| *** alunch has quit IRC | 00:00 | |
| vishy | arosen: still around | 00:01 |
|---|---|---|
| vishy | ? | 00:01 |
| vishy | getting very close but something is missing | 00:01 |
| arosen | yup | 00:02 |
| *** alunch has joined #openstack-dev | 00:02 | |
| vishy | arosen: so i don't see the connection between br-int and br-guest (what i called my provider bridge) on my controller | 00:03 |
| vishy | it is there on the host | 00:03 |
| vishy | arosen: http://paste.openstack.org/show/38219/ | 00:03 |
| vishy | something is supposed to create the magic interface int-br-guest | 00:04 |
| vishy | the only thing i can figure is that i'm not running the agent on the controller | 00:04 |
| vishy | could that be it? | 00:04 |
| arosen | vishy: yup the agent does it | 00:04 |
| *** seanrob has joined #openstack-dev | 00:05 | |
| arosen | vishy: I guess you are using eth0.15 as a flat network interface? | 00:05 |
| arosen | vishy: if you do ovs-ofctl dump-flows br-int ; ovs-vsctl dump-flows br-guest ; you should be able to see the flows that and doing the vlan mapping and stuff. | 00:06 |
| vishy | yeah | 00:06 |
| vishy | arosen: i see the traffic just fine | 00:06 |
| *** nachi__ has joined #openstack-dev | 00:06 | |
| vishy | it makes it to the controller | 00:06 |
| vishy | but it doesn't hop from br-guest to br-int | 00:06 |
| *** seanrob_ has quit IRC | 00:07 | |
| vishy | and dnsmasq is listening on a tap device in br-int | 00:07 |
| *** seanrob has quit IRC | 00:07 | |
| *** seanrob has joined #openstack-dev | 00:07 | |
| arosen | after you started the agent on the controller node did it create the the phy-br-guest veths for you? | 00:08 |
| arosen | the dnsmasq process should show up as a port connected to br-int so that's fine. | 00:09 |
| *** Mandell has quit IRC | 00:09 | |
| arosen | On your controller node is the interface data0 connected to the same L2 as eth0.15 on the compute node? | 00:10 |
| *** gongysh has joined #openstack-dev | 00:10 | |
| vishy | arosen: i haven't tried it yet, getting the agent on the conroller is a little tricky | 00:11 |
| vishy | arosen: yeah | 00:11 |
| vishy | btw it let me add a bridge as a vswitch port | 00:11 |
| vishy | but i don't know if it was actually doing anything :) | 00:11 |
| *** jimfehlig has joined #openstack-dev | 00:11 | |
| *** seanrob_ has joined #openstack-dev | 00:11 | |
| arosen | hopefully it is :) | 00:12 |
| *** seanrob has quit IRC | 00:12 | |
| arosen | it should just configure the bridges and add flows then just be idle. | 00:12 |
| arosen | if you do ovs-ofctl dump-flows br-int ; you'll see what it did | 00:12 |
| *** markmcclain has quit IRC | 00:12 | |
| *** seanrob_ has quit IRC | 00:12 | |
| vishy | yes it does work | 00:13 |
| *** seanrob has joined #openstack-dev | 00:13 | |
| vishy | well that is pretty cool | 00:13 |
| *** cmark has joined #openstack-dev | 00:13 | |
| arosen | work as in you can now ping between two machines ? | 00:13 |
| vishy | it works fine to add a linux bridge directly as a ovs port | 00:13 |
| *** pcm__ has quit IRC | 00:15 | |
| vishy | arosen: i'm still trying to determine whether i need br-tun | 00:15 |
| arosen | vishy: you only need that if you want to do GRE tunnels | 00:16 |
| *** pmathews has quit IRC | 00:16 | |
| *** pmathews has joined #openstack-dev | 00:16 | |
| vishy | arosen: but i could do tunnels over br-guest for example right | 00:16 |
| *** pcm__ has joined #openstack-dev | 00:16 | |
| vishy | i could use tunnel_bridge = br-tun ? | 00:17 |
| *** seanrob has quit IRC | 00:17 | |
| arosen | vishy: The way that the openvswitch plugin works is it creates a special bridge for tunnels that they terminate on | 00:17 |
| *** seanrob_ has joined #openstack-dev | 00:17 | |
| vishy | ah ok | 00:17 |
| vishy | so do i manually have to add a physical interface to br-tun? | 00:17 |
| vishy | and if so can i add the same interface to br-tun and br-guest? | 00:17 |
| vishy | i.e. eth0.15 | 00:18 |
| arosen | I think you just need: enable_tunneling=True in the [OVS] section and it makes the bridge for you if you don't already have it. | 00:18 |
| *** tjones has joined #openstack-dev | 00:18 | |
| *** tjones has left #openstack-dev | 00:19 | |
| vishy | hmm | 00:19 |
| arosen | vishy: if you are using tunneling you then usually use the l3 agent where you tell it which external_network_bridge to upllink the traffic to | 00:19 |
| arosen | to get you to the outside world | 00:20 |
| vishy | sure | 00:20 |
| *** garyTh has quit IRC | 00:20 | |
| vishy | arosen: hmm | 00:20 |
| *** openstackgerrit has quit IRC | 00:20 | |
| *** cmark has quit IRC | 00:21 | |
| *** openstackgerrit has joined #openstack-dev | 00:21 | |
| *** bknudson has quit IRC | 00:21 | |
| *** cmark has joined #openstack-dev | 00:22 | |
| vishy | arosen: ah so it is going to use the ip address you specify and the hosts normal routing to send the traffic | 00:22 |
| vishy | so if that ip address happens to be on eth0.15 then it will go across the same interface | 00:22 |
| arosen | For tunneling when the l2-agent starts it sends the local_ip value (from the configuration file) to quantum-server and then quantum-server tells the agent about the local_ip values that he knows about. Then each agent builds up a tunnel to each local_ip; yup it will just use the hosts routing | 00:25 |
| arosen | (the local_ip values from the other hypervisors) | 00:25 |
| *** melwitt has quit IRC | 00:25 | |
| arosen | i'm not 100% sure what you mean by , if that ip address happens to be on eth0.15 then it will go across the same interface | 00:25 |
| arosen | which ip address. The management ? | 00:26 |
| *** seanrob has joined #openstack-dev | 00:26 | |
| *** seanrob_ has quit IRC | 00:27 | |
| arosen | Usually you'd create a separate static route in the hypervisors to route the tunnel traffic over a different interface if you have multiple (as it just takes the default routing table of the hv). | 00:28 |
| *** topol has joined #openstack-dev | 00:28 | |
| *** pmathews has quit IRC | 00:32 | |
| *** sthaha has joined #openstack-dev | 00:33 | |
| *** sthaha has joined #openstack-dev | 00:33 | |
| *** bdpayne has joined #openstack-dev | 00:34 | |
| *** cmark_ has joined #openstack-dev | 00:34 | |
| *** yjiang5 has quit IRC | 00:35 | |
| *** salv-orlando has quit IRC | 00:35 | |
| *** cmark has quit IRC | 00:38 | |
| *** zzs has left #openstack-dev | 00:40 | |
| *** cmark_ has quit IRC | 00:40 | |
| *** cmark has joined #openstack-dev | 00:40 | |
| *** cdub has quit IRC | 00:45 | |
| *** cmark_ has joined #openstack-dev | 00:46 | |
| *** cmark has quit IRC | 00:48 | |
| *** sushils has quit IRC | 00:50 | |
| *** cmark has joined #openstack-dev | 00:50 | |
| *** cmark has quit IRC | 00:52 | |
| *** cmark has joined #openstack-dev | 00:52 | |
| *** seanrob_ has joined #openstack-dev | 00:53 | |
| *** cmark_ has quit IRC | 00:53 | |
| openstackgerrit | A change was merged to openstack-dev/devstack: Run heat tempest tests if heat is enabled in devstack. https://review.openstack.org/31590 | 00:54 |
| *** vipul is now known as vipul|away | 00:54 | |
| *** vipul|away is now known as vipul | 00:54 | |
| *** bdpayne has quit IRC | 00:55 | |
| *** nachi__ has quit IRC | 00:56 | |
| *** seanrob has quit IRC | 00:57 | |
| *** timello has quit IRC | 00:57 | |
| *** seanrob_ has quit IRC | 00:58 | |
| *** michchap has quit IRC | 00:58 | |
| *** michchap has joined #openstack-dev | 00:58 | |
| *** timello has joined #openstack-dev | 00:58 | |
| *** pcm_ has joined #openstack-dev | 01:00 | |
| *** pcm_ has quit IRC | 01:00 | |
| *** PaulM has joined #openstack-dev | 01:01 | |
| *** pcm_ has joined #openstack-dev | 01:01 | |
| *** giulivo has quit IRC | 01:02 | |
| openstackgerrit | A change was merged to openstack-dev/devstack: Set variables in rsyslog.conf to disable rate-limiting. https://review.openstack.org/30783 | 01:02 |
| openstackgerrit | A change was merged to openstack-dev/devstack: Add missing .debs required by Debian https://review.openstack.org/30546 | 01:02 |
| *** pcm__ has quit IRC | 01:02 | |
| *** lbragstad has joined #openstack-dev | 01:04 | |
| *** novas0x2a|laptop has quit IRC | 01:08 | |
| *** terry7 has quit IRC | 01:12 | |
| *** kaushikc has joined #openstack-dev | 01:12 | |
| *** litong has quit IRC | 01:13 | |
| *** seanrob has joined #openstack-dev | 01:23 | |
| *** rkukura has joined #openstack-dev | 01:27 | |
| *** Tross has joined #openstack-dev | 01:28 | |
| *** seanrob has quit IRC | 01:28 | |
| *** jog0 is now known as jog0-away | 01:29 | |
| *** esheffield has quit IRC | 01:29 | |
| *** otherwiseguy has joined #openstack-dev | 01:32 | |
| *** kaushikc has quit IRC | 01:34 | |
| *** alop has quit IRC | 01:37 | |
| *** kaushikc has joined #openstack-dev | 01:44 | |
| *** kaushikc has quit IRC | 01:44 | |
| *** erkules has quit IRC | 01:47 | |
| *** portante has joined #openstack-dev | 01:47 | |
| *** erkules_ has joined #openstack-dev | 01:47 | |
| *** seanrob has joined #openstack-dev | 01:48 | |
| *** gyee has quit IRC | 01:50 | |
| openstackgerrit | A change was merged to openstack/nova: Keypair API test cleanup https://review.openstack.org/31841 | 01:51 |
| *** zodiak has quit IRC | 01:57 | |
| *** nati_ueno has quit IRC | 01:59 | |
| *** galstrom_zzz is now known as galstrom | 02:04 | |
| *** yaguang has joined #openstack-dev | 02:04 | |
| yamahata | gongysh, pong | 02:05 |
| gongysh | yamahata: hi | 02:05 |
| yamahata | hi | 02:05 |
| gongysh | have u compared the floodlight and ryu? | 02:06 |
| yamahata | Do you mean controller side or plugin side? | 02:07 |
| *** zodiak has joined #openstack-dev | 02:07 | |
| yamahata | I had a very rough look at floodlight code. Not detail. | 02:07 |
| yamahata | The architecture of it looked quite different from Ryu. | 02:07 |
| *** souvik has quit IRC | 02:08 | |
| *** CaptTofu has quit IRC | 02:08 | |
| gongysh | one is java, one is python. | 02:08 |
| gongysh | ryu is adopting ovsdb and openflow | 02:08 |
| *** CaptTofu has joined #openstack-dev | 02:08 | |
| *** bing_bu has joined #openstack-dev | 02:09 | |
| yamahata | Ah yes. Ryu is trying to adopt many protocols if it's useful. | 02:09 |
| gongysh | do u have a plan to support the 'openflow cofig'? | 02:11 |
| yamahata | Ryu will provide RPC support for other language. It's now under development. | 02:11 |
| yamahata | Ryu already supports OF-config. :-) | 02:11 |
| gongysh | ok, good | 02:12 |
| gongysh | does RYU have a topology UI just like floodlight? | 02:13 |
| yamahata | Yes. | 02:13 |
| *** mnaser has quit IRC | 02:13 | |
| gongysh | how to access it? | 02:13 |
| yamahata | I think the slide for ONS2013 is helpful. | 02:13 |
| yamahata | http://osrg.github.io/ryu/slides/ONS2013-april-ryu-intro.pdf | 02:13 |
| yamahata | The patch is available on the ML. And the branch is available on github, I think. | 02:15 |
| gongysh | http://ryu.readthedocs.org/en/latest/ | 02:15 |
| gongysh | the doc is too simple | 02:15 |
| yamahata | The page 10 of the slide shows how the GUI looks like. | 02:15 |
| yamahata | github wiki is more detailed. | 02:16 |
| yamahata | https://github.com/osrg/ryu/wiki | 02:16 |
| yamahata | Right now, we've just started moving to readthedocs. | 02:17 |
| yamahata | For now, the wiki page is better. | 02:17 |
| *** nati_ueno has joined #openstack-dev | 02:17 | |
| *** zb has joined #openstack-dev | 02:19 | |
| gongysh | is the GUI code in trunk? | 02:20 |
| *** davidkranz has joined #openstack-dev | 02:20 | |
| yamahata | Not yet. I think it will be merged very soon. | 02:20 |
| *** galstrom is now known as galstrom_zzz | 02:20 | |
| *** zbitter has joined #openstack-dev | 02:21 | |
| *** zaneb has quit IRC | 02:21 | |
| *** mnaser has joined #openstack-dev | 02:22 | |
| *** zb has quit IRC | 02:23 | |
| *** davidkranz has quit IRC | 02:24 | |
| *** zbitter has quit IRC | 02:25 | |
| *** PaulM has left #openstack-dev | 02:28 | |
| *** darjeeling has joined #openstack-dev | 02:30 | |
| *** anfrolov has joined #openstack-dev | 02:30 | |
| *** cp16net is now known as cp16net|away | 02:34 | |
| *** cp16net|away is now known as cp16net | 02:34 | |
| *** chandankumar has joined #openstack-dev | 02:35 | |
| *** chandankumar has joined #openstack-dev | 02:36 | |
| jamielennox | ayoung: still awake? | 02:37 |
| *** _ffio_ has joined #openstack-dev | 02:37 | |
| *** ffio_ has quit IRC | 02:37 | |
| openstackgerrit | A change was merged to openstack/cinder: Add missing tests for iscsi_* methods https://review.openstack.org/31640 | 02:37 |
| *** ben_duyujie has joined #openstack-dev | 02:38 | |
| adam_g | dhellmann-away, guess i dont have access to +2 this in, but should pass now. https://review.openstack.org/#/c/32038/ | 02:39 |
| *** nati_ueno has quit IRC | 02:40 | |
| *** yjiang5 has joined #openstack-dev | 02:42 | |
| *** ladquin is now known as ladquin_away | 02:43 | |
| *** jbresnah has quit IRC | 02:44 | |
| *** boris-42 has joined #openstack-dev | 02:46 | |
| *** ndipanov has quit IRC | 02:48 | |
| *** ben_duyujie has quit IRC | 02:50 | |
| *** darjeeling has quit IRC | 02:59 | |
| *** sungju has joined #openstack-dev | 03:00 | |
| *** terriyu has quit IRC | 03:01 | |
| *** neelashah has joined #openstack-dev | 03:04 | |
| *** kaushikc has joined #openstack-dev | 03:04 | |
| *** neelashah1 has joined #openstack-dev | 03:08 | |
| *** neelashah has quit IRC | 03:08 | |
| *** rcrit has quit IRC | 03:09 | |
| *** enikanorov has joined #openstack-dev | 03:09 | |
| *** zhiyan has joined #openstack-dev | 03:14 | |
| *** otherwiseguy has quit IRC | 03:16 | |
| *** pixelbeat has quit IRC | 03:18 | |
| *** rcrit has joined #openstack-dev | 03:21 | |
| *** Ryan_Lane has quit IRC | 03:22 | |
| *** sungju has quit IRC | 03:29 | |
| *** marun has quit IRC | 03:30 | |
| *** michchap has quit IRC | 03:34 | |
| *** michchap has joined #openstack-dev | 03:34 | |
| *** edewata has quit IRC | 03:36 | |
| *** yjiang5 has quit IRC | 03:38 | |
| *** edewata has joined #openstack-dev | 03:38 | |
| *** michchap has quit IRC | 03:39 | |
| *** michchap_ has joined #openstack-dev | 03:39 | |
| *** bdpayne has joined #openstack-dev | 03:40 | |
| *** afazekas has joined #openstack-dev | 03:45 | |
| *** chandankumar has quit IRC | 03:45 | |
| *** seanrob has quit IRC | 03:48 | |
| *** seanrob has joined #openstack-dev | 03:49 | |
| *** seanrob has quit IRC | 03:52 | |
| *** seanrob has joined #openstack-dev | 03:52 | |
| *** seanrob has quit IRC | 03:54 | |
| *** souvik has joined #openstack-dev | 03:54 | |
| *** seanrob has joined #openstack-dev | 03:54 | |
| *** abhishekkr has quit IRC | 03:57 | |
| *** nati_ueno has joined #openstack-dev | 03:59 | |
| *** seanrob has quit IRC | 03:59 | |
| *** SergeyLukjanov has joined #openstack-dev | 04:00 | |
| *** kaushikc has quit IRC | 04:01 | |
| *** pmathews has joined #openstack-dev | 04:02 | |
| *** nati_ueno has quit IRC | 04:02 | |
| *** nunosantos has quit IRC | 04:03 | |
| *** kaushikc has joined #openstack-dev | 04:03 | |
| *** dims has quit IRC | 04:08 | |
| *** souvik has quit IRC | 04:10 | |
| *** bdpayne has quit IRC | 04:11 | |
| openstackgerrit | A change was merged to openstack/keystone: Require keystone-user/-group for pki_setup https://review.openstack.org/29644 | 04:11 |
| openstackgerrit | A change was merged to openstack/keystone: Import eventlet patch from oslo. https://review.openstack.org/30999 | 04:11 |
| openstackgerrit | A change was merged to openstack/cinder: Unset all stubs before running other cleanups. https://review.openstack.org/30720 | 04:11 |
| openstackgerrit | A change was merged to openstack-infra/devstack-gate: Add ZeroMQ support https://review.openstack.org/31686 | 04:11 |
| *** jjmb has joined #openstack-dev | 04:14 | |
| *** yjiang5 has joined #openstack-dev | 04:14 | |
| *** chandankumar has joined #openstack-dev | 04:16 | |
| *** neelashah1 has quit IRC | 04:18 | |
| *** anfrolov has quit IRC | 04:21 | |
| *** zaneb has joined #openstack-dev | 04:23 | |
| *** seanrob has joined #openstack-dev | 04:25 | |
| *** abhishekkr has joined #openstack-dev | 04:27 | |
| *** seanrob has quit IRC | 04:30 | |
| *** Aarti has joined #openstack-dev | 04:31 | |
| *** anfrolov has joined #openstack-dev | 04:44 | |
| *** chandankumar has quit IRC | 04:52 | |
| *** anfrolov has quit IRC | 04:53 | |
| *** bdpayne has joined #openstack-dev | 04:56 | |
| *** enikanorov has quit IRC | 04:56 | |
| *** jimfehlig has quit IRC | 04:57 | |
| *** pmathews has quit IRC | 05:00 | |
| openstackgerrit | A change was merged to openstack/nova: Allocate networks in the background https://review.openstack.org/29288 | 05:05 |
| *** SergeyLukjanov has quit IRC | 05:08 | |
| *** nati_ueno has joined #openstack-dev | 05:09 | |
| *** anfrolov has joined #openstack-dev | 05:09 | |
| *** SergeyLukjanov has joined #openstack-dev | 05:10 | |
| *** nati_ueno has quit IRC | 05:11 | |
| *** anfrolov has quit IRC | 05:11 | |
| *** SergeyLukjanov has quit IRC | 05:12 | |
| *** Mandell has joined #openstack-dev | 05:12 | |
| *** susanne-balle has joined #openstack-dev | 05:13 | |
| *** xga has joined #openstack-dev | 05:13 | |
| *** lastidiot has quit IRC | 05:17 | |
| *** susanne-balle has quit IRC | 05:17 | |
| *** comay has quit IRC | 05:17 | |
| *** amerine has quit IRC | 05:19 | |
| *** pmathews has joined #openstack-dev | 05:22 | |
| *** pmathews has quit IRC | 05:22 | |
| *** bdpayne has quit IRC | 05:22 | |
| *** amerine has joined #openstack-dev | 05:23 | |
| *** asalkeld is now known as asalkeld_sleepin | 05:25 | |
| *** asalkeld_sleepin is now known as asalkeld_headcol | 05:25 | |
| *** seanrob has joined #openstack-dev | 05:25 | |
| *** asalkeld_headcol is now known as asalkeld_afk | 05:26 | |
| *** pmathews has joined #openstack-dev | 05:26 | |
| *** jbresnah has joined #openstack-dev | 05:26 | |
| *** navid_ has joined #openstack-dev | 05:27 | |
| *** nati_ueno has joined #openstack-dev | 05:29 | |
| *** kaushikc has left #openstack-dev | 05:29 | |
| *** seanrob has quit IRC | 05:30 | |
| *** gongysh has quit IRC | 05:32 | |
| *** markwash has quit IRC | 05:38 | |
| *** saurabh_ has joined #openstack-dev | 05:40 | |
| saurabh_ | can anyone tell me how the cpu_util for instance calculation is done on ceilometer? | 05:41 |
| *** xga has quit IRC | 05:45 | |
| *** jcoufal has joined #openstack-dev | 05:47 | |
| *** pcm____ has joined #openstack-dev | 05:48 | |
| *** pcm_ has quit IRC | 05:51 | |
| *** freedomhui has joined #openstack-dev | 05:52 | |
| *** _ffio_ has quit IRC | 05:53 | |
| *** darjeeling has joined #openstack-dev | 05:55 | |
| *** _ffio_ has joined #openstack-dev | 05:58 | |
| openstackgerrit | A change was merged to openstack/horizon: Remove "Edit VIP" button when there is no VIP https://review.openstack.org/31967 | 05:59 |
| *** darjeeli_ has joined #openstack-dev | 06:00 | |
| *** nati_ueno has quit IRC | 06:00 | |
| *** nati_ueno has joined #openstack-dev | 06:01 | |
| *** vsergeyev has quit IRC | 06:02 | |
| *** darjeeling has quit IRC | 06:04 | |
| *** jprovazn has joined #openstack-dev | 06:06 | |
| *** pcm____ has quit IRC | 06:08 | |
| *** vartom12 has joined #openstack-dev | 06:08 | |
| *** vartom13 has joined #openstack-dev | 06:10 | |
| *** flaper87 has joined #openstack-dev | 06:10 | |
| *** Yuan_ has joined #openstack-dev | 06:10 | |
| *** pmathews has quit IRC | 06:11 | |
| *** vartom12 has quit IRC | 06:13 | |
| openstackgerrit | A change was merged to openstack/oslo-incubator: Add fileutils maintainer https://review.openstack.org/32077 | 06:15 |
| *** freedomhui has quit IRC | 06:18 | |
| *** nati_ueno has quit IRC | 06:21 | |
| *** kaushikc has joined #openstack-dev | 06:22 | |
| *** seanrob has joined #openstack-dev | 06:26 | |
| *** anfrolov has joined #openstack-dev | 06:27 | |
| *** seanrob has quit IRC | 06:31 | |
| *** sthaha has quit IRC | 06:36 | |
| *** rcrit has quit IRC | 06:36 | |
| *** yolanda has joined #openstack-dev | 06:38 | |
| *** romcheg has joined #openstack-dev | 06:38 | |
| *** romcheg has left #openstack-dev | 06:39 | |
| *** enikanorov has joined #openstack-dev | 06:40 | |
| cmark | Can anyone tell me when/how the SQL database tables are created in keystone? | 06:43 |
| *** freedomhui has joined #openstack-dev | 06:46 | |
| *** jtomasek has joined #openstack-dev | 06:47 | |
| *** topol has quit IRC | 06:48 | |
| *** chandankumar has joined #openstack-dev | 06:52 | |
| *** nati_ueno has joined #openstack-dev | 06:56 | |
| *** reidrac has joined #openstack-dev | 07:00 | |
| *** ifarkas has joined #openstack-dev | 07:03 | |
| *** eglynn has joined #openstack-dev | 07:13 | |
| *** darjeeli_ has quit IRC | 07:14 | |
| *** afazekas_ has joined #openstack-dev | 07:14 | |
| *** psedlak has joined #openstack-dev | 07:17 | |
| *** egallen has joined #openstack-dev | 07:20 | |
| *** SergeyLukjanov has joined #openstack-dev | 07:21 | |
| *** jgallard has joined #openstack-dev | 07:25 | |
| *** seanrob has joined #openstack-dev | 07:26 | |
| *** mrunge has joined #openstack-dev | 07:29 | |
| *** seanrob has quit IRC | 07:31 | |
| *** JordanP has joined #openstack-dev | 07:32 | |
| *** aloga has quit IRC | 07:34 | |
| *** sandeepr has quit IRC | 07:34 | |
| *** aloga has joined #openstack-dev | 07:34 | |
| *** bogdando has quit IRC | 07:37 | |
| *** bogdando has joined #openstack-dev | 07:38 | |
| *** AnilV4 has joined #openstack-dev | 07:40 | |
| *** shardy_afk is now known as shardy | 07:41 | |
| *** ndipanov has joined #openstack-dev | 07:43 | |
| *** nati_ueno has quit IRC | 07:43 | |
| *** sungju has joined #openstack-dev | 07:44 | |
| *** michchap_ has quit IRC | 07:47 | |
| *** michchap has joined #openstack-dev | 07:47 | |
| *** jamielennox is now known as jamielennox|away | 07:49 | |
| *** fbo has joined #openstack-dev | 07:59 | |
| *** darjeeling has joined #openstack-dev | 07:59 | |
| *** giulivo has joined #openstack-dev | 08:01 | |
| *** rnirmal has joined #openstack-dev | 08:03 | |
| *** mmagr has joined #openstack-dev | 08:05 | |
| *** johnthetubaguy has joined #openstack-dev | 08:07 | |
| *** lucasagomes has joined #openstack-dev | 08:07 | |
| *** derekh has joined #openstack-dev | 08:10 | |
| *** kashyap` is now known as kashyap | 08:11 | |
| *** dcasier_ has joined #openstack-dev | 08:13 | |
| openstackgerrit | A change was merged to openstack/horizon: Fix bug in port creation form https://review.openstack.org/31562 | 08:15 |
| *** iartarisi has joined #openstack-dev | 08:16 | |
| *** zhiyan has quit IRC | 08:16 | |
| *** freedomhui has quit IRC | 08:18 | |
| *** jpich has joined #openstack-dev | 08:19 | |
| *** pschaef has joined #openstack-dev | 08:22 | |
| *** darjeeling has quit IRC | 08:23 | |
| *** seanrob has joined #openstack-dev | 08:27 | |
| *** jbresnah has quit IRC | 08:27 | |
| *** sushils has joined #openstack-dev | 08:29 | |
| *** mrunge_ has joined #openstack-dev | 08:29 | |
| *** mrunge has quit IRC | 08:29 | |
| *** kaushikc has quit IRC | 08:30 | |
| *** kaushikc has joined #openstack-dev | 08:31 | |
| *** pixelbeat has joined #openstack-dev | 08:31 | |
| *** seanrob has quit IRC | 08:32 | |
| *** mrunge_ has quit IRC | 08:33 | |
| *** mrunge_ has joined #openstack-dev | 08:33 | |
| *** mrunge_ is now known as mrunge | 08:33 | |
| *** johnthetubaguy1 has joined #openstack-dev | 08:39 | |
| *** johnthetubaguy has quit IRC | 08:40 | |
| *** garyk has joined #openstack-dev | 08:43 | |
| openstackgerrit | A change was merged to openstack/glance: Prevent E500 when listing with null values https://review.openstack.org/30859 | 08:44 |
| *** chandankumar has quit IRC | 08:45 | |
| *** erkules_ is now known as erkules | 08:49 | |
| *** terriyu has joined #openstack-dev | 08:50 | |
| *** yjiang5 has quit IRC | 08:54 | |
| *** anfrolov has quit IRC | 08:58 | |
| *** anfrolov has joined #openstack-dev | 08:58 | |
| *** terriyu has quit IRC | 09:00 | |
| *** danpb has joined #openstack-dev | 09:01 | |
| *** egallen has quit IRC | 09:02 | |
| *** sungju has quit IRC | 09:03 | |
| *** egallen has joined #openstack-dev | 09:04 | |
| *** bing_bu has quit IRC | 09:09 | |
| *** andreaf has quit IRC | 09:10 | |
| *** xBsd has joined #openstack-dev | 09:10 | |
| *** athomas has joined #openstack-dev | 09:14 | |
| *** mjfork has quit IRC | 09:17 | |
| *** jruzicka has joined #openstack-dev | 09:18 | |
| dcasier_ | Hi | 09:24 |
| *** navid_ has quit IRC | 09:25 | |
| dcasier_ | Is there any work to allow to rename directory in swift? | 09:26 |
| *** navid_ has joined #openstack-dev | 09:27 | |
| *** seanrob has joined #openstack-dev | 09:27 | |
| *** seanrob has quit IRC | 09:32 | |
| openstackgerrit | A change was merged to openstack/horizon: switching to use limits instead of quotas https://review.openstack.org/29427 | 09:38 |
| *** zhiyan has joined #openstack-dev | 09:39 | |
| *** garyk has quit IRC | 09:49 | |
| *** dripton has quit IRC | 09:51 | |
| *** dstanek has joined #openstack-dev | 09:51 | |
| *** cmark has quit IRC | 09:51 | |
| *** cmark has joined #openstack-dev | 09:52 | |
| *** openstack has quit IRC | 09:52 | |
| *** openstack has joined #openstack-dev | 09:52 | |
| *** zhiyan1 has joined #openstack-dev | 09:53 | |
| *** dripton has joined #openstack-dev | 09:53 | |
| *** zhiyan has quit IRC | 09:53 | |
| *** PhongLy has quit IRC | 10:00 | |
| *** Ruetobas has quit IRC | 10:01 | |
| *** Ruetobas has joined #openstack-dev | 10:01 | |
| *** sushils has quit IRC | 10:02 | |
| *** jgallard has quit IRC | 10:02 | |
| *** jgallard has joined #openstack-dev | 10:03 | |
| *** sushils has joined #openstack-dev | 10:03 | |
| *** Alexei_987 has joined #openstack-dev | 10:03 | |
| *** sdake_ has quit IRC | 10:04 | |
| *** zaneb has quit IRC | 10:05 | |
| *** zaneb has joined #openstack-dev | 10:06 | |
| *** sushils has quit IRC | 10:07 | |
| *** navid_ has quit IRC | 10:08 | |
| *** navid_ has joined #openstack-dev | 10:08 | |
| *** enikanorov has quit IRC | 10:09 | |
| *** enikanorov has joined #openstack-dev | 10:10 | |
| *** zb has joined #openstack-dev | 10:10 | |
| *** sushils has joined #openstack-dev | 10:11 | |
| *** _ffio_ has quit IRC | 10:13 | |
| *** zaneb has quit IRC | 10:13 | |
| *** athomas has quit IRC | 10:15 | |
| *** pcm_ has joined #openstack-dev | 10:16 | |
| *** pcm_ has quit IRC | 10:16 | |
| *** pcm_ has joined #openstack-dev | 10:17 | |
| *** salv-orlando has joined #openstack-dev | 10:17 | |
| *** athomas has joined #openstack-dev | 10:25 | |
| *** xBsd has quit IRC | 10:27 | |
| *** seanrob has joined #openstack-dev | 10:28 | |
| *** romcheg has joined #openstack-dev | 10:30 | |
| *** abhishekkr has quit IRC | 10:32 | |
| *** seanrob has quit IRC | 10:33 | |
| *** zbitter has joined #openstack-dev | 10:34 | |
| *** zb has quit IRC | 10:37 | |
| *** zbitter has quit IRC | 10:39 | |
| *** zb has joined #openstack-dev | 10:39 | |
| *** markmc has joined #openstack-dev | 10:40 | |
| *** johnthetubaguy1 has quit IRC | 10:41 | |
| openstackgerrit | A change was merged to openstack/glance: Patch changes Fedora 16 to 18 on install page. https://review.openstack.org/32072 | 10:41 |
| *** johnthetubaguy has joined #openstack-dev | 10:42 | |
| *** zbitter has joined #openstack-dev | 10:44 | |
| *** zb has quit IRC | 10:46 | |
| *** zbitter has quit IRC | 10:48 | |
| *** jgallard has quit IRC | 10:48 | |
| *** cmark_ has joined #openstack-dev | 10:52 | |
| *** fbo has quit IRC | 10:52 | |
| *** freedomhui has quit IRC | 10:53 | |
| *** martyntaylor has joined #openstack-dev | 10:54 | |
| *** cmark has quit IRC | 10:55 | |
| *** iartarisi has quit IRC | 10:58 | |
| *** SergeyLukjanov has quit IRC | 11:02 | |
| *** blues-man has joined #openstack-dev | 11:04 | |
| *** dims has joined #openstack-dev | 11:04 | |
| *** fbo has joined #openstack-dev | 11:04 | |
| *** zaneb has joined #openstack-dev | 11:04 | |
| *** mihgen has joined #openstack-dev | 11:06 | |
| *** zb has joined #openstack-dev | 11:07 | |
| *** Yuan_ has quit IRC | 11:08 | |
| *** SergeyLukjanov has joined #openstack-dev | 11:08 | |
| *** sride has joined #openstack-dev | 11:08 | |
| *** zbitter has joined #openstack-dev | 11:09 | |
| *** zaneb has quit IRC | 11:10 | |
| *** topol has joined #openstack-dev | 11:12 | |
| *** zaneb has joined #openstack-dev | 11:12 | |
| *** zb has quit IRC | 11:12 | |
| *** eglynn is now known as eglynn-afk | 11:13 | |
| *** zbitter has quit IRC | 11:14 | |
| *** dstanek has quit IRC | 11:17 | |
| *** johnthetubaguy1 has joined #openstack-dev | 11:19 | |
| *** johnthetubaguy has quit IRC | 11:19 | |
| *** zb has joined #openstack-dev | 11:19 | |
| *** kaushikc1 has joined #openstack-dev | 11:21 | |
| *** zaneb has quit IRC | 11:21 | |
| *** kaushikc has quit IRC | 11:22 | |
| *** litong has joined #openstack-dev | 11:26 | |
| *** zeriouz has joined #openstack-dev | 11:28 | |
| *** asalkeld_afk is now known as asalkeld | 11:28 | |
| *** seanrob has joined #openstack-dev | 11:28 | |
| *** seanrob has quit IRC | 11:33 | |
| *** tmclaugh[work] has joined #openstack-dev | 11:34 | |
| *** phschwartz has joined #openstack-dev | 11:34 | |
| *** zul has joined #openstack-dev | 11:34 | |
| *** litong has quit IRC | 11:35 | |
| *** zbitter has joined #openstack-dev | 11:37 | |
| *** eharney has joined #openstack-dev | 11:38 | |
| *** zb has quit IRC | 11:38 | |
| *** zb has joined #openstack-dev | 11:39 | |
| *** zeriouz has quit IRC | 11:40 | |
| *** xBsd has joined #openstack-dev | 11:41 | |
| *** zbitter has quit IRC | 11:41 | |
| *** yaguang has quit IRC | 11:43 | |
| *** fbo has quit IRC | 11:43 | |
| *** dstanek has joined #openstack-dev | 11:47 | |
| *** sandywalsh has quit IRC | 11:48 | |
| *** salv-orlando has quit IRC | 11:48 | |
| *** vsergeyev has joined #openstack-dev | 11:50 | |
| *** zeriouz has joined #openstack-dev | 11:52 | |
| *** mjfork has joined #openstack-dev | 11:53 | |
| *** johnthetubaguy has joined #openstack-dev | 11:54 | |
| *** johnthetubaguy1 has quit IRC | 11:57 | |
| *** sandywalsh has joined #openstack-dev | 11:59 | |
| *** johnthetubaguy has quit IRC | 12:00 | |
| *** ndipanov has quit IRC | 12:00 | |
| *** zeriouz has quit IRC | 12:01 | |
| *** johnthetubaguy has joined #openstack-dev | 12:02 | |
| *** anfrolov has quit IRC | 12:05 | |
| *** dstanek has quit IRC | 12:06 | |
| *** dstanek has joined #openstack-dev | 12:08 | |
| *** pschaef has quit IRC | 12:09 | |
| *** anfrolov has joined #openstack-dev | 12:10 | |
| *** zb has quit IRC | 12:10 | |
| *** zb has joined #openstack-dev | 12:11 | |
| *** souvik has joined #openstack-dev | 12:12 | |
| *** sride has quit IRC | 12:12 | |
| *** jgallard has joined #openstack-dev | 12:13 | |
| *** dstanek has quit IRC | 12:13 | |
| *** jjmb has quit IRC | 12:14 | |
| *** ndipanov has joined #openstack-dev | 12:15 | |
| *** zb has quit IRC | 12:16 | |
| *** souvik has quit IRC | 12:16 | |
| *** mrunge has quit IRC | 12:17 | |
| *** markvoelker has joined #openstack-dev | 12:17 | |
| *** martine_ has joined #openstack-dev | 12:18 | |
| *** dims has quit IRC | 12:19 | |
| *** zb has joined #openstack-dev | 12:21 | |
| *** fbo has joined #openstack-dev | 12:21 | |
| *** salv-orlando has joined #openstack-dev | 12:22 | |
| *** galstrom_zzz is now known as galstrom | 12:23 | |
| *** aloga has quit IRC | 12:23 | |
| *** zhiyan1 has quit IRC | 12:24 | |
| *** mkollaro has joined #openstack-dev | 12:26 | |
| *** drewlander has joined #openstack-dev | 12:26 | |
| *** seanrob has joined #openstack-dev | 12:29 | |
| *** dims has joined #openstack-dev | 12:33 | |
| *** seanrob has quit IRC | 12:33 | |
| *** johnthetubaguy1 has joined #openstack-dev | 12:33 | |
| *** dprince has joined #openstack-dev | 12:34 | |
| *** zbitter has joined #openstack-dev | 12:34 | |
| *** tzumainn has joined #openstack-dev | 12:35 | |
| *** kaushikc1 has quit IRC | 12:35 | |
| *** kaushikc has joined #openstack-dev | 12:35 | |
| *** johnthetubaguy has quit IRC | 12:36 | |
| *** zb has quit IRC | 12:37 | |
| *** zbitter has quit IRC | 12:39 | |
| *** markvoelker has quit IRC | 12:41 | |
| *** xBsd has quit IRC | 12:42 | |
| *** freedomhui has joined #openstack-dev | 12:42 | |
| *** bknudson has joined #openstack-dev | 12:42 | |
| *** darjeeling has joined #openstack-dev | 12:42 | |
| *** darjeeling has quit IRC | 12:43 | |
| *** ewindisch has quit IRC | 12:45 | |
| *** bknudson has quit IRC | 12:47 | |
| *** Aarti has quit IRC | 12:47 | |
| *** ewindisch has joined #openstack-dev | 12:47 | |
| *** gordc-away is now known as gordc | 12:48 | |
| *** shang has joined #openstack-dev | 12:49 | |
| *** anniec has joined #openstack-dev | 12:54 | |
| *** anniec has left #openstack-dev | 12:54 | |
| *** markvoelker has joined #openstack-dev | 12:54 | |
| *** romcheg has left #openstack-dev | 12:54 | |
| *** jjmb has joined #openstack-dev | 12:56 | |
| *** rcrit has joined #openstack-dev | 12:56 | |
| *** anniec_ has joined #openstack-dev | 12:57 | |
| *** xBsd has joined #openstack-dev | 12:57 | |
| *** ewindisch has quit IRC | 12:58 | |
| *** zaneb has joined #openstack-dev | 12:59 | |
| *** afazekas_ has quit IRC | 13:00 | |
| *** ewindisch has joined #openstack-dev | 13:00 | |
| *** jjmb has quit IRC | 13:01 | |
| *** SergeyLukjanov has quit IRC | 13:01 | |
| *** bknudson has joined #openstack-dev | 13:02 | |
| *** galstrom is now known as galstrom_zzz | 13:02 | |
| *** eglynn-afk is now known as eglynn | 13:03 | |
| *** rharwood has joined #openstack-dev | 13:03 | |
| *** zaneb has quit IRC | 13:05 | |
| *** neelashah has joined #openstack-dev | 13:07 | |
| *** kbringard has joined #openstack-dev | 13:08 | |
| *** jecarey has joined #openstack-dev | 13:08 | |
| *** kaushikc has quit IRC | 13:09 | |
| *** mestery_ has joined #openstack-dev | 13:09 | |
| *** andrewbogott_afk is now known as andrewbogott | 13:09 | |
| *** alunch has quit IRC | 13:09 | |
| *** jjmb has joined #openstack-dev | 13:11 | |
| *** dhellmann-away is now known as dhellmann | 13:11 | |
| *** mestery has quit IRC | 13:12 | |
| *** networkstatic has quit IRC | 13:12 | |
| *** egallen has quit IRC | 13:12 | |
| *** davidkranz has joined #openstack-dev | 13:12 | |
| *** egallen has joined #openstack-dev | 13:13 | |
| *** jayg|g0n3 is now known as jayg | 13:13 | |
| *** westmaas is now known as westmau5 | 13:14 | |
| *** radez_g0n3 is now known as radez | 13:15 | |
| *** ndipanov has quit IRC | 13:17 | |
| *** souvik has joined #openstack-dev | 13:18 | |
| *** marun has joined #openstack-dev | 13:22 | |
| *** msmedved has joined #openstack-dev | 13:22 | |
| *** mihgen has quit IRC | 13:23 | |
| *** souvik has quit IRC | 13:24 | |
| *** mihgen has joined #openstack-dev | 13:24 | |
| *** alunduil has quit IRC | 13:26 | |
| *** mihgen has quit IRC | 13:26 | |
| *** mjfs has joined #openstack-dev | 13:27 | |
| *** zyluo has joined #openstack-dev | 13:28 | |
| *** adalbas has joined #openstack-dev | 13:29 | |
| *** vartom14 has joined #openstack-dev | 13:29 | |
| *** seanrob has joined #openstack-dev | 13:29 | |
| *** martine_ has quit IRC | 13:30 | |
| *** ndipanov has joined #openstack-dev | 13:30 | |
| *** lbragstad has quit IRC | 13:30 | |
| *** krtaylor has quit IRC | 13:31 | |
| *** vartom13 has quit IRC | 13:32 | |
| *** portante has quit IRC | 13:32 | |
| *** vartom14 has quit IRC | 13:34 | |
| *** seanrob has quit IRC | 13:34 | |
| *** souvik has joined #openstack-dev | 13:36 | |
| *** ben_duyujie has joined #openstack-dev | 13:38 | |
| *** souvik has quit IRC | 13:40 | |
| *** dosaboy has joined #openstack-dev | 13:41 | |
| *** mtreinish has joined #openstack-dev | 13:44 | |
| *** michchap has quit IRC | 13:44 | |
| *** rcleere has joined #openstack-dev | 13:45 | |
| *** zaneb has joined #openstack-dev | 13:48 | |
| *** jjmb has quit IRC | 13:51 | |
| *** alexxu has joined #openstack-dev | 13:51 | |
| *** ewindisch has quit IRC | 13:51 | |
| *** markmcclain has joined #openstack-dev | 13:51 | |
| *** anniec_ has quit IRC | 13:52 | |
| *** otherwiseguy has joined #openstack-dev | 13:52 | |
| *** PhongLy has joined #openstack-dev | 13:53 | |
| *** zhiyan has joined #openstack-dev | 13:55 | |
| *** anniec has joined #openstack-dev | 13:56 | |
| *** SergeyLukjanov has joined #openstack-dev | 13:56 | |
| *** redbeard2 has quit IRC | 13:56 | |
| *** cp16net is now known as cp16net|away | 13:57 | |
| *** markvoelker has quit IRC | 13:58 | |
| *** jimfehlig has joined #openstack-dev | 13:58 | |
| *** xBsd has quit IRC | 13:58 | |
| *** alexxu has quit IRC | 13:58 | |
| *** mihgen has joined #openstack-dev | 13:58 | |
| *** adalbas has quit IRC | 13:59 | |
| *** alunch has joined #openstack-dev | 14:00 | |
| *** topol has quit IRC | 14:00 | |
| *** xBsd has joined #openstack-dev | 14:01 | |
| *** lbragstad has joined #openstack-dev | 14:02 | |
| *** rcrit has quit IRC | 14:04 | |
| *** jgallard has quit IRC | 14:04 | |
| *** jgallard has joined #openstack-dev | 14:06 | |
| *** rcleere has quit IRC | 14:07 | |
| *** ewindisch has joined #openstack-dev | 14:08 | |
| *** HenryG has joined #openstack-dev | 14:11 | |
| *** adalbas has joined #openstack-dev | 14:11 | |
| *** marun has quit IRC | 14:12 | |
| *** danpb has quit IRC | 14:12 | |
| *** drewland_ has joined #openstack-dev | 14:12 | |
| *** martyntaylor has quit IRC | 14:13 | |
| *** marun has joined #openstack-dev | 14:13 | |
| *** drewland_ has quit IRC | 14:13 | |
| *** drewland_ has joined #openstack-dev | 14:13 | |
| *** susanne-balle has joined #openstack-dev | 14:14 | |
| *** martyntaylor has joined #openstack-dev | 14:14 | |
| *** michchap has joined #openstack-dev | 14:15 | |
| *** drewlander has quit IRC | 14:15 | |
| *** rcleere has joined #openstack-dev | 14:16 | |
| *** dolphm has joined #openstack-dev | 14:17 | |
| *** radez is now known as radez_g0n3 | 14:17 | |
| *** HenryG has quit IRC | 14:17 | |
| *** HenryG has joined #openstack-dev | 14:17 | |
| *** rcrit has joined #openstack-dev | 14:18 | |
| *** danpb has joined #openstack-dev | 14:18 | |
| *** radez_g0n3 is now known as radez | 14:18 | |
| *** markvoelker has joined #openstack-dev | 14:19 | |
| *** jpeeler has joined #openstack-dev | 14:19 | |
| *** jpeeler has quit IRC | 14:19 | |
| *** jpeeler has joined #openstack-dev | 14:19 | |
| *** andrewbogott is now known as andrewbogott_afk | 14:19 | |
| *** krtaylor has joined #openstack-dev | 14:19 | |
| *** rcleere has quit IRC | 14:20 | |
| *** rcleere has joined #openstack-dev | 14:20 | |
| *** dolphm has quit IRC | 14:20 | |
| *** dolphm has joined #openstack-dev | 14:20 | |
| *** danpb has quit IRC | 14:20 | |
| *** danpb has joined #openstack-dev | 14:20 | |
| *** markvoelker has quit IRC | 14:20 | |
| *** markvoelker has joined #openstack-dev | 14:20 | |
| *** portante has joined #openstack-dev | 14:22 | |
| *** michchap has quit IRC | 14:25 | |
| *** jvrbanac has joined #openstack-dev | 14:26 | |
| *** Yuan has joined #openstack-dev | 14:27 | |
| *** afazekas has quit IRC | 14:28 | |
| *** hartsocks has joined #openstack-dev | 14:30 | |
| *** seanrob has joined #openstack-dev | 14:30 | |
| *** lastidiot has joined #openstack-dev | 14:31 | |
| *** martyntaylor1 has joined #openstack-dev | 14:31 | |
| *** mestery_ is now known as mestery | 14:32 | |
| *** jjmb has joined #openstack-dev | 14:33 | |
| *** afazekas has joined #openstack-dev | 14:34 | |
| *** garyTh has joined #openstack-dev | 14:34 | |
| *** jjmb1 has joined #openstack-dev | 14:34 | |
| *** seanrob has quit IRC | 14:34 | |
| *** martyntaylor has quit IRC | 14:34 | |
| *** esheffield has joined #openstack-dev | 14:34 | |
| *** SergeyLukjanov has quit IRC | 14:35 | |
| *** Ruetobas has quit IRC | 14:36 | |
| *** eglynn has quit IRC | 14:36 | |
| *** eglynn has joined #openstack-dev | 14:36 | |
| *** galstrom_zzz is now known as galstrom | 14:37 | |
| *** jjmb has quit IRC | 14:37 | |
| *** Ruetobas has joined #openstack-dev | 14:38 | |
| *** SergeyLukjanov has joined #openstack-dev | 14:38 | |
| *** litong has joined #openstack-dev | 14:39 | |
| *** marun has quit IRC | 14:39 | |
| *** marun has joined #openstack-dev | 14:41 | |
| *** eglynn has quit IRC | 14:41 | |
| *** eglynn_ has joined #openstack-dev | 14:41 | |
| *** mihgen has quit IRC | 14:42 | |
| *** xBsd has quit IRC | 14:42 | |
| *** topol has joined #openstack-dev | 14:42 | |
| *** otherwiseguy has quit IRC | 14:43 | |
| *** darjeeling has joined #openstack-dev | 14:43 | |
| *** lastidiot has quit IRC | 14:44 | |
| *** ifarkas has quit IRC | 14:45 | |
| *** anniec_ has joined #openstack-dev | 14:46 | |
| *** anniec_ has quit IRC | 14:46 | |
| *** anniec has quit IRC | 14:48 | |
| *** Ryan_Lane has joined #openstack-dev | 14:51 | |
| *** michchap has joined #openstack-dev | 14:51 | |
| *** pabelanger_ has joined #openstack-dev | 14:53 | |
| *** zhiyan has quit IRC | 14:56 | |
| *** datsun180b has joined #openstack-dev | 14:56 | |
| *** darjeeling has quit IRC | 14:57 | |
| *** pabelanger_ has quit IRC | 14:58 | |
| *** pabelanger_ has joined #openstack-dev | 14:58 | |
| *** pabelanger has quit IRC | 14:58 | |
| *** pabelanger_ is now known as pabelanger | 14:58 | |
| *** pabelanger_ has joined #openstack-dev | 14:59 | |
| *** andrewbogott_afk is now known as andrewbogott | 14:59 | |
| *** zyluo has quit IRC | 15:01 | |
| *** anfrolov has quit IRC | 15:01 | |
| *** PhongLy has quit IRC | 15:01 | |
| *** michchap has quit IRC | 15:02 | |
| *** gordc has quit IRC | 15:03 | |
| *** jecarey has quit IRC | 15:05 | |
| *** zyluo has joined #openstack-dev | 15:05 | |
| zyluo | harlowja, ping | 15:05 |
| *** ppyy has joined #openstack-dev | 15:08 | |
| *** jecarey has joined #openstack-dev | 15:09 | |
| *** pabelanger has quit IRC | 15:09 | |
| *** pabelanger has joined #openstack-dev | 15:09 | |
| openstackgerrit | A change was merged to openstack/python-novaclient: Use Python 3.x compatible except: construct https://review.openstack.org/32140 | 15:10 |
| *** cmark_ has quit IRC | 15:10 | |
| *** michchap has joined #openstack-dev | 15:11 | |
| *** cmark has joined #openstack-dev | 15:11 | |
| *** yjiang5 has joined #openstack-dev | 15:11 | |
| *** reidrac has quit IRC | 15:12 | |
| *** michchap_ has joined #openstack-dev | 15:12 | |
| *** lastidiot has joined #openstack-dev | 15:13 | |
| *** Tedster has joined #openstack-dev | 15:16 | |
| *** michchap has quit IRC | 15:16 | |
| *** michchap_ has quit IRC | 15:16 | |
| *** gordc has joined #openstack-dev | 15:17 | |
| yjiang5 | boris-42: hi | 15:18 |
| *** galstrom is now known as galstrom_zzz | 15:18 | |
| boris-42 | yjiang5 hi | 15:18 |
| *** matiu has quit IRC | 15:18 | |
| *** michchap has joined #openstack-dev | 15:18 | |
| *** colinmcnamara has joined #openstack-dev | 15:18 | |
| yjiang5 | we are looking on patch https://review.openstack.org/#/c/27130/ , will you split the patch into smaller one? | 15:18 |
| boris-42 | I am working on splitting | 15:19 |
| boris-42 | actually I spliced it | 15:19 |
| boris-42 | spliced it | 15:19 |
| boris-42 | autocorrect sucks.. | 15:19 |
| boris-42 | So I will publish code at this weekend | 15:19 |
| *** michchap_ has joined #openstack-dev | 15:19 | |
| *** jpeeler has quit IRC | 15:19 | |
| boris-42 | I am not sure when exactly but it will be ready before Monday | 15:20 |
| yjiang5 | boris-42: great. We are considering to add the auto-discovery PCI devices support based on your patch, how do you think about it? | 15:20 |
| *** networkstatic has joined #openstack-dev | 15:20 | |
| boris-42 | yjiang5 I could say only one thing | 15:20 |
| boris-42 | it will be nice=) | 15:20 |
| *** rcrit has quit IRC | 15:20 | |
| boris-42 | I will add some extra fields in PciDevice table | 15:21 |
| *** michcha__ has joined #openstack-dev | 15:21 | |
| boris-42 | dev_type <- | 15:21 |
| *** jgallard has quit IRC | 15:21 | |
| boris-42 | So just wait for monday yjiang5 ^ | 15:21 |
| yjiang5 | boris-42: sure. | 15:21 |
| *** jgallard has joined #openstack-dev | 15:22 | |
| boris-42 | Ok nice=) | 15:22 |
| yjiang5 | boris-42: we are also considering to enhance the pci label to be alias, | 15:22 |
| *** mlavalle has joined #openstack-dev | 15:22 | |
| boris-42 | Yes I thought also that labels are bad name=) | 15:22 |
| yjiang5 | like https://blueprints.launchpad.net/nova/+spec/epa-for-pcie-devices | 15:22 |
| *** jjmb1 has quit IRC | 15:22 | |
| *** johnthetubaguy1 is now known as johnthetubaguy | 15:22 | |
| boris-42 | so I will rename label to alias yjiang5 | 15:22 |
| *** michchap has quit IRC | 15:23 | |
| yjiang5 | alias is more like a rule to define or still be a simple assignment like current label? | 15:23 |
| *** terriyu has joined #openstack-dev | 15:24 | |
| *** Yuan has quit IRC | 15:24 | |
| boris-42 | They will be used in the same way as lables | 15:24 |
| *** michchap_ has quit IRC | 15:24 | |
| boris-42 | for filtering and attaching to instance in db yjiang5 ^] | 15:24 |
| *** michcha__ has quit IRC | 15:25 | |
| boris-42 | I don't think that we should change anything in this logic yjiang5 ^ | 15:25 |
| yjiang5 | boris-42: we are also considering that alias can be defined as QuickAssist=capabilities: devices: (capability_type=pci & capability: (product_id=0x443 & vendor_id=0x8086),, like like https://wiki.openstack.org/wiki/Enhanced-platform-awareness-pcie , | 15:25 |
| boris-42 | I will be back in half of hour ok? | 15:26 |
| *** nati_ueno has joined #openstack-dev | 15:26 | |
| yjiang5 | sure | 15:26 |
| openstackgerrit | A change was merged to openstack/oslo-incubator: Use Python 3.x compatible except construct https://review.openstack.org/32135 | 15:26 |
| *** epim has quit IRC | 15:27 | |
| *** vsergeyev has quit IRC | 15:28 | |
| *** nati_ueno has quit IRC | 15:28 | |
| *** cmark has quit IRC | 15:28 | |
| *** krtaylor has quit IRC | 15:29 | |
| *** monst has joined #openstack-dev | 15:30 | |
| *** seanrob has joined #openstack-dev | 15:31 | |
| *** pasquier-s has quit IRC | 15:31 | |
| *** xBsd has joined #openstack-dev | 15:31 | |
| *** bswartz has quit IRC | 15:32 | |
| *** aelkikhia has joined #openstack-dev | 15:33 | |
| *** rcrit has joined #openstack-dev | 15:33 | |
| *** jprovazn is now known as jprovazn_afk | 15:33 | |
| *** michchap has joined #openstack-dev | 15:33 | |
| openstackgerrit | A change was merged to openstack/oslo-incubator: Changed processing unique constraint name. https://review.openstack.org/32148 | 15:33 |
| *** darjeeling has joined #openstack-dev | 15:34 | |
| *** michchap_ has joined #openstack-dev | 15:34 | |
| *** devoid has joined #openstack-dev | 15:34 | |
| *** alunduil has joined #openstack-dev | 15:34 | |
| *** devoid has left #openstack-dev | 15:35 | |
| *** seanrob has quit IRC | 15:35 | |
| *** jpeeler has joined #openstack-dev | 15:36 | |
| *** jpeeler has quit IRC | 15:36 | |
| *** jpeeler has joined #openstack-dev | 15:36 | |
| *** michcha__ has joined #openstack-dev | 15:37 | |
| *** michchap has quit IRC | 15:37 | |
| *** aelkikhia has left #openstack-dev | 15:37 | |
| *** michchap_ has quit IRC | 15:39 | |
| *** michchap has joined #openstack-dev | 15:39 | |
| *** aelkikhia has joined #openstack-dev | 15:40 | |
| *** ben_duyujie has quit IRC | 15:40 | |
| *** SergeyLukjanov has quit IRC | 15:40 | |
| *** mmagr has quit IRC | 15:40 | |
| *** ben_duyujie has joined #openstack-dev | 15:40 | |
| *** stevemar has joined #openstack-dev | 15:40 | |
| *** michchap_ has joined #openstack-dev | 15:41 | |
| stevemar | dtroyer: ping | 15:41 |
| *** michcha__ has quit IRC | 15:41 | |
| *** otherwiseguy has joined #openstack-dev | 15:41 | |
| *** mrodden has quit IRC | 15:42 | |
| *** krtaylor has joined #openstack-dev | 15:42 | |
| *** michch___ has joined #openstack-dev | 15:43 | |
| *** galstrom_zzz is now known as galstrom | 15:43 | |
| *** michchap has quit IRC | 15:44 | |
| *** michchap_ has quit IRC | 15:45 | |
| dtroyer_zz | stevemar: dtroyer is alive | 15:45 |
| *** anfrolov has joined #openstack-dev | 15:46 | |
| *** AnilV4 has quit IRC | 15:47 | |
| stevemar | dtroyer_zz: can you take a look @ https://review.openstack.org/#/c/27143/, my long comment | 15:47 |
| stevemar | dtroyer_zz: in openstackclient, we don't take into account any domain when trying to find a resource by name right? | 15:47 |
| *** xBytez has quit IRC | 15:48 | |
| boris-42 | yjiang5 ping | 15:48 |
| *** xBytez has joined #openstack-dev | 15:48 | |
| *** michch___ has quit IRC | 15:48 | |
| giulivo | does anyone know where is hosted the devstack site to propose small changes? | 15:49 |
| dtroyer_zz | stevemar: THB I don't know the v3 api well enough to know, that's one reason I haven't jumped in on that one. | 15:49 |
| dtroyer_zz | guitarzan: devstack is handled exactly like the other OpenStack projects, proposals via Gerrit | 15:49 |
| dtroyer_zz | guitarzan: ^^sorry, meant giulivo | 15:50 |
| guitarzan | g<tab> fail! giulivo | 15:50 |
| guitarzan | :) | 15:50 |
| dtroyer_zz | precisely! | 15:50 |
| giulivo | dtroyer_zz, hi | 15:50 |
| giulivo | yeah I can't find the website sources | 15:50 |
| stevemar | dtroyer_zz: i'm asking because i'm confused, i did something I thought should fail, and didn't :) | 15:50 |
| *** katylava has joined #openstack-dev | 15:51 | |
| stevemar | dtroyer_zz: i'm on the master branch of openstack client, and did the following: | 15:51 |
| stevemar | openstack role add role_2 --user admin --project newProject | 15:51 |
| stevemar | which i figured would fail | 15:51 |
| stevemar | because it shouldn't be able to find the project name | 15:52 |
| *** jgallard has quit IRC | 15:52 | |
| boris-42 | yjiang5 Could you explain me in more words about capabilities | 15:52 |
| *** Yuan has joined #openstack-dev | 15:53 | |
| boris-42 | yjiang5 I just don't understand how to integrate it with my current code, and more then I don't understand why I should do this=) | 15:53 |
| stevemar | dtroyer_zz: when I take a look at screen, it's looks like it trys to translate the names to IDs by doing a list | 15:53 |
| *** jjmb has joined #openstack-dev | 15:53 | |
| stevemar | dtroyer_zz: GET /v3/roles?name=role_2, GET /v3/users?name=admin, GET http://10.0.2.15:35357/v3/projects?name=newProject | 15:54 |
| stevemar | dtroyer_zz: and that renders the PUT url correctly | 15:54 |
| boris-42 | yjiang5 But if you explain me, and If I found it interesting and useful I will change my current solution | 15:54 |
| dtroyer_zz | giulivo: git clone https://github.com/openstack-dev/devstack.git and use git review to propose, which sends it to Gerrit per https://wiki.openstack.org/wiki/How_To_Contribute | 15:54 |
| giulivo | dtroyer_zz, thanks for helping, really | 15:54 |
| giulivo | I'm used to the gerrit workflow | 15:54 |
| *** bdpayne has joined #openstack-dev | 15:55 | |
| giulivo | but can't find the devstack website sources ; I was looking into the devstack sources but still can't find it | 15:55 |
| stevemar | dtroyer_zz: I know you changed up something in keystone client code to add find methods, i'm wondering if that is the reason it's working | 15:55 |
| *** monst has quit IRC | 15:55 | |
| *** Mandell has quit IRC | 15:56 | |
| *** jpeeler1 has joined #openstack-dev | 15:56 | |
| dtroyer_zz | stevemar: the v3 api includes better search but I don't think manager.find() does it correctly until https://review.openstack.org/26323 landed yesterday in keystoneclient…it did a brute-force list and searched the result. | 15:57 |
| *** jpeeler has quit IRC | 15:58 | |
| dtroyer_zz | giulivo: oh, you mean to edit devstack.org?? | 15:58 |
| openstackgerrit | A change was merged to openstack/tempest: Default to m1.micro for heat test flavor. https://review.openstack.org/30042 | 15:58 |
| *** ffio has joined #openstack-dev | 15:59 | |
| giulivo | dtroyer_zz, yeah the website | 16:00 |
| *** jprovazn_afk has quit IRC | 16:00 | |
| dtroyer_zz | stevemar: ah, yes, probably that is why | 16:01 |
| stevemar | dtroyer_zz: so i'm wondering if we even need to worry about domain scoping then... | 16:01 |
| *** mrodden has joined #openstack-dev | 16:01 | |
| dtroyer_zz | stevemar: that I can't answer…I don't know what makes sense WRT domains | 16:01 |
| stevemar | dtroyer_zz: should be not be depending on the find method for some reason? brute-force being a no-no? | 16:02 |
| *** networkstatic has quit IRC | 16:02 | |
| stevemar | dtroyer_zz: we need to bug dolphm then | 16:02 |
| *** jpeeler1 has quit IRC | 16:02 | |
| stevemar | dtroyer_zz, but it's lunch, and my stomach calls. | 16:02 |
| * dolphm reading back | 16:02 | |
| stevemar | noooo | 16:02 |
| * dolphm waits | 16:02 | |
| stevemar | dolphm: go ahead :) | 16:03 |
| stevemar | dolphm: you have a knack of resolving things quickly... no pressure | 16:03 |
| *** rcleere has quit IRC | 16:03 | |
| dtroyer_zz | giulivo: this goes back to the days of fire and bronze... git clone https://github.com/cloudbuilders/devstack and edit the gh-pages branch. a github pull request is what it takes to get stuff in…a ping to me to look for it will be helpful as I don't check there too often these days. | 16:03 |
| *** gyee has joined #openstack-dev | 16:03 | |
| yjiang5 | boris-42: didn't notice your message. I think the key thing is how the user will specify the label | 16:04 |
| giulivo | oh thanks a lot! | 16:04 |
| giulivo | :) | 16:04 |
| *** BLZbubba has quit IRC | 16:04 | |
| *** alunduil has quit IRC | 16:04 | |
| dolphm | stevemar: caught up... so, i would suggest having openstackclient assume the default domain if the user doesn't specify one | 16:04 |
| *** redbeard2 has joined #openstack-dev | 16:04 | |
| dolphm | stevemar: dtroyer_zz: so, the 'find a project' call would look like, GET http://10.0.2.15:35357/v3/projects?name=newProject&domain_id=default | 16:05 |
| dolphm | stevemar: dtroyer_zz: that way it's automatically scoped (and safely, for the base use case) and you're guaranteed exactly one project | 16:05 |
| dolphm | stevemar: dtroyer_zz: i think the find() method impl is fine though... it shouldn't make any such assumptions | 16:05 |
| dolphm | stevemar: dtroyer_zz: in fact, assume domain_id=default everywhere if the user doesn't specify one | 16:06 |
| *** gary_th has joined #openstack-dev | 16:07 | |
| yjiang5 | boris-42: Considering when user want to assign the device to VM, they of course don't care for the BDF, I think that's the reason of your label. but how can they know about the exact meaning of the label? | 16:07 |
| *** NobodyCam has quit IRC | 16:07 | |
| boris-42 | yjiang5 Administrator should make explanantation about what means each label | 16:07 |
| *** martyntaylor1 has quit IRC | 16:07 | |
| *** garyTh has quit IRC | 16:08 | |
| boris-42 | yjiang5 Or you would like to store these info in nova.db also?) | 16:08 |
| *** markvoelker has quit IRC | 16:08 | |
| boris-42 | yjiang5 Add one more extra table for aliases?) | 16:09 |
| yjiang5 | boris-42: How can admin explain that to user. Condiering what happens in real world, when we buy a NIC, we will state it's a Intel XXX card, right? | 16:09 |
| stevemar | dolphm: so you still want the user to be able to pass a specific domain? | 16:09 |
| *** NobodyCam has joined #openstack-dev | 16:09 | |
| *** pmathews has joined #openstack-dev | 16:09 | |
| stevemar | dolphm: cause you're thinking of the case where find return > 1 result | 16:09 |
| boris-42 | yjiang5 Yeah and what is problem to make label with such name?) | 16:09 |
| dolphm | stevemar: yes to the first question | 16:10 |
| *** PhongLy has joined #openstack-dev | 16:11 | |
| boris-42 | yjiang5 Btw are you agree that we should have new table for PCI devices?) and the mechanism to specify works through instance_type_extra_specs? | 16:11 |
| dolphm | stevemar: if you assume a domain when conducting the find() in a multi-domain deployment with conflicting user/project names, you guarantee yourself one result and a better ux | 16:11 |
| yjiang5 | boris-42: Yes, I agree the new table and also he mechanism of extra_specs. | 16:11 |
| dolphm | stevemar: in the real world, a lot of deployments will only have the default domain, and those users are saved an extra cli arg... or, it'll be a multidomain deployment likely without a 'default' domain, in which case, those users should be domain-aware and specify their domains appropriately | 16:12 |
| boris-42 | yjiang5 nice so the problem is only how to store lable details?) | 16:12 |
| dolphm | stevemar: an env[OS_DEFAULT_DOMAIN_ID] might even make sense for all CLI users wherever a domain_id is required, as *most* users should be working within the confines of a single domain, at least most of the time | 16:13 |
| yjiang5 | boris-42: The problem of label is how to specify the label. I.e. how to connect the input from user, i.e. the label, to the real device in the compute node. I agree that your current specification through config file can be a first step, and we can show your our patch to enhancement. | 16:14 |
| *** michchap has joined #openstack-dev | 16:14 | |
| stevemar | dolphm: yeah, i was thinking of adding that as an env variable... might be better | 16:14 |
| dolphm | stevemar: (where a user's single domain might have a domain_id of something other than 'default') | 16:14 |
| dolphm | stevemar: i'd be in favor... it's configurable in keystone.conf as a static variable, it might as well be just as configurable on the client side | 16:14 |
| boris-42 | yjiang5 the input from user or from administrator? | 16:14 |
| yjiang5 | boris-42: from the user | 16:14 |
| boris-42 | Hmmmm. user is able just to create instance with specific instance_type that has specific instance_type_extra_specs that contains all labels | 16:15 |
| boris-42 | I don't see any problem in this part (also instance_type could have name Tiny with Intel XXX card) | 16:16 |
| *** navid_ has left #openstack-dev | 16:16 | |
| *** freedomhui has quit IRC | 16:16 | |
| boris-42 | yjiang5 could you say me about your use cases?) | 16:17 |
| boris-42 | yjiang5 where this approach sucks=) | 16:17 |
| *** Alexei_987 has left #openstack-dev | 16:17 | |
| *** flaper87 has quit IRC | 16:17 | |
| *** JordanP has quit IRC | 16:18 | |
| *** blues-man has quit IRC | 16:20 | |
| *** Yuan has quit IRC | 16:20 | |
| *** terryh has joined #openstack-dev | 16:21 | |
| *** hemnafk is now known as hemna_ | 16:21 | |
| *** fbo has quit IRC | 16:21 | |
| *** rods has joined #openstack-dev | 16:21 | |
| *** jecarey_ has joined #openstack-dev | 16:21 | |
| bnemec | I'm looking to sync the fix for https://bugs.launchpad.net/quantum/+bug/1175808 from Oslo to Quantum, but there are a bunch of ZMQ changes that are getting pulled in too. | 16:22 |
| uvirtbot | Launchpad bug 1175808 in nova/grizzly "Qpid cannot serialize objects containing strings longer than 65535 characters" [Critical,In progress] | 16:22 |
| *** jecarey has quit IRC | 16:22 | |
| bnemec | Should I just go ahead and include them, or remove them from this commit since it's not related to the bug I'm trying to fix? | 16:22 |
| boris-42 | yjiang5 As I understand our situation, instance_type_extra_specs could be changed only by admin, instance_type also. | 16:23 |
| boris-42 | yjiang5 So user is able to specify only flavor and some other data. | 16:23 |
| yjiang5 | boris-42: user want a network pass-through support in the instance. He has driver for intel_8259 driver in the image, then he want to create the instance through instance type. The cloud has two machines , one have an Intel 82598AT card which is using intel 82598 controller, another one is Supermicro AOC-UTG-i2. What do you think the admin will specify the label for these two devices and how can user know that these two devices are intel 82598 base | 16:24 |
| *** colinmcnamara has quit IRC | 16:24 | |
| yjiang5 | AOC-UTG-i2 is also based on Intel 82598 controller | 16:24 |
| *** markmc has quit IRC | 16:25 | |
| *** terriyu has quit IRC | 16:25 | |
| *** michchap has quit IRC | 16:25 | |
| boris-42 | yjiang5 great example, now we could think about extending my solution to cover this case=) | 16:25 |
| *** ppyy has quit IRC | 16:25 | |
| *** bswartz has joined #openstack-dev | 16:26 | |
| *** dstanek has joined #openstack-dev | 16:26 | |
| boris-42 | yjiang5 is there any difference for user what card will he get?) | 16:27 |
| dolphm | ttx: there's a small patch attached to bug 1177630 by an author who I don't see on the CLA list... can i put the commit up for review with the actual --author ? | 16:27 |
| uvirtbot | Launchpad bug 1177630 in keystone "LDAP group search doesn't use filter" [Medium,New] https://launchpad.net/bugs/1177630 | 16:27 |
| clarkb | dolphm: are you looking at the launchpad CLA group? we don't use that anymore and may actually need to query gerrit for that info | 16:28 |
| boris-42 | yjiang5 if there is then we should specify 2 different labels for these 2 devices, if there is no difference then we should create only with one | 16:28 |
| clarkb | fungi: ^ do you know if that is the case? | 16:28 |
| *** dansmith is now known as Steely_Dan | 16:28 | |
| *** jecarey has joined #openstack-dev | 16:28 | |
| *** timello has quit IRC | 16:28 | |
| dolphm | clarkb: yeah, i was looking at https://launchpad.net/~openstack-cla/ | 16:28 |
| boris-42 | yjiang5 But I understand that it will better to use to able to specify features separated=) | 16:28 |
| *** chandankumar has joined #openstack-dev | 16:28 | |
| boris-42 | better to be able* | 16:29 |
| *** jecarey_ has quit IRC | 16:29 | |
| *** colinmcnamara has joined #openstack-dev | 16:29 | |
| yjiang5 | boris-42: If from device itself, it should have no difference, that means these two card should be covered by same label specification when assignment/filter. But for NIC, there are something more, like these two card are connected to different switch or different port. | 16:29 |
| clarkb | dolphm: when we switched to using Gerrit to manage the CLA signing we effectively stopped using that group. It is possible the author has signed a CLA since then | 16:29 |
| *** mrodden1 has joined #openstack-dev | 16:30 | |
| * clarkb looks up the bug to see who wrote the patch | 16:30 | |
| boris-42 | yjiang5 Ok I understand the idea, we should move from abstract name to dynamically added features | 16:30 |
| yjiang5 | exactly. | 16:31 |
| yjiang5 | glad we are on the same page now. | 16:31 |
| *** seanrob has joined #openstack-dev | 16:31 | |
| boris-42 | yjiang5 finally=) | 16:31 |
| *** markmcclain has quit IRC | 16:31 | |
| *** chandankumar has quit IRC | 16:32 | |
| *** mrodden has quit IRC | 16:32 | |
| yjiang5 | Ok, I will wait for your spliced patchsets, and also you can consider this more. | 16:33 |
| *** zodiak has quit IRC | 16:33 | |
| *** chandankumar has joined #openstack-dev | 16:33 | |
| boris-42 | yjiang5 wait a sec=) | 16:33 |
| *** ffio has quit IRC | 16:33 | |
| clarkb | dolphm: looks like Brandon does have a gerrit account but has not signed a CLA | 16:33 |
| boris-42 | yjiang5 I would like to try to find some way to implement (our make it easier) to implement this approach | 16:33 |
| *** zodiak has joined #openstack-dev | 16:33 | |
| boris-42 | yjiang5 Did you see my approach? | 16:34 |
| yjiang5 | boris-42: ? | 16:34 |
| yjiang5 | boris-42: I didn't get your idea, sorry. | 16:34 |
| dolphm | clarkb: how are you checking? i don't see like a 'gerrit ls-users' documented | 16:35 |
| *** comay has joined #openstack-dev | 16:35 | |
| clarkb | dolphm: unfortunately by querying the DB directly | 16:35 |
| boris-42 | yjiang5 At this moment I don't know how to improve (refactor) my solution to make it easier to implement feature-based filtering | 16:35 |
| openstackgerrit | A change was merged to openstack/nova: Replace openstack-common with oslo in HACKING.rst https://review.openstack.org/32114 | 16:35 |
| dolphm | clarkb: boo... well, thanks | 16:35 |
| clarkb | gerrit doesn't expose user details very well on its own | 16:35 |
| *** seanrob has quit IRC | 16:36 | |
| boris-42 | yjiang5 Current solution is pretty simple, each device have compute_node_id and label and status so it is pretty easy to specify device/schedule/attach to instance | 16:36 |
| *** markvoelker has joined #openstack-dev | 16:37 | |
| boris-42 | yjiang5 And for example one of question is what changes should I make at DB layer to provide features?) | 16:37 |
| *** aeperezt has joined #openstack-dev | 16:37 | |
| boris-42 | yjiang5 add something like PCIDeviceExtraSpecs? | 16:37 |
| *** anfrolov has quit IRC | 16:37 | |
| *** lloydde has joined #openstack-dev | 16:37 | |
| *** timello has joined #openstack-dev | 16:37 | |
| boris-42 | yjiang5 so for the devices with same label we will be able to have some extra info that could be used at scheduler level? | 16:38 |
| clarkb | which I suppose is a cheap way of doing things securely. don't expose anything | 16:38 |
| *** networkstatic has joined #openstack-dev | 16:39 | |
| boris-42 | yjiang5 As I am refactoring code, it will be not so complex for me to improve a little bit solution | 16:39 |
| boris-42 | yjiang5 what do you think about new table with extra specs?) | 16:39 |
| *** terry7 has joined #openstack-dev | 16:39 | |
| yjiang5 | boris-42: Yes, agree. Otherthan that, I'm not sure if we can set the product_id and vendor_id in the table, which is standard PCI information, and the extra spec can be used to specify more device specifc information like network connection information. | 16:40 |
| *** eglynn_ has quit IRC | 16:41 | |
| *** alop has joined #openstack-dev | 16:41 | |
| *** freedomhui has joined #openstack-dev | 16:41 | |
| boris-42 | ok I will add also product_id and vendor_id in main table, btw I think that we should add also type_of_device in main table? | 16:42 |
| *** alop has quit IRC | 16:42 | |
| ayoung | derekh, can you +2 this one ? https://review.openstack.org/#/c/32018/ | 16:42 |
| *** jtomasek has quit IRC | 16:42 | |
| boris-42 | gpu, usb, pci etc yjiang5 ^ | 16:42 |
| ayoung | derekh, I think you are the only puppet core available for a second review | 16:42 |
| yjiang5 | boris-42: when you talking about type_of_device, you mean bus type? or the pci device type as in configuration space, or you mean VF/PF etc? | 16:43 |
| *** alop has joined #openstack-dev | 16:43 | |
| *** pixelbeat has quit IRC | 16:43 | |
| boris-42 | yjiang5 I mean pci, usb or gnu | 16:43 |
| boris-42 | gpu* | 16:43 |
| *** terriyu has joined #openstack-dev | 16:44 | |
| boris-42 | for example usb-passthrough should be made not through hostdev | 16:44 |
| boris-42 | in libvirt | 16:44 |
| boris-42 | so type should specify the way of passthrough at virt layer | 16:45 |
| yjiang5 | boris-42: I think GPU and USB is different. usually GPU is in PCI bus, while USB device is in usb bus. I only thought of PCI devices yet. But at least, I don't think we should keep information that apply only to libvirt in the table, do you agree? | 16:45 |
| *** jpich has quit IRC | 16:46 | |
| *** kaushikc has joined #openstack-dev | 16:46 | |
| *** gary_th has quit IRC | 16:46 | |
| boris-42 | yjiang5 Ok I think that usb-passthrough could wait a little bit=) | 16:46 |
| yjiang5 | boris-42: :) | 16:47 |
| boris-42 | yjiang5 Ok so I will fix db layer so you will be able to extend scheduler power=) | 16:47 |
| yjiang5 | boris-42: thanks. | 16:48 |
| boris-42 | yjiang5 Also did you see my approach to add devices through nova.conf | 16:48 |
| boris-42 | yjiang5 I will extend it to make it possible to specify also extra_specs, vendor and product id ok? | 16:48 |
| *** markmcclain has joined #openstack-dev | 16:48 | |
| boris-42 | yjiang5 As I understand you would like to have a another way to specify available devices? | 16:49 |
| boris-42 | yjiang5 in automode? | 16:49 |
| derekh | ayoung: reviewed, I think the user/goup flags need to be added | 16:49 |
| derekh | https://review.openstack.org/#/c/29644/3 | 16:49 |
| yjiang5 | boris-42: I read your patch. Yes, it's ok to add more specificication in the conf. I'm considering the auto discovery. | 16:49 |
| *** alop has quit IRC | 16:50 | |
| *** alop_ has joined #openstack-dev | 16:50 | |
| *** ewindisch has quit IRC | 16:50 | |
| yjiang5 | What I'm considered is, the hypervisor (or libvirt) knows about the devices that can be assigned, and administrator can provide a whte list that should not be assigned. | 16:50 |
| *** athomas has quit IRC | 16:51 | |
| yjiang5 | boris-42: not all device in the system is available for assignment, I think. | 16:51 |
| *** chandankumar has quit IRC | 16:51 | |
| *** mlavalle has quit IRC | 16:51 | |
| yjiang5 | boris-42: the auto discovery mechanism will get all devices that is assignable to guest. Do you agree? | 16:52 |
| *** michchap has joined #openstack-dev | 16:52 | |
| *** kaushikc has quit IRC | 16:52 | |
| *** krtaylor has quit IRC | 16:52 | |
| boris-42 | yjiang5 I agree that such functionality is great, but I am developer, so for me is most interesting is it easy to implement such functionality over current solution (with new changes such as extra specs and etc.) ? | 16:53 |
| yjiang5 | sure, we can add it later | 16:53 |
| *** markwash has joined #openstack-dev | 16:54 | |
| yjiang5 | boris-42: just use your static method currently and we can condier how to enhance later. | 16:54 |
| boris-42 | yjiang5 so we will add some new conf parameter, and change a little bit pci/__init__.py to cover two way of specifying devices?) | 16:54 |
| yjiang5 | yes. | 16:55 |
| boris-42 | yjiang5 Ok nice, so know as developer I understand what should I do to help you with extending pci devices in the way that you need. | 16:55 |
| boris-42 | now* | 16:55 |
| boris-42 | yjiang5 thanks for cooperating | 16:56 |
| yjiang5 | I'm developer also, not provider, although I always want to be at cloudprovider :) | 16:56 |
| yjiang5 | see you. | 16:56 |
| boris-42 | I hope I will have enough time to make all this change before monday =) | 16:56 |
| boris-42 | these changes* | 16:56 |
| *** esp has joined #openstack-dev | 16:56 | |
| yjiang5 | anyway, wait for your patches. | 16:57 |
| *** kbrierly has joined #openstack-dev | 16:58 | |
| *** esp has left #openstack-dev | 16:59 | |
| vishy | arosen: if you are available today and you can handle a few more questions i have them :) | 17:01 |
| *** rnirmal has quit IRC | 17:01 | |
| *** markvoelker1 has joined #openstack-dev | 17:02 | |
| *** michchap has quit IRC | 17:04 | |
| *** danpb has quit IRC | 17:04 | |
| *** nati_ueno has joined #openstack-dev | 17:04 | |
| *** markvoelker has quit IRC | 17:04 | |
| *** markvoelker has joined #openstack-dev | 17:05 | |
| *** derekh has quit IRC | 17:05 | |
| *** rcleere has joined #openstack-dev | 17:06 | |
| *** markvoelker1 has quit IRC | 17:07 | |
| *** gary_th has joined #openstack-dev | 17:07 | |
| *** rcleere has quit IRC | 17:07 | |
| vishy | markmcclain: was it you who wrote the quantum metadata agent? | 17:07 |
| *** darjeeling has quit IRC | 17:09 | |
| markmcclain | vishy: yep | 17:10 |
| vishy | markmcclain: got a sec so i can ask you a few questions? | 17:10 |
| markmcclain | sure..what's up? | 17:10 |
| vishy | ok | 17:10 |
| vishy | so assuming I have created a quantum provider network | 17:11 |
| vishy | and i'm running the dhcp agent | 17:11 |
| vishy | but it is passing out a gateway that is external to openstack | 17:11 |
| vishy | i.3. the l3 agent is not doing router duties | 17:11 |
| vishy | * i.e. | 17:12 |
| vishy | will the quantum metadata proxy still work? | 17:12 |
| *** yjiang5 has quit IRC | 17:13 | |
| markmcclain | right now no because there's an assumption that the upstream gateway is the L3 agent | 17:14 |
| vishy | markmcclain: bummer | 17:14 |
| markmcclain | that's a case we should fix | 17:14 |
| vishy | markmcclain: is there any way i can hack around it? | 17:14 |
| vishy | I would like to run the agent on the nodes | 17:14 |
| vishy | and basicallly have it intercept traffic before it leaves the node | 17:15 |
| harlowja | zyluo pong | 17:15 |
| *** prad_ has joined #openstack-dev | 17:15 | |
| markmcclain | vishy: when the DHCP agent detects there is not a gateway defined for the network it sets a static route to itself for the meta IP | 17:16 |
| *** krtaylor has joined #openstack-dev | 17:16 | |
| markmcclain | you could override that functionality and basically force it to set that route always | 17:16 |
| vishy | markmcclain: but there is a gateway defined for the network | 17:16 |
| vishy | you mean detects there is no l3 agent? | 17:16 |
| *** Yuan has joined #openstack-dev | 17:17 | |
| arosen | the subnet on the network was created specifying that it doesn't have a gateway_ip | 17:17 |
| *** sdake_ has joined #openstack-dev | 17:18 | |
| *** sdake_ has joined #openstack-dev | 17:18 | |
| vishy | arosen: ah i didn't realize there was a property for gateway on subnet | 17:18 |
| *** lucasagomes has quit IRC | 17:19 | |
| arosen | yup it's exposed via the python-client as --no-gateway on create-subnet() | 17:19 |
| * arosen subnet-create | 17:20 | |
| markmcclain | vishy: could enable_isolated_metadata=true but you'll probably still need to tweak some code until we add support for forcing the metadata routes always | 17:21 |
| vishy | looking back | 17:21 |
| vishy | so i did use gateway when i created the ntetwork | 17:21 |
| vishy | * subnet | 17:21 |
| vishy | but this router is an external entity | 17:21 |
| *** vipul is now known as vipul|away | 17:22 | |
| fungi | dolphm: clarkb: yeah, we need to delete the old cla group off lp i guess | 17:22 |
| *** jjmb has quit IRC | 17:22 | |
| vishy | markmcclain: i guess i don't fully grasp how the metadata is getting the traffic in the first place | 17:22 |
| vishy | markmcclain: does it add a route to the l3 config for 169.254.169.254? | 17:23 |
| *** dstanek has quit IRC | 17:23 | |
| markmcclain | for isolated networks it adds the route | 17:23 |
| vishy | and non-isolated? | 17:23 |
| *** jcoufal has quit IRC | 17:23 | |
| markmcclain | when the L3 agent is running.. it abuses a quirk or linux routing where even link local traffic is sent to the default gw | 17:24 |
| vishy | interesting | 17:25 |
| vishy | what i want to do is mirror what i have curently with nova network | 17:26 |
| vishy | which is basically intercept traffic on the host that is destined for 169.254.169.254 and send it to the metadata server | 17:26 |
| vishy | it seems like you could do that fairly easily with a flow rule, no? | 17:26 |
| *** ndipanov has quit IRC | 17:26 | |
| *** monst has joined #openstack-dev | 17:27 | |
| arosen | yup, if you do ip netns exec <namespace> ip addr ; i think you'll see the 169.254 address also on the default gateway interface. That's how it picks up the packets . | 17:27 |
| vishy | markmcclain: i will look into that in more detail soon | 17:28 |
| *** Yuan has quit IRC | 17:28 | |
| vishy | markmcclain/arosen: another set of questions. So if i have multiple l3 agents that i want to serve a single network | 17:28 |
| arosen | vishy: the issue is overlapping ips. we need to send it to the agent first to do the look for the header that is sent to nova. | 17:28 |
| markmcclain | yes except when you're using overlapping IPs..the traffic needs to be run through a proxy to add context otherwise the source ip is ambiguous for a project | 17:29 |
| *** vipul|away is now known as vipul | 17:29 | |
| vishy | markmcclain: sure i meant send it to the proxy | 17:29 |
| *** jjmb has joined #openstack-dev | 17:29 | |
| *** nati_ueno has quit IRC | 17:30 | |
| vishy | markmcclain/arosen: so multiple l3 agents, do i need to create a separate quantum router for each of the l3 agents? | 17:30 |
| *** monst has quit IRC | 17:30 | |
| arosen | I think yongs work will just pick one agent. I'm not sure if it does any kind of active active / active-standby right now. | 17:30 |
| *** michchap has joined #openstack-dev | 17:30 | |
| markmcclain | it only picks one agent | 17:30 |
| markmcclain | and does not alter the default gateway yet | 17:31 |
| *** nati_ueno has joined #openstack-dev | 17:31 | |
| *** markvoelker1 has joined #openstack-dev | 17:31 | |
| *** seanrob has joined #openstack-dev | 17:32 | |
| vishy | markmcclain: I just found this: http://docs.openstack.org/trunk/openstack-network/admin/content/under_the_hood_openvswitch.html | 17:32 |
| vishy | very useful :) | 17:32 |
| arosen | cool i haven't seen that either :) | 17:32 |
| *** markvoelker has quit IRC | 17:33 | |
| vishy | arosen: I'm a little confused by --router:external=True . Any idea what it does? | 17:34 |
| arosen | vishy: one sec i made a powerpoint that explains this that could be helpful since it visualizes the steps. | 17:34 |
| arosen | it's shared on google one sec* | 17:35 |
| *** jog0-away is now known as flashgordon | 17:35 | |
| arosen | https://docs.google.com/presentation/d/18uuHfa5pq6VMrh8gpwXXiomC7hM1Dzy7t7CHF9ktgUk/edit#slide=id.p | 17:35 |
| *** mrunge has joined #openstack-dev | 17:36 | |
| arosen | slide 5 | 17:36 |
| *** armax has joined #openstack-dev | 17:36 | |
| *** seanrob has quit IRC | 17:36 | |
| arosen | vishy: this went along with this tutorial if you hadn't seen it: http://blog.aaronorosen.com/building-a-multi-tier-application-with-openstack/ | 17:37 |
| *** drewlander has joined #openstack-dev | 17:37 | |
| vishy | ah i gotcha, so the external means tenant routers can use it for uplink? | 17:38 |
| arosen | yup | 17:38 |
| arosen | you can only do ; quantum router-gateway-set on external networks. | 17:38 |
| arosen | i guess it also kind of denotes that it's routed address space in your infrastructure | 17:39 |
| * arosen or at least should be :) | 17:39 | |
| *** markvoelker has joined #openstack-dev | 17:39 | |
| *** pabelanger has quit IRC | 17:39 | |
| *** nunosantos has joined #openstack-dev | 17:40 | |
| *** drewland_ has quit IRC | 17:40 | |
| *** drewlander has quit IRC | 17:41 | |
| *** michchap has quit IRC | 17:41 | |
| *** markvoelker1 has quit IRC | 17:42 | |
| *** lbragstad has quit IRC | 17:43 | |
| *** changbl has joined #openstack-dev | 17:43 | |
| *** markwash has quit IRC | 17:44 | |
| *** sstent has quit IRC | 17:44 | |
| *** bradjones|away has quit IRC | 17:45 | |
| *** bradjones|away has joined #openstack-dev | 17:45 | |
| *** _sifusam_ has quit IRC | 17:45 | |
| *** armax has quit IRC | 17:46 | |
| *** dolphm has quit IRC | 17:46 | |
| *** mlavalle has joined #openstack-dev | 17:46 | |
| *** armax has joined #openstack-dev | 17:47 | |
| *** jvrbanac has quit IRC | 17:47 | |
| *** markwash has joined #openstack-dev | 17:47 | |
| *** terryh has quit IRC | 17:47 | |
| *** dhellmann has quit IRC | 17:48 | |
| *** savid has joined #openstack-dev | 17:48 | |
| savid | Hi guys, I'm a bit confused about services in keystone. In the v2 API, services are created with name, type, description, however in the v3 API, services are created with name, type, and an "enabled". What exactly does "enabled" do to an entry in the service catalog? | 17:49 |
| ayoung | savid, we added the abstraction of allowing services to be disabled. It does absolutly nothing. | 17:49 |
| savid | ayoung, Ah, I see. Can description still be set in the v3 api? | 17:50 |
| *** yjiang5 has joined #openstack-dev | 17:50 | |
| ayoung | savid, actually, I don't even see it in the database tables | 17:50 |
| *** freedomhui has quit IRC | 17:50 | |
| ayoung | https://github.com/openstack/keystone/blob/master/keystone/catalog/backends/sql.py | 17:50 |
| savid | don't see what, description? | 17:51 |
| *** ewindisch has joined #openstack-dev | 17:51 | |
| savid | I'm on devstack right now, and when calling keystone service-list it returns id, name, type, and description | 17:51 |
| *** Ryan_Lane has quit IRC | 17:51 | |
| *** Ryan_Lane has joined #openstack-dev | 17:52 | |
| ayoung | savid, I believe you | 17:52 |
| ayoung | I don't see enabled or description in the database for service | 17:52 |
| *** NobodyCam has quit IRC | 17:52 | |
| *** eglynn_ has joined #openstack-dev | 17:53 | |
| savid | so, question is, *should* they be there? | 17:53 |
| *** jvrbanac has joined #openstack-dev | 17:53 | |
| ayoung | savid, lets see what the API says | 17:53 |
| *** sifusam has joined #openstack-dev | 17:53 | |
| ayoung | savid https://github.com/openstack/identity-api/blob/master/openstack-identity-api/src/markdown/identity-api-v3.md | 17:54 |
| ayoung | https://github.com/openstack/identity-api/blob/master/openstack-identity-api/src/markdown/identity-api-v3.md#services-v3services | 17:55 |
| ayoung | savid, I think it is a bug | 17:55 |
| *** carl_baldwin has joined #openstack-dev | 17:55 | |
| ayoung | it is inheriting enabled and description from some base class, but it isn';t in the DB.... | 17:55 |
| ayoung | It is V3ism | 17:55 |
| ayoung | description doesn't seem to be there, though | 17:56 |
| ayoung | savid, open a bug | 17:56 |
| ayoung | I'm guessing dolphm will take it. He's kindof the master of all things V3 | 17:56 |
| savid | hah, a V3ism. nice. | 17:56 |
| savid | ok, thanks | 17:56 |
| *** zaitcev has joined #openstack-dev | 17:56 | |
| *** eglynn_ has quit IRC | 17:57 | |
| *** seanrob has joined #openstack-dev | 17:59 | |
| *** deeps___ has joined #openstack-dev | 18:03 | |
| *** seanrob has quit IRC | 18:04 | |
| *** alop_ has quit IRC | 18:04 | |
| *** andrewbogott is now known as andrewbogott_afk | 18:06 | |
| *** michchap has joined #openstack-dev | 18:08 | |
| *** lbragstad has joined #openstack-dev | 18:09 | |
| *** terriyu has quit IRC | 18:11 | |
| *** hugokuo has quit IRC | 18:12 | |
| *** Brainspa1kle has quit IRC | 18:12 | |
| *** swifterdarrell has quit IRC | 18:12 | |
| *** terriyu has joined #openstack-dev | 18:13 | |
| *** easter has joined #openstack-dev | 18:13 | |
| *** seanrob has joined #openstack-dev | 18:13 | |
| *** seanrob has quit IRC | 18:13 | |
| *** easter has left #openstack-dev | 18:13 | |
| *** johnthetubaguy has quit IRC | 18:13 | |
| *** cdub has joined #openstack-dev | 18:14 | |
| *** imsplitbit is now known as splitster | 18:15 | |
| *** seanrob has joined #openstack-dev | 18:15 | |
| *** johnthetubaguy has joined #openstack-dev | 18:15 | |
| *** swifterdarrell has joined #openstack-dev | 18:16 | |
| *** Brainspackle has joined #openstack-dev | 18:16 | |
| *** markmcclain has quit IRC | 18:17 | |
| *** hugokuo has joined #openstack-dev | 18:17 | |
| *** sstent has joined #openstack-dev | 18:19 | |
| *** michchap has quit IRC | 18:20 | |
| *** sdake_ has quit IRC | 18:20 | |
| *** reed has joined #openstack-dev | 18:21 | |
| *** alop has joined #openstack-dev | 18:23 | |
| *** Mandell has joined #openstack-dev | 18:24 | |
| *** comstud is now known as bearhands | 18:24 | |
| *** markmcclain has joined #openstack-dev | 18:24 | |
| *** rharwood has quit IRC | 18:24 | |
| *** krtaylor has quit IRC | 18:25 | |
| *** rharwood has joined #openstack-dev | 18:25 | |
| *** jruzicka has quit IRC | 18:27 | |
| *** cp16net|away is now known as cp16net | 18:27 | |
| savid | ayoung, it looks like I can pass description in the V3 API afterall. It must store it in the "extra" JSON blob column. | 18:29 |
| *** dolphm has joined #openstack-dev | 18:29 | |
| *** dhellmann has joined #openstack-dev | 18:30 | |
| *** vipul is now known as vipul|away | 18:30 | |
| *** vipul|away is now known as vipul | 18:30 | |
| *** cmark has joined #openstack-dev | 18:31 | |
| *** seanrob has quit IRC | 18:31 | |
| *** jjmb has quit IRC | 18:35 | |
| *** dosaboy has quit IRC | 18:37 | |
| *** jpeeler has joined #openstack-dev | 18:38 | |
| *** jpeeler has quit IRC | 18:38 | |
| *** jpeeler has joined #openstack-dev | 18:38 | |
| *** jjmb has joined #openstack-dev | 18:38 | |
| *** marun has quit IRC | 18:44 | |
| *** michchap has joined #openstack-dev | 18:47 | |
| *** dhellmann is now known as dhellmann-away | 18:49 | |
| *** timello has quit IRC | 18:49 | |
| *** cmark has quit IRC | 18:49 | |
| *** cmark has joined #openstack-dev | 18:50 | |
| *** krtaylor has joined #openstack-dev | 18:50 | |
| *** edewata has quit IRC | 18:55 | |
| openstackgerrit | A change was merged to openstack/swift: Remove <cond> and <val1> or <val2> systax https://review.openstack.org/31943 | 18:55 |
| *** jbresnah has joined #openstack-dev | 18:55 | |
| *** edewata has joined #openstack-dev | 18:55 | |
| *** novas0x2a|laptop has joined #openstack-dev | 18:56 | |
| *** shang has quit IRC | 18:57 | |
| *** sushils has quit IRC | 18:57 | |
| *** michchap has quit IRC | 18:58 | |
| *** pabelanger has joined #openstack-dev | 18:58 | |
| *** andrewbogott_afk is now known as andrewbogott | 18:59 | |
| *** radez is now known as radez_g0n3 | 19:00 | |
| *** rods has left #openstack-dev | 19:01 | |
| *** SergeyLukjanov has joined #openstack-dev | 19:03 | |
| *** andrew_plunk has quit IRC | 19:04 | |
| *** cmark has quit IRC | 19:04 | |
| *** sstent has quit IRC | 19:05 | |
| *** flaper87 has joined #openstack-dev | 19:06 | |
| *** mrunge has quit IRC | 19:09 | |
| *** colinmcnamara has quit IRC | 19:09 | |
| *** egallen has quit IRC | 19:12 | |
| *** bswartz has quit IRC | 19:12 | |
| *** timello has joined #openstack-dev | 19:13 | |
| *** marun has joined #openstack-dev | 19:16 | |
| *** SergeyLukjanov has quit IRC | 19:17 | |
| *** dolphm has quit IRC | 19:21 | |
| *** colinmcnamara has joined #openstack-dev | 19:21 | |
| *** prad_ has quit IRC | 19:21 | |
| *** sandywalsh has quit IRC | 19:23 | |
| *** dstanek has joined #openstack-dev | 19:24 | |
| *** prad_ has joined #openstack-dev | 19:25 | |
| *** eharney has quit IRC | 19:25 | |
| *** michchap has joined #openstack-dev | 19:25 | |
| *** markmcclain has quit IRC | 19:27 | |
| *** alop has quit IRC | 19:31 | |
| *** SergeyLukjanov has joined #openstack-dev | 19:32 | |
| *** markmcclain has joined #openstack-dev | 19:32 | |
| *** kagan has joined #openstack-dev | 19:33 | |
| *** dolphm has joined #openstack-dev | 19:33 | |
| *** garyTh has joined #openstack-dev | 19:35 | |
| *** sandywalsh has joined #openstack-dev | 19:36 | |
| *** michchap has quit IRC | 19:36 | |
| *** gary_th has quit IRC | 19:37 | |
| *** monst has joined #openstack-dev | 19:38 | |
| *** sstent has joined #openstack-dev | 19:39 | |
| *** sdake_ has joined #openstack-dev | 19:39 | |
| *** sdake_ has quit IRC | 19:39 | |
| *** sdake_ has joined #openstack-dev | 19:39 | |
| *** jpeeler has quit IRC | 19:48 | |
| *** dprince has quit IRC | 19:52 | |
| *** rkukura has quit IRC | 19:52 | |
| *** devvesa has quit IRC | 19:52 | |
| openstackgerrit | A change was merged to openstack/nova: Rename unique constraints due to new convention. https://review.openstack.org/30108 | 19:53 |
| *** edewata has quit IRC | 19:53 | |
| *** edewata has joined #openstack-dev | 19:55 | |
| *** sstent has quit IRC | 19:57 | |
| *** bswartz has joined #openstack-dev | 19:57 | |
| *** sushils has joined #openstack-dev | 20:01 | |
| *** otherwiseguy has quit IRC | 20:01 | |
| *** michchap has joined #openstack-dev | 20:04 | |
| *** melwitt has joined #openstack-dev | 20:04 | |
| *** eharney has joined #openstack-dev | 20:07 | |
| *** eharney has quit IRC | 20:07 | |
| *** eharney has joined #openstack-dev | 20:07 | |
| *** mkollaro has quit IRC | 20:07 | |
| *** dstanek has quit IRC | 20:12 | |
| *** jayg is now known as jayg|g0n3 | 20:14 | |
| *** jayg|g0n3 is now known as jayg | 20:14 | |
| *** yidclare has quit IRC | 20:15 | |
| *** michchap has quit IRC | 20:17 | |
| *** monst has quit IRC | 20:21 | |
| *** pabelanger has quit IRC | 20:21 | |
| *** bdpayne has quit IRC | 20:21 | |
| *** monst has joined #openstack-dev | 20:23 | |
| *** lastidiot has quit IRC | 20:23 | |
| *** bdpayne has joined #openstack-dev | 20:23 | |
| *** monst has quit IRC | 20:24 | |
| *** ladquin_away is now known as ladquin | 20:26 | |
| *** ayoung is now known as ayoung-out | 20:26 | |
| *** mkollaro has joined #openstack-dev | 20:26 | |
| *** monst has joined #openstack-dev | 20:26 | |
| *** stevemar has quit IRC | 20:26 | |
| *** sushils has quit IRC | 20:27 | |
| *** dstanek has joined #openstack-dev | 20:28 | |
| *** sushils has joined #openstack-dev | 20:28 | |
| *** monst has quit IRC | 20:29 | |
| *** monst has joined #openstack-dev | 20:29 | |
| *** kagan has quit IRC | 20:30 | |
| *** monst has joined #openstack-dev | 20:30 | |
| *** ayoung-out is now known as ayoung | 20:30 | |
| *** SergeyLukjanov has quit IRC | 20:30 | |
| *** yidclare has joined #openstack-dev | 20:30 | |
| *** armax has left #openstack-dev | 20:32 | |
| *** ctracey is now known as ctracey|away | 20:32 | |
| *** seanrob has joined #openstack-dev | 20:33 | |
| *** mtreinis1 has joined #openstack-dev | 20:33 | |
| *** maurosr- has joined #openstack-dev | 20:34 | |
| *** alunduil has joined #openstack-dev | 20:34 | |
| *** bknudson1 has joined #openstack-dev | 20:35 | |
| *** seanrob has quit IRC | 20:35 | |
| *** seanrob has joined #openstack-dev | 20:36 | |
| *** mtreinish has quit IRC | 20:36 | |
| *** seanrob has quit IRC | 20:36 | |
| *** tmclaugh[work] has quit IRC | 20:36 | |
| *** maurosr| has joined #openstack-dev | 20:36 | |
| *** seanrob has joined #openstack-dev | 20:36 | |
| *** gary_th has joined #openstack-dev | 20:36 | |
| *** bknudson has quit IRC | 20:36 | |
| *** maurosr has quit IRC | 20:37 | |
| *** bknudson has joined #openstack-dev | 20:37 | |
| *** otherwiseguy has joined #openstack-dev | 20:37 | |
| *** garyTh has quit IRC | 20:37 | |
| *** galstrom is now known as galstrom_zzz | 20:38 | |
| *** maurosr| is now known as maurosr | 20:38 | |
| *** mtreinish has joined #openstack-dev | 20:38 | |
| *** bknudson1 has quit IRC | 20:39 | |
| *** mtreinis1 has quit IRC | 20:40 | |
| *** maurosr- has quit IRC | 20:40 | |
| *** yolanda has quit IRC | 20:42 | |
| *** michchap has joined #openstack-dev | 20:42 | |
| *** yidclare has quit IRC | 20:45 | |
| openstackgerrit | A change was merged to openstack-dev/devstack: Support for the ml2 quantum plugin. https://review.openstack.org/27576 | 20:46 |
| *** michchap has quit IRC | 20:47 | |
| openstackgerrit | A change was merged to openstack/keystone: Check schema when dropping constraints. https://review.openstack.org/31249 | 20:47 |
| *** dhellmann-away is now known as dhellmann | 20:47 | |
| SpamapS | wth I thought sqlalchemy was supposed to know when I'm doing something bad and do something smarter.. :-/ | 20:47 |
| SpamapS | sqlalchemy.exc.NotSupportedError: (NotSupportedError) (1235, "This version of MySQL doesn't yet support 'LIMIT & IN/ALL/ANY/SOME subquery'") 'DELETE FROM token WHERE token.id IN (SELECT token.id \nFROM token \nWHERE token.expires < %s \n LIMIT %s)' (datetime.datetime(2013, 6, 7, 20, 46, 47, 570583), 10) | 20:47 |
| *** topol has quit IRC | 20:48 | |
| *** dstanek has quit IRC | 20:48 | |
| *** yidclare has joined #openstack-dev | 20:50 | |
| *** krtaylor has quit IRC | 20:56 | |
| *** aelkikhia has quit IRC | 20:59 | |
| SpamapS | durn it, sqlalchemy doesn't support DELETE ... LIMIT | 21:03 |
| *** mjfs has quit IRC | 21:06 | |
| *** bswartz has quit IRC | 21:08 | |
| *** bdpayne has quit IRC | 21:09 | |
| *** alop has joined #openstack-dev | 21:10 | |
| *** rharwood has quit IRC | 21:10 | |
| *** jjmb has quit IRC | 21:12 | |
| *** michchap has joined #openstack-dev | 21:13 | |
| *** bdpayne has joined #openstack-dev | 21:15 | |
| *** mjfs has joined #openstack-dev | 21:16 | |
| *** jjmb has joined #openstack-dev | 21:16 | |
| *** mkollaro has quit IRC | 21:16 | |
| *** dolphm has quit IRC | 21:19 | |
| *** alop has quit IRC | 21:20 | |
| *** jjmb has quit IRC | 21:22 | |
| *** jasdeepH has joined #openstack-dev | 21:23 | |
| *** andrew_plunk has joined #openstack-dev | 21:23 | |
| *** davidkranz has quit IRC | 21:23 | |
| *** portante has quit IRC | 21:24 | |
| *** jjmb has joined #openstack-dev | 21:24 | |
| *** michchap has quit IRC | 21:27 | |
| *** mjfs has quit IRC | 21:27 | |
| *** pcm_ has quit IRC | 21:28 | |
| *** alop has joined #openstack-dev | 21:29 | |
| *** deeps___ has quit IRC | 21:29 | |
| *** dolphm has joined #openstack-dev | 21:30 | |
| *** gyee has quit IRC | 21:31 | |
| *** jjmb has quit IRC | 21:32 | |
| *** seanrob has quit IRC | 21:32 | |
| *** anfrolov has joined #openstack-dev | 21:32 | |
| *** seanrob has joined #openstack-dev | 21:33 | |
| *** kbringard has quit IRC | 21:34 | |
| *** markvoelker has quit IRC | 21:35 | |
| *** mkollaro has joined #openstack-dev | 21:36 | |
| *** pixelbeat has joined #openstack-dev | 21:36 | |
| *** dolphm has quit IRC | 21:37 | |
| *** boris-42 has quit IRC | 21:40 | |
| *** katylava has quit IRC | 21:42 | |
| *** alunch has quit IRC | 21:43 | |
| *** alop has quit IRC | 21:43 | |
| *** esheffield has quit IRC | 21:47 | |
| *** seanrob has quit IRC | 21:49 | |
| *** seanrob has joined #openstack-dev | 21:49 | |
| *** alop has joined #openstack-dev | 21:49 | |
| *** PaulM has joined #openstack-dev | 21:50 | |
| *** seanrob has quit IRC | 21:50 | |
| *** PaulM has left #openstack-dev | 21:50 | |
| *** seanrob has joined #openstack-dev | 21:50 | |
| *** monst has quit IRC | 21:51 | |
| alop | any cinder guys here today? | 21:51 |
| *** kagan has joined #openstack-dev | 21:52 | |
| *** dhellmann is now known as dhellmann-away | 21:53 | |
| *** michchap has joined #openstack-dev | 21:53 | |
| *** seanrob has quit IRC | 21:53 | |
| *** seanrob has joined #openstack-dev | 21:54 | |
| *** jimfehlig has quit IRC | 21:55 | |
| *** mlavalle has quit IRC | 21:56 | |
| *** seanrob has quit IRC | 21:58 | |
| *** seanrob has joined #openstack-dev | 21:58 | |
| *** datsun180b has quit IRC | 21:58 | |
| *** morazi has quit IRC | 22:00 | |
| *** bdpayne has quit IRC | 22:01 | |
| *** mkollaro has quit IRC | 22:03 | |
| *** colinmcnamara has quit IRC | 22:04 | |
| jgriffith | alop: what's up? | 22:04 |
| alop | I was hoping someone could take a look at https://bugs.launchpad.net/cinder/+bug/1188422 | 22:04 |
| uvirtbot | Launchpad bug 1188422 in cinder "attributes missing in sizelimit" [Undecided,New] | 22:04 |
| alop | like, it's either broken or I'm missing something | 22:05 |
| jgriffith | alop: I saw that last night... | 22:05 |
| *** markvoelker has joined #openstack-dev | 22:05 | |
| alop | it it real or did I just screw up? | 22:05 |
| jgriffith | alop: my first thought was you screwed up, but that's not fair :) | 22:06 |
| alop | could be | 22:06 |
| jgriffith | alop: doesn't sound like you did anything out of the ordinary though | 22:06 |
| *** michchap has quit IRC | 22:06 | |
| jgriffith | alop: so I assumed it was a bug, just don't know how to repro yet | 22:06 |
| *** vipul is now known as vipul|away | 22:06 | |
| *** vipul|away is now known as vipul | 22:06 | |
| alop | yeah, I'll put a little background info in the report | 22:07 |
| jgriffith | alop: BTW, that's the log output when you try to restart I'm assuming? | 22:07 |
| alop | yes | 22:07 |
| jgriffith | strange | 22:07 |
| jgriffith | alop: Grizzly? | 22:08 |
| alop | yes, stable/grizzly | 22:08 |
| *** mlavalle has joined #openstack-dev | 22:08 | |
| *** vipul is now known as vipul|away | 22:08 | |
| *** vipul|away is now known as vipul | 22:08 | |
| jgriffith | thingee: ping | 22:09 |
| *** andrew_plunk has quit IRC | 22:09 | |
| vishy | arosen: so for external networks | 22:10 |
| vishy | am I responsible for giving the host an ip on the external network? | 22:10 |
| jgriffith | alop: there was one thing I thought of but if everytyhing was fine until the force delete that throws out my theory | 22:10 |
| jgriffith | alop: ahh... wait | 22:10 |
| alop | well, what was your thought? | 22:10 |
| *** markvoelker has quit IRC | 22:10 | |
| *** jjmb has joined #openstack-dev | 22:11 | |
| *** changbl has quit IRC | 22:12 | |
| jgriffith | alop: is there anything further up in those logs or is that it? | 22:12 |
| alop | I'll put a pastie up | 22:13 |
| *** mtreinish has quit IRC | 22:13 | |
| *** nunosantos has quit IRC | 22:13 | |
| *** gary_th has quit IRC | 22:14 | |
| *** lastidiot has joined #openstack-dev | 22:14 | |
| jgriffith | alop: BTW what I was initially thinking was the fault-wrapper deprecation, but you've got the new/correct one it looks like | 22:15 |
| jgriffith | alop: while youre pasting :) maybe your api-paste.ini file if you could | 22:16 |
| *** alop has quit IRC | 22:18 | |
| *** mjfork_ has joined #openstack-dev | 22:19 | |
| *** marun has quit IRC | 22:19 | |
| *** carl_baldwin has quit IRC | 22:19 | |
| *** mjfork has quit IRC | 22:20 | |
| *** mjfork_ is now known as mjfork | 22:20 | |
| *** hartsocks has quit IRC | 22:21 | |
| *** shardy has quit IRC | 22:21 | |
| *** mrodden1 has quit IRC | 22:23 | |
| *** alop has joined #openstack-dev | 22:23 | |
| *** vipul is now known as vipul|away | 22:23 | |
| *** seanrob has quit IRC | 22:25 | |
| *** seanrob has joined #openstack-dev | 22:25 | |
| vishy | markmcclain: so I noticed my router just picked a random ip from my external range. Is there a way to force it to pick a specific ip? | 22:25 |
| *** jjmb1 has joined #openstack-dev | 22:27 | |
| *** jvrbanac has quit IRC | 22:28 | |
| *** dims has quit IRC | 22:28 | |
| *** jjmb has quit IRC | 22:29 | |
| *** markmcclain has quit IRC | 22:29 | |
| *** seanrob has quit IRC | 22:30 | |
| *** jjmb1 has quit IRC | 22:30 | |
| *** vipul|away is now known as vipul | 22:30 | |
| thingee | jgriffith: pong | 22:30 |
| *** mdomsch has quit IRC | 22:32 | |
| *** stevemar has joined #openstack-dev | 22:32 | |
| *** savid has quit IRC | 22:32 | |
| *** michchap has joined #openstack-dev | 22:32 | |
| *** adepti37 has quit IRC | 22:33 | |
| *** pabelanger_ has quit IRC | 22:35 | |
| *** pabelanger has joined #openstack-dev | 22:36 | |
| *** markvoelker has joined #openstack-dev | 22:36 | |
| jgriffith | thingee: I was going to ask if the bug by alop rang any bells for you: https://bugs.launchpad.net/cinder/+bug/1188422 | 22:37 |
| uvirtbot | Launchpad bug 1188422 in cinder "attributes missing in sizelimit" [Undecided,New] | 22:37 |
| *** vuntz has quit IRC | 22:37 | |
| *** jecarey has quit IRC | 22:37 | |
| jgriffith | alop: hey wait... the extension load is what's failing there | 22:38 |
| thingee | jgriffith: lookin' | 22:39 |
| *** markvoelker has quit IRC | 22:41 | |
| arosen | vishy: nope the dhcp agent will do taht for you as well on the external network | 22:41 |
| vishy | arosen: ok i got it to work but is there any way to force it to pick a specific ip? | 22:42 |
| vishy | it just pulled the first ip from the network | 22:42 |
| *** alop has quit IRC | 22:42 | |
| *** prad_ has quit IRC | 22:42 | |
| *** dims has joined #openstack-dev | 22:43 | |
| arosen | vishy: yup if you do quantum port-create --fixed-ip <ip> network ; then nova boot --nic port-id=<port_uuid> | 22:44 |
| vishy | arosen: not for the vm | 22:44 |
| vishy | arosen: for the router | 22:44 |
| vishy | so i have an external network and i plugged a qrouter into it | 22:44 |
| *** michchap has quit IRC | 22:44 | |
| *** bdpayne has joined #openstack-dev | 22:44 | |
| vishy | and it used the first available ip from the external network | 22:44 |
| *** dcasier_ has quit IRC | 22:45 | |
| arosen | Oh sorry i thought we were talking about provider networks... doh... | 22:45 |
| arosen | no when you do: quantum router-gateway-set it picks the first ip out of external network. :/ | 22:45 |
| vishy | arosen: yeah that is annoying | 22:46 |
| arosen | Is there any reason why you care about that? | 22:46 |
| vishy | arosen: i have other things on that network | 22:46 |
| arosen | vishy: that's the ip that your vms are going to use for nat. | 22:46 |
| vishy | arosen: that ip address is in use | 22:46 |
| arosen | Oh that's a different case :) | 22:46 |
| arosen | So when you create the subnet it has something called allocation_pools | 22:46 |
| vishy | arosen: fun fun | 22:46 |
| arosen | so you can say quantum subnet-create --allocation-pool start=10.0.0.20,end=10.0.0.22 --allocation-pool start=10.0.0.33,end=10.0.0.55 | 22:47 |
| vishy | ah cool | 22:47 |
| arosen | and then it will only pick ips out of that range | 22:47 |
| vishy | ok that works | 22:47 |
| vishy | arosen: so the only other issue is that the metadata proxy is still not working | 22:48 |
| vishy | although i think i have l3 working properly now | 22:48 |
| arosen | vishy: how so? What's it doing? | 22:48 |
| vishy | well nothing :) | 22:48 |
| arosen | hrm so if you do in your vm curl http://169.254.169.254 does it get a 500 error back or just hangs? | 22:48 |
| *** alop has joined #openstack-dev | 22:49 | |
| Apsu` | Metadata proxying is quite annoying to setup. | 22:49 |
| *** zzs has joined #openstack-dev | 22:49 | |
| arosen | vishy: I'd tail the metadata agent log while you hit it with curl | 22:50 |
| vishy | i can't yet | 22:50 |
| vishy | :) | 22:50 |
| vishy | wget: can't connect to remote host (169.254.169.254): Connection timed out | 22:50 |
| vishy | cloudsetup: failed 2/30: up 196.32. request failed | 22:50 |
| *** vuntz has joined #openstack-dev | 22:50 | |
| vishy | but the metadata agent log shows nothing | 22:50 |
| vishy | so it isn't getting there | 22:50 |
| arosen | Can you ping the default gw from within the vm? | 22:50 |
| vishy | arosen: i can't login to the vm yet but i can ping the vm from the gateway just fine | 22:51 |
| *** neelashah has quit IRC | 22:51 | |
| vishy | it doesn't set up ssh until after it times out :) | 22:51 |
| vishy | maybe i can get in through the console somehow though :) | 22:51 |
| vishy | let me try | 22:51 |
| arosen | yea i don't think you can since this is part of the init process. It would be nice if it ran the background. | 22:51 |
| *** alop has quit IRC | 22:51 | |
| *** hemna_ is now known as hemnafk | 22:52 | |
| arosen | Is anything happening in the metadata agent log? | 22:52 |
| vishy | arosen: no | 22:52 |
| vishy | i'm not really sure how to start tcpdumping to find out what is happening to the traffic | 22:52 |
| Apsu` | Metadata agent log doesn't spit out anything unless it's failing to actually proxy. I assume this is a routing issue. | 22:52 |
| arosen | this subnet has a gateway_ip right? | 22:52 |
| Apsu` | vishy: Are you using namespaces? | 22:52 |
| *** flaper87 has quit IRC | 22:52 | |
| *** mlavalle has quit IRC | 22:52 | |
| vishy | yeah | 22:53 |
| vishy | so far i've tried this | 22:53 |
| vishy | http://paste.openstack.org/show/38266/ | 22:53 |
| arosen | vishy: if you go to your l3 agent node and run; ip netns ; it will disable all the namespaces; | 22:53 |
| vishy | hmm that might help | 22:53 |
| Apsu` | s/disable/display/ | 22:53 |
| arosen | vishy: i'd expect to see the 169.254 requests in your tcpdump | 22:54 |
| Apsu` | "ip netns" lists them. You can look at the interfaces/routes in the namespace with "ip netns exec <namespace> <command>" | 22:54 |
| vishy | arosen: yeah that would make sense | 22:54 |
| vishy | let me make sure it is leaving the node properly | 22:54 |
| *** neelashah has joined #openstack-dev | 22:55 | |
| vishy | arosen: i may have just missed it | 22:55 |
| vishy | arosen: the timeout is really freaking long | 22:55 |
| arosen | vishy: after the node comes it should be easer to debug :/ | 22:55 |
| arosen | Yea i know :( | 22:55 |
| *** giulivo has quit IRC | 22:55 | |
| *** anfrolov has quit IRC | 22:56 | |
| vishy | arosen: interesting it looks like the unicast bootp is failing | 22:57 |
| vishy | i wonder why that would be | 22:57 |
| vishy | 10.0.0.2 > 10.0.0.3: ICMP 10.0.0.2 udp port 68 unreachable, length 354 | 22:57 |
| arosen | i'm not sure. Does the instance have an ip when you login? (i'm guessing it does since you said you can ping it) | 22:57 |
| vishy | yeah the broadcast is fine | 22:57 |
| vishy | but when it tries to renew it sends a unicast and fails | 22:57 |
| vishy | arosen: ah i see why i don't get the 169 traffic | 22:58 |
| vishy | Request who-has 10.0.0.1 tell 10.0.0.2, length 28 | 22:58 |
| vishy | there is no arp-reply coming back | 22:58 |
| arosen | It can't talk to the gateway for some reason :/ | 22:58 |
| Apsu` | Did you setup a provider network? | 22:58 |
| vishy | well it is arping but the gateway isn't responding | 22:58 |
| vishy | yeah i have two networks | 22:58 |
| vishy | an external network and an internal network | 22:59 |
| vishy | the external is 10.10.10.10/24 | 22:59 |
| Apsu` | If the provider can't communicate between the compute host and wherever the L3 agent's router bridge is, that'd be your issue. | 22:59 |
| arosen | Do you have the quantum openvswitch agent running on the same box as the l3 agent? | 22:59 |
| *** darjeeling has joined #openstack-dev | 22:59 | |
| vishy | the internal is 10.0.0.0/24 | 22:59 |
| vishy | yeah | 22:59 |
| arosen | vishy: yup all that sounds fine. | 22:59 |
| Apsu` | Which means you'd need the OVS agent in both places, and the physical network between them would have to be setup appropriately. | 22:59 |
| vishy | the traffic is definitely getting to the l3 host | 22:59 |
| vishy | i can see the traffic | 22:59 |
| Apsu` | Alrighty | 22:59 |
| vishy | but there is no arp reply from the router | 23:00 |
| *** JonnyNomad has quit IRC | 23:00 | |
| arosen | can you paste the output of ovs-vsctl show ; on both boxes? | 23:00 |
| Apsu` | Would help to see.. yes, that. And also your provider network setup. net-show on the network and subnet-show on the subnet in question. | 23:00 |
| Apsu` | Should also verify the tap interface in the qrouter-* namespace is appropriate | 23:00 |
| vishy | i can ping the vm from qrouter ns | 23:01 |
| arosen | so you can ping in one direction and not the other? | 23:02 |
| *** samalba has joined #openstack-dev | 23:02 | |
| vishy | arosen: http://paste.openstack.org/show/38268/ | 23:02 |
| vishy | no i think i can probably ping just fine | 23:02 |
| vishy | i suspect i can't arp though | 23:03 |
| *** darjeeling has quit IRC | 23:03 | |
| *** neelashah has quit IRC | 23:04 | |
| *** jfriedly has joined #openstack-dev | 23:04 | |
| arosen | vishy: these aren't overlay networks are they? | 23:04 |
| vishy | arosen: no they are both physical | 23:04 |
| vishy | flat | 23:04 |
| *** jfriedly has left #openstack-dev | 23:04 | |
| *** afazekas has quit IRC | 23:05 | |
| arosen | vishy: i don't think you are going to be able to ping if you can't arp | 23:05 |
| vishy | arosen: perhaps not | 23:05 |
| arosen | could i also see a quantum port-list ? | 23:06 |
| Apsu` | vishy: There's different tags on those ports on either side... your provider may be physical, but it's tagging the ports | 23:06 |
| arosen | Apsu`: those tags aren't the tags that it's actually using. | 23:06 |
| arosen | the tags are in the flow table. | 23:06 |
| vishy | those are tags between the br-int and the phys right? | 23:06 |
| Apsu` | arosen: I realize they're not provider tags | 23:06 |
| vishy | so if i ping from controller to vm i can see the traffic on both sides | 23:07 |
| vishy | so i think the tagging is just fine | 23:07 |
| *** jf-jenni has joined #openstack-dev | 23:07 | |
| *** markvoelker has joined #openstack-dev | 23:07 | |
| *** jf-jenni has quit IRC | 23:08 | |
| *** jf-jenni has joined #openstack-dev | 23:08 | |
| *** gyee has joined #openstack-dev | 23:08 | |
| *** xBsd has quit IRC | 23:08 | |
| arosen | vishy: oops sorry, so you are using flat networking and you want metadata to work? | 23:08 |
| vishy | arosen: well i want metadata to work yes | 23:09 |
| vishy | did i miss a config option somewhere? | 23:09 |
| vishy | it might be faster to find an image without cloud init and boot it | 23:10 |
| arosen | vishy: i'm not sure if the metadata agent handles this case. I think you might have to run nova-api-metadata on the flat network to get that to work. | 23:10 |
| Apsu` | Flat provider + metadata agent works fine. So long as the metadata agent's node can hit the nova-api-metadata IP; doesn't have to be on the same flat network. | 23:11 |
| *** michchap has joined #openstack-dev | 23:11 | |
| arosen | Apsu`: I don't think the metadata agent creates an interface though on the physical node to catch these packets. | 23:11 |
| *** markvoelker has quit IRC | 23:12 | |
| arosen | to get the flat networking and metadata working i think you might need to have nova-api-metadata running on that node and the this iptables rule: | 23:12 |
| arosen | iptables -t nat -A PREROUTING -d 169.254.169.254 -p tcp -m tcp --dport 80 -j DNAT --to-destination 127.0.0.1:8775 | 23:12 |
| Apsu` | arosen: It does, I have it working. No custom iptables or routing required. | 23:13 |
| Apsu` | Can do it without the L3 agent, in fact | 23:13 |
| *** gordc has quit IRC | 23:14 | |
| Apsu` | You need namespaces enabled, and the metadata-agent will spawn the quantum-ns-metadata-proxy attached to 169.254.169.254/16, which is added to the namespaced tap device dnsmasq has an IP on as well. | 23:14 |
| Apsu` | At least in conjunction with dhcp-agent on a DHCP-enabled subnet. | 23:14 |
| Apsu` | (in the qdhcp-* namespace) | 23:14 |
| arosen | Apsu`: yes but what about in the case of using a provider network that is flat? | 23:14 |
| Apsu` | Exactly in that case is what I'm testing. VLAN works too. | 23:15 |
| vishy | Apsu`: wait you can do it without l3 agent? | 23:15 |
| *** otherwiseguy has quit IRC | 23:15 | |
| Apsu` | Yes. | 23:15 |
| Apsu` | vishy: Check your PM btw :P | 23:15 |
| vishy | Apsu`: can you run it on the host? | 23:15 |
| arosen | I'm confused why the metadata agent uses namespaces if running with vlans or flat networks. | 23:15 |
| Apsu` | Compute host? multi-host style? | 23:15 |
| *** eglynn_ has joined #openstack-dev | 23:15 | |
| Apsu` | arosen: Because it's not smart enough to combine metadata IPs and technically they overlap. The proxy is a namespaced proxy. | 23:16 |
| arosen | Apsu`: okay. So how do packets from one HV to the other enter the namespace? | 23:17 |
| arosen | Apsu`: you could be right. I've haven't really played around with the OVS plugin using VLANS +metadata. I've only tested it using overlay. | 23:18 |
| Apsu` | arosen: The simplest setup logically is: 1 network node with dhcp/metadata/ovs agents, 1 non-quantum management interface shared with controller (to hit nova-api-metadata on), and 1 provider interface on the same physical network as compute nodes. | 23:18 |
| *** tzumainn has quit IRC | 23:18 | |
| Apsu` | You make a separate provider bridge, and specify the mapping in the ovs plugin agent config | 23:18 |
| Apsu` | network_vlan_ranges = ph-eth1, bridge_mappings = ph-eth1:br-eth1 | 23:18 |
| Apsu` | Create the ovs bridge, add eth1 to it | 23:19 |
| Apsu` | Enable namespaces in the dhcp agent, and enable_isolated_metadata | 23:19 |
| Apsu` | quantum net-create --provider:physical_network=ph-eth1 --provider:network_type=flat blah | 23:20 |
| Apsu` | You have to cheat a little when you make the subnet so you'll get a valid metadata route into instances, because the agent logic is dumb around gateways | 23:20 |
| Apsu` | But it's very doable. | 23:20 |
| vishy | Apsu`: ah i don't have enable_isolated_metadata on | 23:21 |
| arosen | what does your subnet craete look like? | 23:21 |
| Apsu` | Ultimately you end up with a qdhcp-* namespace for your provider network, with a tap* in it that has the dnsmasq-attached IP for the subnet, and a 169.254.169.254/16 | 23:21 |
| Apsu` | subnet-create --name blah --no-gateway --host-route destination=0.0.0.0/0,nexthop=10.0.0.1 --allocation-pool start=10.0.0.2,end=10.0.0.254 blah 10.0.0.0/24 | 23:22 |
| Apsu` | The problem is... without an L3 agent, the dhcp-agent is entirely responsible for routes | 23:22 |
| arosen | Apsu`: Yes. That works because you are doing --no-gateway. If you have a gateway then. | 23:22 |
| Apsu` | arosen: No, look carefully at my host-route | 23:22 |
| Apsu` | A "gateway" or "default" route is just a 0.0.0.0/0 route. | 23:23 |
| Apsu` | I'm cheating by telling quantum there's no gateway (so the dhcp-agent injects the metadata route for me) | 23:23 |
| arosen | Apsu`: I see. That's a work around :) | 23:23 |
| Apsu` | But passing an additional static route (through DHCP option 121) that is actually a default :) | 23:23 |
| *** markwash has quit IRC | 23:23 | |
| vishy | Apsu`: a that is very interesting | 23:23 |
| arosen | yea i totally understand that. | 23:23 |
| Apsu` | So you get A) an external gateway default route, and B) a metadata route | 23:23 |
| *** michchap has quit IRC | 23:23 | |
| *** bswartz has joined #openstack-dev | 23:24 | |
| Apsu` | The L3 agent gives you some more options | 23:24 |
| Apsu` | But I dislike it because its failure modes are bad, and the current quantum scheduler is dumb :P | 23:24 |
| *** changbl has joined #openstack-dev | 23:24 | |
| *** pmathews has quit IRC | 23:24 | |
| arosen | Apsu`: That's funny that that works using the host route trick. Quantum should have a much better work around for that though. | 23:25 |
| *** matiu has joined #openstack-dev | 23:26 | |
| arosen | Apsu`: the reason why one wouldn't want to use host_routes like that is because that requires the dhcp client in the instance to support that. | 23:26 |
| Apsu` | arosen: I agree. And yeah, it does. Most do by default though. | 23:26 |
| Apsu` | 121 isn't an odd or crazy option. | 23:26 |
| arosen | Apsu`: If you use that iptables rule ; and run nova-api-metadata locally then you don't need to worry about that. | 23:26 |
| Apsu` | Sure. No need to proxy then. But that means more out-of-quantum mungery :P | 23:27 |
| *** kagan has quit IRC | 23:27 | |
| Apsu` | My goal was to solve it with as little mungery as possible, and push fixes upstream to make it less painful, hehe | 23:28 |
| vishy | arosen: is it safe to mix iptables and flows? | 23:28 |
| Apsu` | They don't interact at all. | 23:28 |
| Apsu` | OVS bridges don't hook into netfilter at all. | 23:29 |
| Apsu` | That's why if you want security groups, quantum creates a native bridge and hooks ports into it | 23:29 |
| Apsu` | (the qbr* connected with a qvo* and qvb* "veth pair" as they call it, into your br-int) | 23:29 |
| vishy | arosen: so i could steal the traffic using an iptables rule on the hosts | 23:29 |
| arosen | vishy: yea you could do that. Or using Apsu` work around. | 23:30 |
| *** cdub has quit IRC | 23:30 | |
| *** stevemar has quit IRC | 23:31 | |
| *** utlemming has quit IRC | 23:33 | |
| *** stevebaker has joined #openstack-dev | 23:34 | |
| *** stevebaker is now known as stevebaker_ | 23:34 | |
| *** andrewbogott is now known as andrewbogott_afk | 23:34 | |
| *** msmedved has quit IRC | 23:38 | |
| *** markvoelker has joined #openstack-dev | 23:38 | |
| *** salv-orlando has quit IRC | 23:39 | |
| *** kbrierly has quit IRC | 23:39 | |
| *** jasdeepH has quit IRC | 23:40 | |
| *** eglynn_ has quit IRC | 23:40 | |
| arosen | btw if using the overlay networking the metadata stuff works without that. | 23:40 |
| Apsu` | Can you elaborate on that a bit? | 23:41 |
| Apsu` | You mean non-provider nets? | 23:41 |
| arosen | yea non-provider | 23:41 |
| openstackgerrit | A change was merged to openstack/quantum: Imported Translations from Transifex https://review.openstack.org/31956 | 23:41 |
| vishy | Apsu`: ad router request: Subnet for router interface must have a gateway IP | 23:42 |
| vishy | :( | 23:42 |
| arosen | I don't think flat networks are very often used with quantum | 23:42 |
| arosen | we should have better metadata support for them though. | 23:42 |
| vishy | that is unfortunate | 23:42 |
| *** cmark_ has joined #openstack-dev | 23:42 | |
| *** markvoelker has quit IRC | 23:43 | |
| Apsu` | Well just injecting the metadata route always would solve it. | 23:43 |
| arosen | vishy: what's the big use case? I guess if you have legacy stuff? | 23:43 |
| vishy | arosen: i need to migrate nova-network to quantum | 23:43 |
| vishy | so i need to replicate the same config | 23:43 |
| arosen | I see. | 23:44 |
| Apsu` | Moreso when you want to directly connect to a provider. If your top-of-rack network doesn't expose an overlay endpoint, and you don't want to route all your traffic through the same machine... | 23:44 |
| *** cmark_ has quit IRC | 23:45 | |
| *** otherwiseguy has joined #openstack-dev | 23:45 | |
| *** cmark has joined #openstack-dev | 23:46 | |
| *** zzs has left #openstack-dev | 23:46 | |
| vishy | Apsu`: so the workaround doesn't work when you are actually using a qrouter | 23:47 |
| vishy | i think i just realized why the arp wasn't working though | 23:47 |
| Apsu` | vishy: Indeed. You have to do this differently for L3+metadata | 23:47 |
| *** dcasier_ has joined #openstack-dev | 23:47 | |
| vishy | i have another box with 10.0.0.1 | 23:47 |
| arosen | Apsu`: why do you have to do this differently for L3+metadata? | 23:47 |
| arosen | It seems you can use your same trick if you want. | 23:48 |
| Apsu` | arosen: https://lists.launchpad.net/openstack/msg22882.html | 23:48 |
| vishy | arosen: the reason is what i just pasted above | 23:48 |
| vishy | you can't connnect a subnet to a router that doesn't have a gateway | 23:48 |
| vishy | arosen: Bad router request: Subnet for router interface must have a gateway IP | 23:49 |
| Apsu` | Right. Not only that, but the gateway you specify... is the one the qrouter routes metadata to | 23:49 |
| *** michchap has joined #openstack-dev | 23:49 | |
| *** sushils has quit IRC | 23:49 | |
| *** eharney has quit IRC | 23:50 | |
| arosen | ah i see the api checks against that... | 23:50 |
| Apsu` | It makes your life hard. But you can make a metadata access network! | 23:51 |
| *** xBsd has joined #openstack-dev | 23:51 | |
| Apsu` | Attach your qrouter to it, and all other networks into that qrouter | 23:51 |
| Apsu` | And toggle the enable_metadata_network flag | 23:51 |
| Apsu` | Then it will have appropriate routes in both directions | 23:51 |
| Apsu` | But you're still routing all your traffic (on the network(s) that needs metadata) through that qrouter | 23:52 |
| vishy | Apsu`: well everything is going through a router right now anyway | 23:52 |
| Apsu` | Which I feel is also deficient compared to the straight provider network model. | 23:52 |
| Apsu` | vishy: But I chose not to use one :) | 23:52 |
| vishy | Apsu`: so in your straight provider network model | 23:53 |
| *** lloydde has quit IRC | 23:53 | |
| vishy | does your metadata proxy have to be reachable from the instances gateway? | 23:53 |
| Apsu` | Nope. | 23:53 |
| vishy | how does the metadata trafic get there then? | 23:53 |
| Apsu` | It will be, because dnsmasq is | 23:53 |
| Apsu` | They're all on the same network. | 23:54 |
| *** psedlak has quit IRC | 23:54 | |
| vishy | oh right so it is using dnsmasq | 23:54 |
| Apsu` | Yep. | 23:54 |
| vishy | yeah that seems much saner | 23:54 |
| vishy | i still can't get this thing to work with the router model | 23:54 |
| Apsu` | It lets every instance hit the upstream gateway directly, which Could be an ECMP route pool, or anycasting off your switching fabric, or whatever you want to do load-balancing and aggregation with | 23:55 |
| *** sushils has joined #openstack-dev | 23:55 | |
| Apsu` | Without having to build that into your network nodes or compute nodes. | 23:55 |
| vishy | Apsu`: does this work for tenant networks as well? | 23:55 |
| vishy | Apsu`: I assume so since they all get their own dnsmasq? | 23:55 |
| Apsu` | vishy: Yeah, think so. | 23:56 |
| vishy | ok will try that model again | 23:56 |
| * vishy goes to delete all of his networks :) | 23:56 | |
| Apsu` | hehe | 23:59 |
| arosen | vishy: Apsu` aright i'm out. Friiidayyy :) | 23:59 |
| arosen | have a good weekend | 23:59 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!