** darraghb has quit IRC | 17:31 | |
| *** Ryan_Lane has joined #openstack-dev | 17:33 | |
| avishay | Hi all. I just got some code merged into cinder, and it was written by me and a co-worker. Only my name appears in the commit and Authors file since I submitted it. How can I best share the credit? | 17:35 |
| jaypipes | avishay: actually a very good question. I'm not sure :( jeblair or mtaylor? | 17:37 |
| avishay | It's 2100 lines of code - don't want to hog all the glory :) | 17:38 |
| bhuvan | avishay: you can add their name in AUTHORS.in file | 17:38 |
| bhuvan | it'll be included to AUTHORS file that goes in release package | 17:38 |
| avishay | bhuvan: where is the file? | 17:39 |
| bhuvan | create one, if it doesn't exist | 17:40 |
| *** thingee has joined #openstack-dev | 17:41 | |
| *** PotHix has joined #openstack-dev | 17:43 | |
| avishay | bhuvan: OK thanks! | 17:44 |
| *** lcheng has joined #openstack-dev | 17:52 | |
| ayoung | s34n, Forbidden is probably not IPTables, but rather a response code from an API due to an Auth problem. Make sure keystone token-get and glance image-list work from the command line first | 17:52 |
| *** Aaton_off is now known as Aaton | 17:56 | |
| *** sulochan has quit IRC | 17:58 | |
| jeblair | avishay: also, add a "Co-authored-by:" tag to the commit msg | 17:59 |
| jeblair | avishay: i believe there's been some talk of using that in the automatic AUTHORS generation, plus it's there in the commit msg for people to see | 18:00 |
| avishay | jeblair: the commit got merged - too late? | 18:00 |
| jeblair | avishay: yep. sorry. next time you write a huge patch with a coworker. :) | 18:01 |
| vishy | comstud: any idea if you (or anyone) is using return_reservation_id parameter for create servers? | 18:01 |
| avishay | jeblair: would it be OK to add our names to the file itself? it contains only our driver | 18:01 |
| *** dolphm has joined #openstack-dev | 18:02 | |
| *** thingee has quit IRC | 18:03 | |
| *** dolphm_ has joined #openstack-dev | 18:04 | |
| *** dachary has joined #openstack-dev | 18:04 | |
| jeblair | avishay: as bhuvan said, you can add your names to AUTHORS.in for the project. I'm unaware of whether there's policy about adding individual names to files. | 18:05 |
| *** epim has joined #openstack-dev | 18:06 | |
| *** rohitk has joined #openstack-dev | 18:08 | |
| avishay | jeblair: OK - thanks a lot! | 18:09 |
| *** steveb_ has joined #openstack-dev | 18:10 | |
| bhuvan | jaypipes: mtaylor: can you please review/approve https://review.openstack.org/#/c/10345/ and https://review.openstack.org/#/c/10346/? | 18:11 |
| *** thingee has joined #openstack-dev | 18:15 | |
| jeblair | bhuvan: what do you think about waiting for mtaylor to finish his work on PBR and then using PBR instead for that? | 18:16 |
| jeblair | bhuvan: (if you haven't seen it, pbr does all the authors/setuptools work as an external library) | 18:17 |
| *** rohitk_1 has joined #openstack-dev | 18:17 | |
| hemna | when does a volume driver's create_volume method get called? I can't seem to get it called at all | 18:17 |
| jeblair | bhuvan: sorry, to be specific, i was talking about 10345 | 18:18 |
| *** dolphm_ has quit IRC | 18:19 | |
| *** rohitk has quit IRC | 18:19 | |
| *** dolphm has quit IRC | 18:19 | |
| avishay | hemna: see volume/manager.py - create_volume() | 18:20 |
| avishay | self.driver.create_volume(volume_ref) | 18:20 |
| hemna | been poking around in there, but I can't seem to even get a manager.py create_volume() called | 18:21 |
| hemna | I have a LOG.debug at the top of manager.py's create_volume, but it never gets called | 18:21 |
| hemna | I'm trying nova volume-create --display_name foo 11 | 18:21 |
| hemna | nada | 18:21 |
| hemna | using the LoggingVolumeDriver | 18:22 |
| *** danwent has quit IRC | 18:22 | |
| avishay | does it create the volume, or get stuck somewhere before? | 18:22 |
| bhuvan | jeblair: ok, yep it can wait until pbr is complete | 18:23 |
| hemna | I see the entry in the database | 18:23 |
| hemna | and nova volume-create doesn't error out | 18:23 |
| hemna | but the manager.py::create_volume is never called | 18:24 |
| bcwaldon | jeblair: I just pushed python-glanceclient v0.2.0! | 18:24 |
| jeblair | bhuvan: cool, thanks. btw, have you noticed, i'm adding tests? https://review.openstack.org/#/q/status:open+project:openstack-ci/zuul,n,z | 18:24 |
| bcwaldon | jeblair: when should it show up in pypi? | 18:24 |
| bcwaldon | jeblair: it just showd up! | 18:24 |
| jeblair | bcwaldon: Awesome! Like in a minute. | 18:24 |
| bcwaldon | jeblair: never mind! | 18:24 |
| *** andrewbogott_ has joined #openstack-dev | 18:25 | |
| *** andrewbogott_ has joined #openstack-dev | 18:25 | |
| *** andrewbogott has quit IRC | 18:26 | |
| *** andrewbogott_ is now known as andrewbogott | 18:26 | |
| *** andrewbogott is now known as andrewbogott_afk | 18:26 | |
| *** andrewbogott_afk is now known as andrewbogott | 18:26 | |
| avishay | hemna: check the logs - maybe you have a trace somewhere | 18:27 |
| *** rohitk has joined #openstack-dev | 18:27 | |
| hemna | I'm tailing the nova-volume.log | 18:27 |
| hemna | the driver's check_for_setup_error() is called and passes (LoggingVolumeDriver) | 18:27 |
| *** andrewbogott is now known as andrewbogott_afk | 18:29 | |
| *** andrewbogott_afk is now known as andrewbogott | 18:29 | |
| *** andrewbogott_ has joined #openstack-dev | 18:29 | |
| *** andrewbogott_ has quit IRC | 18:30 | |
| *** andrewbogott_ has joined #openstack-dev | 18:30 | |
| *** andrewbogott_ is now known as andrewlurks | 18:30 | |
| *** andrewbogott is now known as andrewbogott_afk | 18:30 | |
| *** rohitk_1 has quit IRC | 18:30 | |
| *** andrewbogott_afk is now known as andrewbogott | 18:30 | |
| *** andrewlurks has quit IRC | 18:30 | |
| hemna | avishay, http://paste.openstack.org/show/19843/ | 18:30 |
| *** dachary has quit IRC | 18:31 | |
| *** rohitk_1 has joined #openstack-dev | 18:32 | |
| *** nati_ueno has joined #openstack-dev | 18:32 | |
| avishay | hemna: don't know... | 18:32 |
| hemna | this is stock essex on ubuntu 12.04 | 18:33 |
| *** Gordonz has quit IRC | 18:34 | |
| *** Gordonz has joined #openstack-dev | 18:34 | |
| *** rohitk has quit IRC | 18:36 | |
| bhuvan | jeblair: tests for zuul ... sounds cool! | 18:36 |
| *** reed has joined #openstack-dev | 18:37 | |
| jeblair | a rhyme every time | 18:37 |
| avishay | hemna, sorry :( | 18:40 |
| *** vuntz has quit IRC | 18:41 | |
| hemna | avishay, what is the driver's get_volume_stats() supposed to return? | 18:42 |
| *** gabrielhurley has joined #openstack-dev | 18:43 | |
| jtran | ping dtroyer_zzz | 18:43 |
| avishay | hemna: good question - no driver implements it :) | 18:44 |
| hemna | the manager.py's _report_driver_status gets called every N minutes and calls the driver's get_volume_stats to check "capabilities" | 18:48 |
| *** vuntz has joined #openstack-dev | 18:48 | |
| hemna | which always returns None, because only the VolumeDriver implements and it's hard coded to return None | 18:49 |
| *** johnpostlethwait has joined #openstack-dev | 18:50 | |
| *** nati_ueno has quit IRC | 18:52 | |
| *** novas0x2a|laptop has joined #openstack-dev | 18:52 | |
| *** nati_ueno has joined #openstack-dev | 18:52 | |
| *** novas0x2a|laptop has quit IRC | 18:52 | |
| *** danwent has joined #openstack-dev | 18:53 | |
| *** novas0x2a|laptop has joined #openstack-dev | 18:55 | |
| avishay | hemna: correct. I assume you're implementing a new driver? | 18:56 |
| hemna | yes | 18:56 |
| hemna | I'm on a new project at HP...writing a volume driver to support 3Par and some of our other SAN devices | 18:56 |
| *** spiffxp has quit IRC | 18:56 | |
| hemna | with the hope of open sourcing the driver back to openstack | 18:56 |
| avishay | hemna: Isn't 3par supported already? | 18:56 |
| hemna | not directly | 18:57 |
| avishay | hemna: do you know duncan thomas? if i remember correctly he's from HP. | 18:57 |
| hemna | not off hand | 18:57 |
| hemna | HP is a huge co. :) | 18:58 |
| hemna | HP has an app called Storage Provisioning Manager (SPM) that sits in front of our SAN devices (3Par, LeftHand, EVA, XP, etc) | 18:58 |
| hemna | SPM is what customers use to create LUNs, and exports those to servers | 18:59 |
| hemna | I'm writing the driver for openstack to talk to SPM | 18:59 |
| hemna | with the primary objective of supporting 3Par | 18:59 |
| avishay | hemna, got it, OK | 19:00 |
| avishay | hemna, try to go into #openstack-cinder and talk to duncan...he reviewed my driver, and since i believe he's from HP, he could probably help you out | 19:00 |
| zykes- | hemna: will it be public ? | 19:00 |
| hemna | zykes-, that's the hope | 19:01 |
| hemna | I'm pushing for it | 19:01 |
| zykes- | anyone from reddwarf around ? | 19:01 |
| *** GheRivero_ has quit IRC | 19:05 | |
| hemna | any idea what nova volume's api.py has in relationship to the manager.py ? | 19:06 |
| hemna | trying to track down the call path to when the manager's create_volume is called | 19:07 |
| tr3buchet | eglynn: can you tell me why you refactored the network cache refresh code while fixing the floating ip bug 1029911 | 19:07 |
| uvirtbot | Launchpad bug 1029911 in nova "Associate an already associated Floating IP to server is returning 202" [Undecided,Fix committed] https://launchpad.net/bugs/1029911 | 19:07 |
| *** avishay has quit IRC | 19:08 | |
| tr3buchet | eglynn: https://review.openstack.org/#/c/10532/ | 19:09 |
| *** novas0x2a|laptop has quit IRC | 19:10 | |
| *** novas0x2a|laptop has joined #openstack-dev | 19:10 | |
| tr3buchet | eglynn: i have to step out, ping me | 19:10 |
| russellb | tr3buchet: he may be out for the day (in Europe) | 19:11 |
| Dr_Who | n00b q: when doing run_tests.sh for say nova, is there a file where stdout is getting saved to ? trying to debug a test case | 19:17 |
| *** rnirmal has quit IRC | 19:20 | |
| comstud | vishy: i do not know of anyone using it | 19:22 |
| *** sulochan has joined #openstack-dev | 19:26 | |
| *** mokas has quit IRC | 19:28 | |
| *** apevec has quit IRC | 19:28 | |
| *** s0mik has joined #openstack-dev | 19:29 | |
| *** sandywalsh has joined #openstack-dev | 19:29 | |
| mcaway | tr3buchet, seemed pretty clear to me - @refresh_cache only applies to the instance arg of associate_floating_ip | 19:29 |
| mcaway | tr3buchet, to fix the bug, you also need to clear the cache for orig_instance | 19:30 |
| *** mokas has joined #openstack-dev | 19:30 | |
| *** markmcclain has quit IRC | 19:31 | |
| *** Aaton is now known as Aaton_off | 19:31 | |
| *** rnirmal has joined #openstack-dev | 19:34 | |
| *** markmcclain has joined #openstack-dev | 19:38 | |
| *** gabrielhurley has quit IRC | 19:39 | |
| *** davidha has quit IRC | 19:40 | |
| *** littleidea_ has joined #openstack-dev | 19:40 | |
| hemna | so it looks like when nova volume-create is called, the api creates an entry in the db and then makes an rpc call to stuff an event into rabbitmq | 19:41 |
| *** markmcclain has quit IRC | 19:42 | |
| *** markmcclain has joined #openstack-dev | 19:42 | |
| cloudfly | that would make sense | 19:42 |
| *** littleidea has quit IRC | 19:42 | |
| *** littleidea_ is now known as littleidea | 19:43 | |
| *** dtroyer has joined #openstack-dev | 19:43 | |
| *** markmcclain1 has joined #openstack-dev | 19:43 | |
| *** davidha has joined #openstack-dev | 19:44 | |
| hemna | I don't see any errors in the rpc call to push the event into rabbitmq | 19:45 |
| hemna | but I'm not seeing anything on the other side...I presume there has to be something polling for events in rabbitmq | 19:46 |
| *** markmcclain has quit IRC | 19:46 | |
| *** jtran has quit IRC | 19:47 | |
| eglynn | tr3buchet: I'm back | 19:47 |
| eglynn | tr3buchet: so the problem is that when a floating IP was reassigned from one instance to another, the refresh_cache decroate *only* refreshed the info cahce for the second instance | 19:49 |
| *** mnewby has joined #openstack-dev | 19:49 | |
| hemna | anyone know what is listening to rabbitmq for events and how those events get dispatched? | 19:49 |
| eglynn | tr3buchet: the fix was simply to force the refresh on the original instance also | 19:49 |
| eglynn | tr3buchet: the refactoring was a small code re-org to make this second (explicit) refresh convenient | 19:50 |
| *** reed has quit IRC | 19:50 | |
| eglynn | tr3buchet: makes sense, or? | 19:51 |
| cloudfly | hemna nova-volume would be responding | 19:51 |
| cloudfly | nova-manage service list | 19:52 |
| cloudfly | looke for nova-volume responding systems with :) | 19:52 |
| hemna | hrmm nova-manage service list craps out | 19:54 |
| hemna | "unable to open database file None None" | 19:55 |
| cloudfly | nova conf problem maybe? | 19:55 |
| cloudfly | or corrupt db | 19:55 |
| hemna | nova volume-list works | 19:55 |
| *** mnewby has quit IRC | 19:55 | |
| cloudfly | interesting | 19:55 |
| cloudfly | you using rabbitmq? | 19:56 |
| cloudfly | try rabbitmqctl list_queues | 19:56 |
| cloudfly | or whatever the command is that is very similar to that | 19:56 |
| hemna | http://paste.openstack.org/show/19848/ | 19:56 |
| hemna | list_queues lists a bunch of queues | 19:57 |
| cloudfly | okay | 19:57 |
| cloudfly | well maybe strace the nova-manage call or check logs | 19:57 |
| cloudfly | something is definitely wonky there | 19:57 |
| hemna | can see the queues in the paste | 19:57 |
| cloudfly | yeah i mean if the queues are there and not going nuts i'd assume the issue isn't rabbit | 19:58 |
| cloudfly | and that means the database | 19:58 |
| cloudfly | or foreign issues | 19:58 |
| *** dolphm has joined #openstack-dev | 19:58 | |
| cloudfly | but without nova-manage working it's hard to localize | 19:58 |
| cloudfly | blindly anyways | 19:59 |
| hemna | /var/log/nova-manage.log is 0 bytes | 19:59 |
| hemna | hrmm | 19:59 |
| *** mnewby has joined #openstack-dev | 19:59 | |
| hemna | strace nova-manage service list | 19:59 |
| hemna | looks like it's trying to open sqllite | 19:59 |
| hemna | my nova db is in mysql...not sure those are related though | 19:59 |
| cloudfly | where is it getting its config from? | 19:59 |
| cloudfly | i would assume it's polling /etc/nova/nova.conf for db data | 20:00 |
| hemna | should be /etc/nova/nova.conf | 20:00 |
| hemna | (updated the paste with the strace info) | 20:00 |
| cloudfly | anything wonky in there? | 20:00 |
| cloudfly | maybe an over write caused by config management or something? | 20:01 |
| *** jtran has joined #openstack-dev | 20:01 | |
| *** anotherjesse_zz is now known as anotherjesse | 20:01 | |
| hemna | man I dunno | 20:02 |
| hemna | it all looks ok | 20:03 |
| hemna | nothing about sqlite | 20:03 |
| *** dolphm has quit IRC | 20:04 | |
| *** mcaway is now known as markmc | 20:07 | |
| cloudfly | got me | 20:08 |
| *** timello has quit IRC | 20:08 | |
| cloudfly | i'd check out in strace see where it's opening it's configs from | 20:08 |
| *** mnewby has quit IRC | 20:11 | |
| tr3buchet | eglynn: there is a recursion bug | 20:12 |
| eglynn | tr3buchet: how does it show up? | 20:13 |
| tr3buchet | eglynn: also i would prefer if you disassociate the floating ip on the original instance, thus refreshing the cache on it, before associating to a new instance | 20:13 |
| eglynn | tr3buchet: well it does currently disassociate the floating ip from the original instance before associating it with the new instance | 20:15 |
| *** lloydde has joined #openstack-dev | 20:15 | |
| tr3buchet | eglynn: it should be two operations | 20:15 |
| tr3buchet | associating with a new instance should fail | 20:15 |
| eglynn | tr3buchet: but IIUC that does not cause the refresh to occur | 20:15 |
| dansmith | russellb: okay, it was less trouble than I thought.. my brain is a bit fried on this now, but I think patch set 5 is everything we talked about :) | 20:15 |
| russellb | yay | 20:16 |
| eglynn | tr3buchet: do you mean that you disagree with the implicit disassociation? | 20:16 |
| eglynn | tr3buchet: (that was pre-existing logic) | 20:16 |
| tr3buchet | eglynn: disassociating a floating ip causes the cache to refresh | 20:16 |
| tr3buchet | for the instance it was associated with | 20:16 |
| eglynn | tr3buchet: IIUC, the refresh only occurs if the disassociate is initiated by the network api service | 20:17 |
| eglynn | tr3buchet: whereas the implicit disassociate occurs within the nova-network service itself | 20:18 |
| eglynn | tr3buchet: (i.e. not decaorated with @refresh_cache) | 20:18 |
| *** wiliam has quit IRC | 20:18 | |
| tr3buchet | eglynn: there should not be any implicit disassociate | 20:19 |
| eglynn | tr3buchet: well that's a different bug to raise if you disagree with it, but as I said that was the pre-existing logic | 20:20 |
| eglynn | tr3buchet: my patch simply fixed the code that was already there ... | 20:20 |
| *** EmilienM has left #openstack-dev | 20:21 | |
| eglynn | tr3buchet: the recursion issue you mention, how does that manifest? | 20:21 |
| tr3buchet | eglynn: i'm not sure exactly. but the decorator calls a decorated function | 20:22 |
| *** cloudvirt has quit IRC | 20:22 | |
| tr3buchet | someone said it was happening | 20:23 |
| tr3buchet | i looked at the code and saw it was quite possible | 20:23 |
| tr3buchet | and then i couldnt' figure out why the refactor of the decorator | 20:23 |
| eglynn | tr3buchet: a-ha, I see | 20:24 |
| *** devcamca- is now known as devcamcar | 20:24 | |
| tr3buchet | i have a mind to undo the refactor becuase I can't see how the original bug has anything to do with it | 20:25 |
| eglynn | tr3buchet: the reason for changing the _get_instance_nw_info() call to get_instance_nw_info() was because the quantumv2.API only supports the later | 20:25 |
| *** blamar_ has joined #openstack-dev | 20:25 | |
| eglynn | tr3buchet: I can propose a patch reverting the get_instance_nw_info() call to _get_instance_nw_info(), so no recursion | 20:26 |
| tr3buchet | then the quantumv2.api is wrong | 20:26 |
| tr3buchet | eglynn: can you name a reason for refactoring the decorator? | 20:26 |
| tr3buchet | i'm curious | 20:27 |
| eglynn | tr3buchet: yep, I'll split the get_instance_nw_info()/_get_instance_nw_info() on quantumv2.API so | 20:27 |
| tr3buchet | huh? | 20:27 |
| *** blamar has quit IRC | 20:27 | |
| *** blamar_ is now known as blamar | 20:27 | |
| *** dprince has quit IRC | 20:28 | |
| eglynn | tr3buchet: that was a reply your previous statement on quantum, not your request for an explantion on the refactor | 20:28 |
| eglynn | trbuchet: OK, on the refactor ... | 20:28 |
| eglynn | tr3buchet: the reason was simply to make it more convenient to allow an *explicit* update of the cache on a particular instance | 20:29 |
| eglynn | tr3buchet: the cache refresh driven by the decorator applies only to the target instance | 20:29 |
| eglynn | tr3buchet: with the result that the nw info for the original instance became stale | 20:29 |
| eglynn | tr3buchet: what exactly isn't clear about the intent? | 20:30 |
| *** mnewby has joined #openstack-dev | 20:32 | |
| *** spiffxp has joined #openstack-dev | 20:37 | |
| tr3buchet | eglynn: ah i see, i missed that you were calling the function from the associate api call | 20:38 |
| tr3buchet | eglynn: nothing isn't clear now, i just very much dislike the business of associating a floating ip that is already associated, instead of failing | 20:38 |
| *** cloudvirt has joined #openstack-dev | 20:39 | |
| tr3buchet | eglynn: i can fix that later | 20:39 |
| tr3buchet | eglynn: if you wouldn't mind, change that to _get_instance_nw_info | 20:39 |
| tr3buchet | eglynn: the quantumv2 api can be broken or it can manage the cache itself | 20:40 |
| russellb | dansmith: test problems ... this is where you go "I LOVE DECORATORS SO MUCH" | 20:40 |
| dansmith | russellb: heh, yeah.. | 20:41 |
| dansmith | I guess this is cinder using rpcapi directly? | 20:41 |
| russellb | dunno, haven't dug into it yet | 20:41 |
| russellb | looked like a decorator related problem | 20:41 |
| russellb | not getting the number of args it expected | 20:41 |
| eglynn | tr3buchet: https://review.openstack.org/10626 | 20:41 |
| dansmith | russellb: ah, yeah, I actually had this fixed and nuked it from my patch by mistake | 20:42 |
| *** gabrielhurley has joined #openstack-dev | 20:42 | |
| dansmith | it's a simple fix | 20:42 |
| eglynn | tr3buchet: sure, its debatable whether an implicit disassociation should occur ... but as I said, that was pre-existing logic, and my intent was simply to fix the code that was there already ... | 20:43 |
| russellb | cool | 20:43 |
| *** johnpostlethwait has quit IRC | 20:45 | |
| dansmith | russellb: you know you've been round-and-round too many times when you don't recognize your own code from your own patch :D | 20:45 |
| russellb | heh yep | 20:45 |
| *** johnpostlethwait has joined #openstack-dev | 20:45 | |
| jaypipes | vishy: is there a comparable option to cpu_allocation_ratio for memory allocation? thx in advance | 20:51 |
| jaypipes | vishy: NM, found it. | 20:51 |
| vishy | jaypipes: cool | 20:51 |
| jaypipes | vishy: for Diablo, is there an equivalent for ram_allocation_ratio? | 20:51 |
| sdague | jaypipes: so, that whitebox testing patch didn't seem to progress further. Is someone actively working that? Would love to see that in. | 20:51 |
| *** dolphm has joined #openstack-dev | 20:51 | |
| *** Aaton_off is now known as Aaton | 20:52 | |
| jaypipes | sdague: yes, sorry man, it's slow going today.. it's my next review. sorry for the delay | 20:52 |
| vishy | jaypipes: i answered that question yesterday | 20:52 |
| jaypipes | vishy: question from me? | 20:52 |
| vishy | jaypipes: diablo doesn't key off of ram for simple_scheduler | 20:52 |
| sdague | jaypipes: well, it needs work, it still fails jenkins | 20:52 |
| vishy | jaypipes: yes | 20:52 |
| vishy | jaypipes: it keys off of cores, you can change max_cores | 20:53 |
| jaypipes | vishy: ok, ty. | 20:53 |
| jaypipes | vishy: yes, I saw your reply about max_cores... didn't know if there was an equiv for ram. thx for the info. | 20:53 |
| sdague | jaypipes: not trying to harass you too much, just trying to figure out if that's going to progress shortly, or if I should take a different approach in digging into some of these bugs :) | 20:53 |
| *** johnpostlethwait has quit IRC | 20:54 | |
| *** johnpostlethwait has joined #openstack-dev | 20:54 | |
| jaypipes | sdague: frankly, I've been asking rohit to make changes to that patchset and he hasn't done the changes yet... it's getting a bit frustrating. | 20:55 |
| jaypipes | sdague: I'm half tempted to fork that patchset and just fix up myself, but I don't want to step on feelings. | 20:57 |
| sdague | jaypipes: definitely understand | 20:57 |
| jaypipes | (most of the code originated from my original smoke test branch anyway) | 20:57 |
| *** Rajesh_ has joined #openstack-dev | 20:57 | |
| dansmith | jaypipes: the giraffe-tendon-wielding developer suddenly cares about feelings? | 20:57 |
| jaypipes | dansmith: I've softened up apparently. | 20:58 |
| dansmith | eesh. | 20:58 |
| *** markmcclain1 has quit IRC | 20:58 | |
| vishy | jgriffith: ping | 20:59 |
| jgriffith | vishy: Yo | 20:59 |
| *** guitarza1 has joined #openstack-dev | 20:59 | |
| *** maurosr has quit IRC | 21:00 | |
| *** guitarzan is now known as guitarzan_ | 21:00 | |
| *** guitarza1 is now known as guitarzan | 21:00 | |
| vishy | jgriffith: so there is one more change required to chuck's iscsi stuff? we need to get it into nova | 21:01 |
| jgriffith | vishy: Yes, he's working on a patch right now | 21:01 |
| jgriffith | vishy: Minor tweak to the tgtadm call | 21:01 |
| *** Rajesh_ has quit IRC | 21:01 | |
| zul | patch is up for review | 21:02 |
| jgriffith | vishy: If he doesn't have time I'll just submit it this afternoon and be done with it | 21:02 |
| jgriffith | zul: Excellent... no email yet, do you have a link? | 21:02 |
| zul | jgriffith/vishy: just running the test stuff first | 21:02 |
| *** Rajesh_ has joined #openstack-dev | 21:02 | |
| jgriffith | zul: Cool... that's what I was going to do but if you've got it covered even better | 21:02 |
| *** guitarzan has quit IRC | 21:03 | |
| jgriffith | zul: Great, those are the changes I had and they worked | 21:03 |
| *** guitarzan has joined #openstack-dev | 21:03 | |
| *** zaitcev has joined #openstack-dev | 21:04 | |
| eglynn | tr3buchet: can you re-approve https://review.openstack.org/10626? ... those devstack failures smell like transients | 21:05 |
| vishy | jgriffith: any progress on db migration code? | 21:05 |
| *** anotherjesse is now known as anotherjesse_zz | 21:06 | |
| jgriffith | vishy: very little | 21:06 |
| jgriffith | vishy: Need to decide, db's on same hosts only etc | 21:06 |
| vishy | jgriffith: I would do it as two steps | 21:06 |
| *** Shrews has quit IRC | 21:07 | |
| vishy | dump of required tables to text | 21:07 |
| vishy | import text | 21:07 |
| vishy | so you can do it anywhere | 21:07 |
| jgriffith | vishy: excellent | 21:07 |
| jgriffith | vishy: Ok, that's next on my list this week | 21:07 |
| vishy | jgriffith: although that may make it a little tough. you could pickle the objects perhaps, but json would clearly be best. | 21:08 |
| *** mnewby has quit IRC | 21:08 | |
| vishy | jgriffith: or we could just provide example scripts for mysqldump and let someone else figure out how to do it for postgres | 21:08 |
| jgriffith | vishy: I was leaning towards json, and having it be an *example* that someone could leverage | 21:09 |
| jgriffith | vishy: Just to alleviate some of the fear that things are so different | 21:09 |
| jgriffith | vishy: The more interesting question I had was live connections etc | 21:09 |
| jgriffith | vishy: So should we require no attachments? Or try to do it live etc | 21:10 |
| vishy | jgriffith: documenting the mysql one seems like an easy first step | 21:10 |
| davidha | Hi | 21:10 |
| vishy | mysqldump + import | 21:10 |
| vishy | jgriffith: have we cleaned up the tables in cinder yet? | 21:10 |
| jgriffith | vishy: Yeah, at leat I took a first pass at it | 21:10 |
| jgriffith | vishy: Ripped out the majority of the excess | 21:11 |
| jgriffith | vishy: I'm good with the mysqldump to start. Just didn't know that would be an option for those not using mysql. | 21:12 |
| jgriffith | vishy: But if it's meant to be an example and not a tool then perfect | 21:12 |
| *** ywu has quit IRC | 21:12 | |
| jgriffith | vishy: I started writing it like any other migration that we do currently, just going to a *new* db | 21:13 |
| jgriffith | vishy: But I like this idea better :) | 21:13 |
| vishy | jgriffith: yeah not that for sure | 21:13 |
| *** lts has quit IRC | 21:13 | |
| vishy | jgriffith: i would implement it as nova-manage + cinder-manage commands | 21:13 |
| *** Rajesh_ has quit IRC | 21:14 | |
| vishy | (but doing the mysqldump first is way better) | 21:14 |
| jgriffith | vishy: Yeah, I'm going to start with that | 21:14 |
| vishy | jgriffith: we did something like that for the auth migration to keystone | 21:14 |
| jgriffith | vishy: I'll send somethign your way tomorrow to look at and use as a starting point | 21:14 |
| vishy | k | 21:14 |
| *** markmc is now known as mcaway | 21:14 | |
| jgriffith | vishy: If you happen to have a pointer to what was used for keystone I'll check it out | 21:14 |
| *** Rajesh_ has joined #openstack-dev | 21:15 | |
| jgriffith | vishy: else I'll poke around and see if I can find it | 21:15 |
| zul | jgriffith: cinder gerrit prop has been updated as well | 21:15 |
| jgriffith | zul: Thanks Chuck! | 21:15 |
| zul | bbl | 21:15 |
| vishy | jgriffith: https://github.com/openstack/nova/blob/stable/essex/bin/nova-manage#L217 | 21:16 |
| jgriffith | vishy: nice... thanks! | 21:17 |
| vishy | jgriffith: although honestly since the tables are exactly the same i think mysqldump is what most people will use | 21:17 |
| *** markvoelker has quit IRC | 21:17 | |
| jgriffith | vishy: Agree, but I like to see how other folks did it as well, and depending on how it goes maybe expand. | 21:17 |
| jgriffith | vishy: I'm more interested in how to switch over the endpoints etc on a live setup | 21:18 |
| jgriffith | vishy: But that can wait until the sqldump is done | 21:18 |
| *** maoy has quit IRC | 21:20 | |
| *** dubsquared has quit IRC | 21:21 | |
| *** rkukura has quit IRC | 21:22 | |
| *** s0mik has quit IRC | 21:24 | |
| *** ohnoimdead has joined #openstack-dev | 21:26 | |
| *** Rajesh_ has quit IRC | 21:26 | |
| *** epim has quit IRC | 21:27 | |
| *** Rajesh_ has joined #openstack-dev | 21:27 | |
| *** Rajesh_ has quit IRC | 21:29 | |
| *** heckj has quit IRC | 21:29 | |
| *** cloudvirt has quit IRC | 21:33 | |
| jgriffith | zul: Sighh... one more problem in the nova-volume code still | 21:33 |
| jgriffith | zul: Let me know when you're back | 21:33 |
| *** s0mik has joined #openstack-dev | 21:33 | |
| vishy | jgriffith: is it an easy fix? | 21:34 |
| jgriffith | vishy: Probably http://paste.openstack.org/show/19865/ | 21:35 |
| jgriffith | vishy: I can just manually delete /dev/mapper files prior to the lvm call | 21:35 |
| jgriffith | vishy: But I should probably figure out what's changed | 21:35 |
| *** s0mik has quit IRC | 21:36 | |
| *** e1mer has joined #openstack-dev | 21:36 | |
| *** s0mik has joined #openstack-dev | 21:37 | |
| vishy | jgriffith: does it eventually succeed? | 21:38 |
| jgriffith | vishy: Nope | 21:38 |
| *** s0mik has quit IRC | 21:38 | |
| *** heckj has joined #openstack-dev | 21:39 | |
| vishy | jgriffith: so the tgatdm command doesn't actually close the volume properly | 21:39 |
| vishy | ? | 21:39 |
| jgriffith | vishy: It's either that or it's not getting called, checking now | 21:40 |
| jgriffith | vishy: think I got it.. testing now | 21:45 |
| *** flaviamissi has quit IRC | 21:46 | |
| *** cloudvirt has joined #openstack-dev | 21:47 | |
| *** arosen has quit IRC | 21:51 | |
| *** sulochan has quit IRC | 21:52 | |
| vishy | jog0: reviewed | 21:52 |
| vishy | jog0: ping me if my question about filtering isn't clear | 21:52 |
| jog0 | vishy: thanks will do | 21:53 |
| *** cloudvirt has quit IRC | 21:54 | |
| *** cloudvirt has joined #openstack-dev | 21:54 | |
| ttx | jgriffith: you're next on #openstack-meeting :) | 21:54 |
| jgriffith | ttx: ready | 21:55 |
| *** rnirmal has quit IRC | 21:56 | |
| *** dolphm has quit IRC | 21:57 | |
| *** kbringard has quit IRC | 22:01 | |
| *** littleidea has quit IRC | 22:07 | |
| *** littleidea has joined #openstack-dev | 22:07 | |
| *** arosen has joined #openstack-dev | 22:10 | |
| *** gabrielhurley has quit IRC | 22:14 | |
| *** steveb_ has quit IRC | 22:16 | |
| tr3buchet | eglynn: thanks btw! | 22:16 |
| *** sacharya has quit IRC | 22:16 | |
| tr3buchet | quick resolution there | 22:16 |
| *** Dr_Who has quit IRC | 22:18 | |
| *** cloudvirt has quit IRC | 22:20 | |
| *** datsun180b has quit IRC | 22:21 | |
| *** heckj has quit IRC | 22:24 | |
| *** Gordonz has quit IRC | 22:25 | |
| *** littleidea_ has joined #openstack-dev | 22:25 | |
| *** lcheng has quit IRC | 22:25 | |
| *** littleidea has quit IRC | 22:25 | |
| *** littleidea_ is now known as littleidea | 22:25 | |
| *** lloydde has quit IRC | 22:28 | |
| *** lloydde has joined #openstack-dev | 22:28 | |
| *** salgado has quit IRC | 22:31 | |
| *** mnewby has joined #openstack-dev | 22:40 | |
| *** andrewbogott is now known as andrewbogott_afk | 22:42 | |
| *** mnewby has quit IRC | 22:47 | |
| *** eglynn has quit IRC | 22:49 | |
| *** ecarlin has joined #openstack-dev | 22:50 | |
| *** danwent has quit IRC | 22:52 | |
| *** alexpilotti has quit IRC | 22:53 | |
| *** blamar_ has joined #openstack-dev | 22:53 | |
| *** blamar has quit IRC | 22:57 | |
| *** blamar_ is now known as blamar | 22:57 | |
| *** blamar has quit IRC | 23:01 | |
| *** roge has quit IRC | 23:01 | |
| *** edygarcia has quit IRC | 23:02 | |
| *** edygarcia has joined #openstack-dev | 23:03 | |
| *** matiu has quit IRC | 23:03 | |
| *** spiffxp has quit IRC | 23:03 | |
| *** dolphm has joined #openstack-dev | 23:08 | |
| *** mnewby has joined #openstack-dev | 23:10 | |
| clarkb | bcwaldon: fyi https://jenkins.openstack.org/job/glance-coverage/565/console failed when the gate test https://jenkins.openstack.org/job/gate-glance-python27/1103/ | 23:11 |
| clarkb | bcwaldon: for change https://review.openstack.org/10630 probably an intermittent issue, but also something to be aware of | 23:12 |
| *** dolphm has quit IRC | 23:12 | |
| *** rods has quit IRC | 23:14 | |
| *** lloydde has quit IRC | 23:14 | |
| *** ohnoimdead has quit IRC | 23:15 | |
| *** cloudvirt has joined #openstack-dev | 23:16 | |
| *** mattstep has quit IRC | 23:19 | |
| *** amotoki_ has joined #openstack-dev | 23:24 | |
| Ryan_Lane | is there no way to get the ec2 style id back when using the nova api? | 23:25 |
| *** amotoki has quit IRC | 23:26 | |
| *** amotoki has joined #openstack-dev | 23:28 | |
| *** amotoki_ has quit IRC | 23:30 | |
| s34n | is anybody else running the current devstack on f17? | 23:34 |
| asalkeld | ya me | 23:34 |
| asalkeld | it didn't behave for a while | 23:34 |
| s34n | asalkeld: and it ran for you out of the box? | 23:34 |
| * s34n is a slow typer | 23:34 | |
| asalkeld | but after re-runing it couple of times is sprung to life | 23:35 |
| s34n | really? | 23:35 |
| asalkeld | I can paste my config | 23:35 |
| s34n | just re-running it somehow made things different? | 23:35 |
| s34n | or was it re-run, tweak, re-run, tweak... ? | 23:36 |
| asalkeld | I know weird | 23:36 |
| asalkeld | http://fpaste.org/09hV/ | 23:36 |
| asalkeld | so better to use qpid on fedora | 23:37 |
| asalkeld | and note the wlan0 - running on my home wireless | 23:38 |
| *** cloudvirt has quit IRC | 23:38 | |
| *** chrisfer has quit IRC | 23:39 | |
| *** zaitcev has quit IRC | 23:40 | |
| jgriffith | vishy: ping | 23:40 |
| *** gyee has quit IRC | 23:41 | |
| *** edygarcia has quit IRC | 23:41 | |
| *** mnewby has quit IRC | 23:42 | |
| *** edygarcia has joined #openstack-dev | 23:42 | |
| s34n | asalkeld: what does ENABLED_SERVICES do? | 23:42 |
| arosen | it tells ./stack.sh what services to install/start | 23:43 |
| arosen | http://devstack.org/stack.sh.html | 23:43 |
| *** mnewby has joined #openstack-dev | 23:47 | |
| *** jakedahn is now known as jakedahn_zz | 23:47 | |
| *** gongys has joined #openstack-dev | 23:48 | |
| *** lzyeval has joined #openstack-dev | 23:49 | |
| hemna | who adds the 'schedule' routing key binding to the nova exchange? | 23:54 |
| *** spiffxp has joined #openstack-dev | 23:54 | |
| *** johnpur has quit IRC | 23:54 | |
| *** steveb_ has joined #openstack-dev | 23:54 | |
| *** johnpur has joined #openstack-dev | 23:54 | |
| *** ChanServ sets mode: +v johnpur | 23:54 | |
| s34n | asalkeld: qpid can't load because port 5672 is being used by beam.smp | 23:55 |
| asalkeld | what's that? | 23:55 |
| asalkeld | beam.smp? | 23:55 |
| *** sacharya has joined #openstack-dev | 23:55 | |
| s34n | asalkeld: I don't know. Google tells me it might be couchdb | 23:55 |
| Ryan_Lane | is there a decent example anywhere on how to load the api extensions? | 23:56 |
| s34n | but I don't know how it got here | 23:56 |
| *** mikal has quit IRC | 23:56 | |
| *** mattstep has joined #openstack-dev | 23:56 | |
| asalkeld | http://ports.my-addr.com/tcp_port-udp_port-application-and-description.php?port=5672 | 23:57 |
| asalkeld | seems like amqp should be on 5671 | 23:57 |
| *** mikal has joined #openstack-dev | 23:58 | |
| *** johnpur has quit IRC | 23:58 | |
| *** johnpur has joined #openstack-dev | 23:59 | |
| *** ChanServ sets mode: +v johnpur | 23:59 | |
| jkoelker | beam.smp is an erlang process, chances are rabbit is running | 23:59 |
| asalkeld | ok can't have rabbit and qpid | 23:59 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!