| *** miclorb has quit IRC | 00:00 | |
| pvo | Orman: made some changes to your doc | 00:02 |
|---|---|---|
| *** dendro-afk is now known as dendrobates | 00:03 | |
| *** sebastianstadil_ has joined #openstack | 00:05 | |
| *** sebastianstadil has quit IRC | 00:08 | |
| *** sebastianstadil_ is now known as sebastianstadil | 00:08 | |
| *** miclorb_ has joined #openstack | 00:08 | |
| *** gustavomzw has quit IRC | 00:09 | |
| *** dragondm has quit IRC | 00:09 | |
| *** aliguori has joined #openstack | 00:10 | |
| *** sparkycollier_ has joined #openstack | 00:11 | |
| *** sparkycollier has quit IRC | 00:13 | |
| *** sparkycollier_ is now known as sparkycollier | 00:13 | |
| *** Orman has joined #openstack | 00:20 | |
| Orman | Anyone on | 00:20 |
| devcamcar | few of us lurkers! | 00:20 |
| Orman | :) | 00:21 |
| Orman | devcamcar: So which dev are you? | 00:21 |
| Orman | Nova or Swift. | 00:22 |
| devcamcar | primarily nova, i'm on the nasa nebula dev team | 00:22 |
| *** dendrobates is now known as dendro-afk | 00:25 | |
| Orman | Really? | 00:25 |
| *** sparkycollier has quit IRC | 00:25 | |
| Orman | devcamcar: I would like to work for Nasa one day in Security. ;D | 00:26 |
| Orman | We have made a Nova Security Notes doc to outline our plans. | 00:26 |
| devcamcar | orman: awesome | 00:26 |
| Orman | devcamcar: Would you be interested in collab your knowledge of Nova to the notes? | 00:27 |
| devcamcar | orman: what do you have so far? | 00:27 |
| Orman | http://etherpad.openstack.org/NovaSecurityNotes | 00:28 |
| pvo | Orman: I made some notes. | 00:29 |
| pvo | Orman: are you coming to the summit? | 00:29 |
| Orman | pvo:Ok awesome | 00:29 |
| Orman | pvo: I wish,but I live in Florida. :( | 00:29 |
| devcamcar | orman: this is a great start | 00:29 |
| devcamcar | i'll be at the summit all next week | 00:30 |
| pvo | Orman: its only a few hours from FL. | 00:30 |
| pvo | : ) | 00:30 |
| alekibango | orman: i striked out one line, have really no sense | 00:30 |
| Orman | anne said there will be some security info talked about their. | 00:30 |
| Orman | alekibango: Ok I'll look. | 00:31 |
| Orman | pvo: I know,but I am still in school. ;) | 00:31 |
| pvo | Orman: likely so. I'd love to talk more to this in person. Some of it makes perfect sense, but others seem contradictory. | 00:31 |
| pvo | Orman: I'm in class right now. :p | 00:31 |
| Orman | Cool | 00:32 |
| Orman | devcamcar Feelfree to add new notes or code. ;) | 00:32 |
| alekibango | pvo: we can all meet using VOIP (mumble) | 00:32 |
| pvo | "Users should not be able to listen to packets coming in from the network" | 00:32 |
| pvo | this one is a bit confusing | 00:32 |
| alekibango | pvo please feel free to edit | 00:33 |
| alekibango | its just piece of something | 00:33 |
| alekibango | to lure people to actually help | 00:33 |
| pvo | well, I dont' mind editing, but I didn't want to go through ripping anything out without notes | 00:33 |
| alekibango | :) | 00:33 |
| Orman | That one line was mine pvo. | 00:33 |
| alekibango | pvo: there is history in worst case, dont be shy, edit | 00:33 |
| alekibango | :) | 00:33 |
| pvo | Orman: you might want to expand that one a bit more. I'm unsure of the intention. | 00:34 |
| alekibango | pvo: he means that user of one virtual guest should not be able to listen to what other VGs are sending over network | 00:34 |
| alekibango | iiuc | 00:34 |
| Orman | I mean just data being sent too one shouldn't be intercepted. | 00:34 |
| alekibango | Orman: you have there different classess of data | 00:35 |
| alekibango | you need to list them | 00:35 |
| Orman | Alekibango:Right I will do that when I get a sec. | 00:35 |
| alekibango | he could listen to public network maybe, but he for sure should not listen to communication with storage center etc | 00:35 |
| Orman | What we need is source code. :P | 00:35 |
| *** MarkAtwood has quit IRC | 00:35 | |
| pvo | https://blueprints.launchpad.net/nova/+spec/xs-host-networking-protections <- applies to XenServer, but the *-tables can apply to any *nix host. | 00:35 |
| alekibango | heh, right, and it differs for evry h-vizor | 00:36 |
| Orman | Good blueprint | 00:36 |
| alekibango | Orman: thats why we need to start with generic rules first | 00:36 |
| pvo | alekibango: not really. Unix hosts should have those 3 tools. | 00:36 |
| Orman | Of course | 00:36 |
| *** dabo has quit IRC | 00:36 | |
| pvo | Windows, yea, off the reservaation | 00:36 |
| Orman | I just excited sometimes. :P | 00:36 |
| alekibango | 3 tools? uh??? | 00:37 |
| pvo | ebtables, iptables, tables | 00:37 |
| pvo | arptables | 00:37 |
| Orman | Ah ok | 00:37 |
| Orman | ;) | 00:37 |
| alekibango | ah, i was not talking about them :) | 00:37 |
| alekibango | anyway, they should | 00:37 |
| pvo | you need to block at the bridge or just outside | 00:37 |
| alekibango | pvo: but thats technical talk | 00:37 |
| alekibango | we need to first identify goals and expectations | 00:38 |
| alekibango | then technical measures to achieve policy | 00:38 |
| *** schisamo has quit IRC | 00:38 | |
| pvo | alekibango: agreed | 00:38 |
| alekibango | great | 00:38 |
| alekibango | please help orman, i have still some other work to do now :) | 00:38 |
| Orman | Ok | 00:38 |
| *** daleolds has quit IRC | 00:38 | |
| Orman | Hmmm what other goals could we add. | 00:39 |
| pvo | secure comm with guest | 00:39 |
| Orman | ;) | 00:39 |
| * pvo wanders over to add it. | 00:39 | |
| alekibango | go for it! :) | 00:40 |
| alekibango | and secure virtual guest groups | 00:40 |
| alekibango | :) | 00:40 |
| Orman | Orman likes that idea! ;) | 00:40 |
| *** kevnfx has joined #openstack | 00:40 | |
| Orman | Permissions of guest groups | 00:42 |
| *** daleolds has joined #openstack | 00:43 | |
| Orman | Oh ahah I already have that one. ;) | 00:43 |
| *** ppetraki has joined #openstack | 00:43 | |
| Orman | Don't be shy guys edit it! | 00:44 |
| Orman | Want to be good. ;) | 00:44 |
| *** ppetraki has quit IRC | 00:45 | |
| *** jc_smith has quit IRC | 00:46 | |
| *** grizzletooth has left #openstack | 00:46 | |
| Orman | pvo: Did you add that one? I don't see it. | 00:47 |
| *** aliguori has quit IRC | 00:47 | |
| pvo | I just changed that line we were discussing a few min ago | 00:47 |
| pvo | The host should filter traffic to the guest to prevent the guest from spoofing traffic or sniffing traffic | 00:47 |
| pvo | sounds more like a "goal" | 00:47 |
| Orman | Ok thats I wanted know. | 00:47 |
| pvo | or maybe "the system shoudl" | 00:48 |
| Orman | Thinking | 00:48 |
| Orman | I guess the system should | 00:48 |
| pvo | yea, may not be the host doing the filtering | 00:49 |
| Orman | Right | 00:49 |
| Orman | The host would be protected from any threat of security from the guest. | 00:50 |
| Orman | Permissions | 00:50 |
| pvo | Orman: agree in principle, but that might fall to some of the hypervisor implementations. | 00:51 |
| *** abecc has quit IRC | 00:51 | |
| Orman | Yeah | 00:52 |
| Orman | I would like to make that universal in which the host could be protected no matter the situation. | 00:52 |
| pvo | I'm not exactly sure you can do that. | 00:52 |
| Orman | Just a thought. | 00:52 |
| Orman | We are doing well with outlining the goals and after that's done we can start on detailing them. ;) | 00:55 |
| Orman | Diagrams,code,and so forth. | 00:55 |
| *** miclorb_ has quit IRC | 00:56 | |
| *** littleidea has quit IRC | 01:01 | |
| *** adjohn has joined #openstack | 01:02 | |
| *** miclorb_ has joined #openstack | 01:03 | |
| *** miclorb_ has quit IRC | 01:05 | |
| *** littleidea has joined #openstack | 01:05 | |
| *** littleidea has quit IRC | 01:05 | |
| *** jdmaturen has quit IRC | 01:09 | |
| eday | pvo: it looks like a number of blueprints you added already had blueprints from other folks. we should probably try to consolidate them | 01:10 |
| pvo | yep, I figured that would be the case. | 01:12 |
| pvo | I was going to dig into that later. easier to throw out dupes. | 01:12 |
| pvo | some may be specific to xenserver while others might be more generic. | 01:13 |
| *** pvo has quit IRC | 01:15 | |
| *** adjohn has quit IRC | 01:18 | |
| *** mdomsch has quit IRC | 01:22 | |
| Orman | devcamcar: Good point with the Virtual guest one. ;) | 01:28 |
| *** metcalfc has quit IRC | 01:29 | |
| Orman | devcamcar: We will have to add some more things to it. | 01:29 |
| *** khussein has joined #openstack | 01:29 | |
| *** jbryce has quit IRC | 01:30 | |
| Orman | +eday: Feel free to edit the notes too. | 01:32 |
| *** daleolds has quit IRC | 01:32 | |
| *** Orman has quit IRC | 01:34 | |
| *** jdmaturen has joined #openstack | 01:37 | |
| *** MarkAtwood has joined #openstack | 01:37 | |
| *** Ryan_Lane has quit IRC | 01:51 | |
| *** Ryan_Lane has joined #openstack | 01:56 | |
| *** pvo has joined #openstack | 01:59 | |
| *** pvo has joined #openstack | 01:59 | |
| *** ChanServ sets mode: +v pvo | 01:59 | |
| *** jdmaturen has quit IRC | 02:05 | |
| pvo | eday: to your point about duplicate blueprints, some of them are generic in "inject files" ( i see we both had that one) but some of the ones I added are xenserver specific. | 02:07 |
| pvo | not sure if we want to make a distinction somewhere. | 02:07 |
| *** ambo has quit IRC | 02:13 | |
| *** ambo has joined #openstack | 02:17 | |
| *** Orman has joined #openstack | 02:20 | |
| Orman | Back online. :) | 02:21 |
| *** mmalgeri has quit IRC | 02:24 | |
| Orman | devcamcar: Thanks for adding some notes to our Nova Security Notes. :D | 02:27 |
| *** miclorb has joined #openstack | 02:27 | |
| eday | pvo: ahh, ok. we should probably try to make them generic (not just xs). Of course there is the qquestion of if a feature is "complete" if it only works for one of the supported hypervisors | 02:29 |
| pvo | yea, I was just talking to ewan about that. | 02:29 |
| pvo | do we have to implement a 'not implemented' feature in other hypervisors or... ? | 02:30 |
| pvo | I dunno | 02:30 |
| pvo | I figured that was good summit fodder | 02:30 |
| eday | yeah | 02:31 |
| * eday sees a feature matrix coming soon | 02:31 | |
| pvo | yea, I think thats going to be a necessity | 02:31 |
| *** schisamo has joined #openstack | 02:33 | |
| *** schisamo_ has joined #openstack | 02:35 | |
| *** schisamo has quit IRC | 02:38 | |
| *** schisamo_ is now known as schisamo | 02:38 | |
| *** mdomsch has joined #openstack | 02:41 | |
| *** arthurc has quit IRC | 02:44 | |
| Orman | Feature matrix ah yes!!! ;) | 02:45 |
| *** arthurc has joined #openstack | 02:47 | |
| *** jdmaturen has joined #openstack | 02:48 | |
| pvo | http://wiki.openstack.org/NovaFeatureMatrix | 02:56 |
| pvo | have at it. : ) | 02:56 |
| pvo | I'll add what I can later. Tonight I have to finish a paper. : / | 02:56 |
| *** jdmaturen has quit IRC | 02:57 | |
| Orman | Okay pvo | 03:01 |
| Orman | Thanks for your help with the security stuff today. | 03:02 |
| Orman | ;) | 03:02 |
| *** jdmaturen has joined #openstack | 03:04 | |
| *** jdmaturen has quit IRC | 03:08 | |
| pvo | np. Hope we can finish the doc. | 03:08 |
| *** blakeyeager has joined #openstack | 03:11 | |
| Orman | +pvo: I hope so too. | 03:15 |
| *** jdmaturen has joined #openstack | 03:16 | |
| Orman | I just can't wait to ee alekibango's diagram. | 03:16 |
| pvo | Orman: what is he diagramming? | 03:16 |
| Orman | He' drawing up a blueprint for the Nova Security Notes I believe. | 03:17 |
| Orman | Outlining some things. | 03:17 |
| pvo | ah | 03:17 |
| *** joearnold has joined #openstack | 03:18 | |
| Orman | Yep ;) | 03:18 |
| Orman | devcamcar made a great point in the notes about the Virtual Guest I wrote about. | 03:19 |
| pvo | which note? | 03:20 |
| Orman | The Nova Security Notes | 03:20 |
| Orman | http://etherpad.openstack.org/NovaSecurityNotes | 03:20 |
| pvo | right, I'm there, but I don't see his color. (purple) | 03:20 |
| Orman | Under the section Virtual Guest. | 03:21 |
| pvo | "Virtual Guest Security Permissions For The User" ? | 03:21 |
| Orman | Yeah | 03:22 |
| Orman | I messed up on that one. | 03:22 |
| pvo | I added the 2 comments in parenthesis | 03:22 |
| Orman | Oh ok it was you my bad. | 03:22 |
| Orman | lol I thought it was devcamcar. | 03:23 |
| Orman | Then I have you to thank. | 03:23 |
| pvo | no worries, just trying to see what he added. : ) | 03:23 |
| Orman | Ok | 03:23 |
| Orman | So with what you put in the second parentheses that's what I need to add. | 03:25 |
| Orman | They would need to be separate. | 03:25 |
| pvo | they don't have to be, but probably should be. Its less of an issue if you're not doing multi-tenent | 03:26 |
| Orman | However it makes sense though with what you added. | 03:26 |
| Orman | It's probably safer as well. ;) | 03:26 |
| Orman | If you can add more about that I'd love that! | 03:27 |
| pvo | I will if I have time. : ) | 03:27 |
| Orman | Ok | 03:27 |
| Orman | I don't mean to rush you. I am just glad you're on the doc with us. | 03:28 |
| Orman | The more people the better. ;) | 03:28 |
| pvo | well, I have a lot more to add, so I'll chop the doc up soonish. | 03:28 |
| Orman | Ok that sounds good because I have class tomorrow so I won't be on till later in the afternoon around 4:00ESt. | 03:30 |
| Orman | Just thought I'd tell you. | 03:30 |
| pvo | ha, won't be before then. | 03:30 |
| Orman | I know | 03:35 |
| Orman | I just wanted tell you | 03:35 |
| *** neogenix has joined #openstack | 03:38 | |
| *** kashyapc has joined #openstack | 03:41 | |
| *** Ryan_Lane has quit IRC | 03:47 | |
| *** Ryan_Lane has joined #openstack | 03:47 | |
| *** ArdRigh has quit IRC | 03:54 | |
| *** jc_smith has joined #openstack | 04:01 | |
| *** gaveen has joined #openstack | 04:02 | |
| *** littleidea has joined #openstack | 04:02 | |
| Orman | alekibango:You on | 04:03 |
| *** littleidea_ has joined #openstack | 04:04 | |
| *** littleidea has quit IRC | 04:07 | |
| *** littleidea_ is now known as littleidea | 04:07 | |
| *** AimanA is now known as HouseAway | 04:16 | |
| sebastianstadil | Where in the launchpad interface do you join Swift? | 04:16 |
| Orman | alekibango: I added some new things under the Virtual Guest Security Permissions. | 04:17 |
| Orman | sebastianstadil: Did you try this https://launchpad.net/~swift | 04:20 |
| *** Ryan_Lane has quit IRC | 04:21 | |
| Orman | +pvo: Check the notes when you get a chance because I elaborated on what you said in the parentheses. ;) | 04:26 |
| Orman | +pvo: Ok | 04:30 |
| *** ambo has quit IRC | 04:30 | |
| pvo | watching me type? : P | 04:30 |
| Orman | Yes | 04:30 |
| Orman | :P | 04:30 |
| Orman | You are pointing out some great ones. | 04:30 |
| sebastianstadil | Orman: Difference between https://launchpad.net/~swift and https://launchpad.net/swift is user / project? I can't any 'Join' on either. | 04:33 |
| *** ambo has joined #openstack | 04:33 | |
| Orman | Ok | 04:34 |
| *** omidhdl has joined #openstack | 04:39 | |
| *** omidhdl1 has joined #openstack | 04:42 | |
| *** omidhdl has quit IRC | 04:42 | |
| *** blakeyeager has quit IRC | 04:43 | |
| Orman | alekibango: you on? | 04:44 |
| Orman | +pvo did added some great pointers. | 04:44 |
| *** kashyapc has quit IRC | 04:45 | |
| sebastianstadil | Seems like the subscription policy is set to closed | 04:46 |
| Orman | :( | 04:49 |
| Orman | sebstainstadil: Are you a dev? | 04:50 |
| sebastianstadil | I am, yes | 04:50 |
| Orman | Do you know Nova at all? | 04:50 |
| sebastianstadil | But for this instance I am helping someone join | 04:51 |
| Orman | Ok | 04:51 |
| *** BK_man has joined #openstack | 04:51 | |
| *** BK_man has joined #openstack | 04:51 | |
| Orman | I am helping to write the Nova Security Notes if you're interested at all. | 04:51 |
| Orman | https://launchpad.net/~swift | 04:52 |
| Orman | ooops worng link haha | 04:52 |
| Orman | Here: http://etherpad.openstack.org/NovaSecurityNotes | 04:52 |
| sebastianstadil | I run a cloud computing group in the Bay area, and encourage members to participate in OpenStack - I was asked how to join and couldn't respond | 04:53 |
| Orman | Ok | 04:53 |
| Orman | Well if you have any security ideas or goals for Nova then please add them in the notes. | 04:54 |
| *** mdomsch has quit IRC | 04:55 | |
| *** alekibango has quit IRC | 04:58 | |
| Orman | Have a blessed night guys. I'm going to hit the hay.;) | 05:02 |
| Orman | See ya guys tomorrow. | 05:02 |
| *** kevnfx has quit IRC | 05:02 | |
| Orman | God bless and night. | 05:02 |
| sebastianstadil | Good night | 05:03 |
| Orman | night | 05:03 |
| *** Orman has quit IRC | 05:03 | |
| *** krish has joined #openstack | 05:13 | |
| *** schisamo has quit IRC | 05:16 | |
| *** krish has quit IRC | 05:20 | |
| *** joearnold has quit IRC | 05:23 | |
| *** krish has joined #openstack | 05:24 | |
| *** kashyapc has joined #openstack | 05:31 | |
| *** khussein has quit IRC | 05:34 | |
| *** kashyapc has quit IRC | 05:35 | |
| *** kashyapc has joined #openstack | 05:35 | |
| *** Dweezahr has quit IRC | 05:38 | |
| *** Dweezahr has joined #openstack | 05:39 | |
| *** MarkAtwood has quit IRC | 05:41 | |
| *** metcalfc has joined #openstack | 05:47 | |
| *** hazmat has quit IRC | 05:47 | |
| *** littleidea has quit IRC | 05:52 | |
| *** ramkrsna has joined #openstack | 05:54 | |
| *** ramkrsna has joined #openstack | 05:54 | |
| *** dmd17 has quit IRC | 05:56 | |
| *** dmd17 has joined #openstack | 05:56 | |
| *** zaitcev has quit IRC | 05:59 | |
| anticw | redbo: awake? | 06:04 |
| *** metcalfc has quit IRC | 06:08 | |
| *** alekibango has joined #openstack | 06:09 | |
| *** jc_smith has quit IRC | 06:10 | |
| *** pvo has quit IRC | 06:11 | |
| *** al-maisan has joined #openstack | 06:23 | |
| *** infernix has quit IRC | 06:32 | |
| *** infernix has joined #openstack | 06:33 | |
| *** ranjib has joined #openstack | 06:33 | |
| ranjib | hi all | 06:38 |
| *** stewart has quit IRC | 06:38 | |
| *** stewart has joined #openstack | 06:50 | |
| *** dmd17 has quit IRC | 06:51 | |
| *** dmd17 has joined #openstack | 06:52 | |
| *** stewart has quit IRC | 07:01 | |
| *** infernix has quit IRC | 07:02 | |
| *** infernix has joined #openstack | 07:11 | |
| *** miclorb has quit IRC | 07:11 | |
| *** alekibango has quit IRC | 07:13 | |
| *** stewart has joined #openstack | 07:14 | |
| *** alekibango has joined #openstack | 07:18 | |
| *** ramkrsna has quit IRC | 07:18 | |
| *** omidhdl1 has quit IRC | 07:22 | |
| *** omidhdl has joined #openstack | 07:23 | |
| *** alekibango has quit IRC | 08:04 | |
| *** termie has quit IRC | 08:12 | |
| *** termie has joined #openstack | 08:17 | |
| *** ramkrsna has joined #openstack | 08:18 | |
| *** alekibango has joined #openstack | 08:22 | |
| *** ibarrera has joined #openstack | 08:24 | |
| *** eldarnugaev has joined #openstack | 08:26 | |
| *** sebastianstadil has quit IRC | 08:28 | |
| *** krish has quit IRC | 08:33 | |
| *** sebastianstadil has joined #openstack | 08:41 | |
| *** krish has joined #openstack | 08:45 | |
| *** allsystemsarego has joined #openstack | 08:46 | |
| *** perestrelka has quit IRC | 08:47 | |
| *** eldarnugaev has quit IRC | 08:50 | |
| *** miclorb_ has joined #openstack | 08:58 | |
| *** dmd17 has quit IRC | 09:01 | |
| *** perestrelka has joined #openstack | 09:11 | |
| *** sebastianstadil_ has joined #openstack | 09:19 | |
| *** openfly has quit IRC | 09:20 | |
| *** sebastianstadil has quit IRC | 09:21 | |
| *** sebastianstadil_ is now known as sebastianstadil | 09:21 | |
| *** irahgel1 has joined #openstack | 09:25 | |
| *** sebastianstadil has quit IRC | 09:26 | |
| *** dizz|away has quit IRC | 09:38 | |
| *** ptremblett has quit IRC | 09:41 | |
| *** Ryan_Lane has joined #openstack | 09:55 | |
| *** f4m8_ is now known as f4m8 | 09:57 | |
| *** omidhdl1 has joined #openstack | 09:58 | |
| *** omidhdl has quit IRC | 09:58 | |
| *** ranjib has quit IRC | 10:01 | |
| *** Ryan_Lane has quit IRC | 10:04 | |
| *** ranjib has joined #openstack | 10:06 | |
| *** ambo has left #openstack | 10:08 | |
| *** ar1 has quit IRC | 10:14 | |
| *** ar1 has joined #openstack | 10:15 | |
| *** aimon has quit IRC | 10:18 | |
| *** aimon has joined #openstack | 10:18 | |
| *** ar1 has quit IRC | 10:20 | |
| *** aimon_ has joined #openstack | 10:23 | |
| *** miclorb_ has quit IRC | 10:25 | |
| *** aimon has quit IRC | 10:26 | |
| *** aimon_ is now known as aimon | 10:26 | |
| *** ptremblett has joined #openstack | 10:55 | |
| *** openstack_newbie has quit IRC | 11:05 | |
| *** omidhdl1 has left #openstack | 11:09 | |
| *** eldarnugaev has joined #openstack | 11:17 | |
| *** openfly has joined #openstack | 11:31 | |
| *** dendro-afk is now known as dendrobates | 11:35 | |
| *** ctennis has quit IRC | 11:37 | |
| *** ddumitriu has joined #openstack | 11:45 | |
| *** stewart has quit IRC | 11:48 | |
| *** ddumitriu has quit IRC | 11:50 | |
| *** stewart has joined #openstack | 11:55 | |
| *** dendrobates is now known as dendro-afk | 12:04 | |
| *** dendro-afk is now known as dendrobates | 12:04 | |
| *** ramkrsna has quit IRC | 12:04 | |
| *** dysinger has quit IRC | 12:09 | |
| *** littleidea has joined #openstack | 12:11 | |
| *** kashyapc has quit IRC | 12:14 | |
| *** littleidea has quit IRC | 12:19 | |
| *** littleidea has joined #openstack | 12:22 | |
| *** schisamo has joined #openstack | 12:23 | |
| *** littleidea_ has joined #openstack | 12:27 | |
| *** mdiver_ has joined #openstack | 12:27 | |
| *** ctennis has joined #openstack | 12:28 | |
| *** littleidea has quit IRC | 12:28 | |
| *** littleidea_ is now known as littleidea | 12:28 | |
| BK_man | ping | 12:31 |
| BK_man | anybody trying to run nova on several hosts? | 12:32 |
| BK_man | mine got problems on second host | 12:32 |
| BK_man | nova-compute running ok, but euca-describe-instances failed to connect to localhost:8773 | 12:33 |
| BK_man | any clue? | 12:33 |
| piken | BK_man: nova-compute or nova-manager? | 12:34 |
| BK_man | piken: nova-compute, according to the http://etherpad.openstack.org/ep/pad/view/NovaMultinodeInstall/latest | 12:34 |
| piken | BK_man: If your having connection issues to localhost:8773, you probally have the ec2_url set wrong on the node that is having the error. | 12:36 |
| piken | ec2_url needs to point to the cloud controller (nova-manager) | 12:37 |
| BK_man | piken: this is what I have: http://paste.openstack.org/show/89/ | 12:37 |
| BK_man | piken: this is from compute (2nd) node | 12:38 |
| piken | BK_man: hmm, something else is going wrong because you have url set right, but your error is stating it is trying to connect to localhost, not to 172.20.102.12 | 12:39 |
| mdiver_ | Trying to start nova-api , I've got an Error, " ImportError : cannot import name pidlockfile " | 12:39 |
| mdiver_ | any ideas ? | 12:39 |
| mdiver_ | Running on Centos5.5. | 12:39 |
| BK_man | Oh! My bad... nova/novarc was outdated. sorry! | 12:39 |
| *** MarkAtwood has joined #openstack | 12:40 | |
| piken | mdiver: did you install python-daemon? | 12:41 |
| *** MarkAtwood has quit IRC | 12:42 | |
| piken | your failure is from an import. | 12:42 |
| piken | from daemon import pidlockfile in server.py | 12:42 |
| piken | The docs for centos5.5 install are not that good as all the python libs installed are 2.4.3 not 2.6 that is needed. | 12:43 |
| *** westmaas1 has joined #openstack | 12:43 | |
| piken | you have to use easy-install2.6 to install the 2.6 version after you get 2.6 installed. Easiest to use epel to do that. | 12:43 |
| BK_man | Interesting... I can now see an instance running on first node from second node but can't reach it by network. What's wrong with my setup? | 12:43 |
| piken | BK_man: not sure there. Might be a routing or iptables issue. | 12:44 |
| *** kevnfx has joined #openstack | 12:44 | |
| *** mmalgeri has joined #openstack | 12:45 | |
| BK_man | piken: I do not even have a route to my VMs network on second node. What service is responsible for managing host's networking? | 12:47 |
| BK_man | (I meant what service in nova) | 12:47 |
| mdiver_ | piken: I've installed Daemon. | 12:47 |
| mdiver_ | The pidlockfile.py is there | 12:47 |
| piken | mdiver: easy test start an interactive python shell and type `from daemon import pidlockfile` | 12:48 |
| piken | BK_man: I am not sure. Let me ask the admin in ops that has been doing the setup while I have been focusing on dev work for it. | 12:48 |
| mdiver_ | piken: in python2.4 I can import.. not a problem... However, running python2.6 it gives me the error. | 12:50 |
| mdiver_ | pike: I've noticed that pidlockfile.py on 2.6 is a separated file | 12:50 |
| mdiver_ | piken: while in 2.4 it is a daemon.py function | 12:50 |
| *** Podilarius has joined #openstack | 12:52 | |
| mdiver_ | piken: I was able to import 'import lockfile' | 12:53 |
| *** pvo has joined #openstack | 12:54 | |
| *** ChanServ sets mode: +v pvo | 12:54 | |
| *** gustavomzw has joined #openstack | 12:54 | |
| *** pvo has quit IRC | 12:59 | |
| piken | mdiver_: when you used easy_install-2.6 to install daemon did you install daemon or python-daemon as they are different? | 13:00 |
| *** BK_man_ has joined #openstack | 13:01 | |
| mdiver_ | piken: I've installed both. | 13:03 |
| *** BK_man has quit IRC | 13:04 | |
| piken | If you installed both, you would not have the issue. | 13:04 |
| *** mdomsch_ has joined #openstack | 13:04 | |
| piken | from daemon import pidlockfile is an import from python-daemon | 13:05 |
| *** BK_man_ has quit IRC | 13:05 | |
| piken | if you easy_install-2.6 python-daemon, it would work with out a failure on the import. | 13:05 |
| mdiver_ | piken: I could import pidlockfile issuing => 'from lockfile import pidlockfile' | 13:08 |
| mdiver_ | I can find the file under /usr/lib/python2.6/site-packages/python_daemon-1.5.5-py2.6.egg/daemon/pidlockfile.py | 13:10 |
| *** eldarnugaev has quit IRC | 13:15 | |
| mdiver_ | piken: It is not supposed to be 'from lockfile import pidlockfile' instead of 'from daemon import pidlockfile' ??? | 13:16 |
| piken | no, it is from daemon | 13:16 |
| mdiver_ | piken: It can't find pidlockfile on daemon, just on lockfile... spooky.. | 13:17 |
| ranjib | mdiver: u there? | 13:19 |
| ranjib | mdiver_: i have recently documented the centos 5.5 installation .. its working smooth on my end.. i too faced the same issue. | 13:20 |
| ranjib | mdiver_: but after issuing easy_install-2.6 daemon, it was resolved | 13:20 |
| mdiver_ | ranjib: Hi, I'm here | 13:20 |
| ranjib | mdiver_:http://wiki.openstack.org/NovaInstall#Nova_installation_on_CentOS_5.5 | 13:22 |
| mdiver_ | ranjib: I can see at least 3 daemon under site-packages ./daemon , ./python_daemon, ./daemon-1.0-py2.6 , should I remove them ? | 13:22 |
| mdiver_ | ranjib: and re-install daemon only ? | 13:22 |
| ranjib | mdiver_:you should have python_daemon 1.5..5 only | 13:25 |
| *** mmalgeri has quit IRC | 13:26 | |
| ranjib | mdiver_:easy_install 2.6 daemon should do the job. but before that clean up rest | 13:26 |
| mdiver_ | ranjib: Python question... can I just delete them Up ? | 13:26 |
| *** theron has joined #openstack | 13:28 | |
| ranjib | mdiver_:yes | 13:28 |
| *** pvo has joined #openstack | 13:30 | |
| *** ChanServ sets mode: +v pvo | 13:30 | |
| ranjib | mdiver_:let me know if that fixes the problem | 13:32 |
| *** pvo has quit IRC | 13:34 | |
| mdiver_ | ranjib: I will. working on that now. | 13:34 |
| *** pvo has joined #openstack | 13:34 | |
| *** ChanServ sets mode: +v pvo | 13:34 | |
| mdiver_ | ranjib: I've removed all Daemons and removed from the easy-install.pth file all daemon references, issued the command as explained in the guide : easy-install-2.6 daemon | 13:39 |
| mdiver_ | ranjib: I can see now that under ./site-packages/daemon-1.0-py2.6.egg was installed | 13:39 |
| mdiver_ | ranjib: However, still not able to resolve pidlockfile | 13:40 |
| mdiver_ | ranjib: Still having the same error. | 13:40 |
| mdiver_ | ranjib: running Python prompt I can't either. | 13:41 |
| *** abecc has joined #openstack | 13:42 | |
| *** kevnfx has quit IRC | 13:43 | |
| *** abecc has joined #openstack | 13:43 | |
| ranjib | mdiver_: let me kickstart a centos and check things out at my end | 13:45 |
| *** mdomsch_ has quit IRC | 13:47 | |
| *** gaveen has quit IRC | 13:48 | |
| mdiver_ | ok | 13:49 |
| piken | the install is not easy-install-2.6 daemon | 13:56 |
| piken | It is easy-install-2.6 python-daemon | 13:56 |
| piken | They are different pypi packages and python-daemon is the required. | 13:57 |
| *** gondoi has joined #openstack | 13:57 | |
| mdiver_ | piken: I've tried this one easy-install-2.6 python-daemon , I didn't work... let me try again... | 13:57 |
| piken | you didn't do it if it did not work. I have tested it on 40+ nodes here that are all centos5.5 and it works without an issue. | 13:58 |
| *** gaveen has joined #openstack | 14:01 | |
| ranjib | mdiver_: hey, piken is right, its python-daemon | 14:01 |
| *** Ryan_Lane has joined #openstack | 14:03 | |
| *** irahgel1 has quit IRC | 14:04 | |
| *** blakeyeager has joined #openstack | 14:09 | |
| *** littleidea_ has joined #openstack | 14:14 | |
| *** littleidea has quit IRC | 14:17 | |
| *** littleidea_ is now known as littleidea | 14:17 | |
| *** metcalfc has joined #openstack | 14:19 | |
| *** irahgel1 has joined #openstack | 14:25 | |
| *** ppetraki has joined #openstack | 14:26 | |
| soren | anotherjesse: | 14:27 |
| soren | whoops | 14:27 |
| soren | anotherjesse: Did you see my comment on the whiteboard of https://blueprints.launchpad.net/nova/+spec/austin-puppet-deployment ? | 14:27 |
| *** Cybodog has joined #openstack | 14:32 | |
| *** johnpur has joined #openstack | 14:41 | |
| *** ChanServ sets mode: +v johnpur | 14:41 | |
| *** gustavomzw has quit IRC | 14:43 | |
| *** neogenix has quit IRC | 14:43 | |
| *** Ryan_Lane has quit IRC | 14:44 | |
| *** metcalfc has quit IRC | 14:44 | |
| piken | ranjib: mdiver_ is having issues due to easy_install on the box he was working on. It was importing some weird new version of python-daemon-1.6 when it is missing have the package. Then when using setuptools and building 1.5.5 directly, it would fail when he would do the import for an issue with the lockfile egg. | 14:47 |
| piken | Seems that system has something weird happening with pypi | 14:47 |
| *** Orman has joined #openstack | 14:51 | |
| *** kevnfx has joined #openstack | 14:53 | |
| ranjib | piken: strange . | 14:53 |
| *** f4m8 is now known as f4m8_ | 14:53 | |
| Orman | Hello | 14:55 |
| *** schisamo has quit IRC | 14:55 | |
| *** Orman has quit IRC | 14:59 | |
| *** grizzletooth has joined #openstack | 15:00 | |
| ranjib | i am forking vishy's nova.sh for centos,,, any contribution is welcome . | 15:02 |
| *** iammartian has left #openstack | 15:05 | |
| *** eldarnugaev has joined #openstack | 15:07 | |
| piken | I would suggest doing it all as one script with a param of the os to be added that way it can be extended to more in the future. | 15:07 |
| piken | ./nova.sh centos | 15:07 |
| piken | ./nova.sh ubuntu | 15:07 |
| piken | etc | 15:07 |
| dendrobates | yep | 15:08 |
| creiht | what about ./nova.sh gentoo? :) | 15:08 |
| piken | We might also think of not doing it as bash | 15:09 |
| grizzletooth | gentoo-ers wouldn't use it. they would just read the lines in the script and and try to run them each manually. | 15:09 |
| piken | Doing it as python using very minor standards. python 2.4 min with no outside modules will give us a little more flexability to the script | 15:09 |
| piken | grizzletooth: lol, good one | 15:09 |
| ranjib | piken: wont that make the script too long... | 15:11 |
| ranjib | piken: currenty i replace the services shebang line with env python2.6 too... | 15:11 |
| ranjib | piken:as i cant change the systems's default python 2.4 for its dependance on yum | 15:12 |
| *** iammartian has joined #openstack | 15:12 | |
| piken | ranjib: I think it would be cleaner. Have the script be really basic and know how to install items for each os (and be in the base python installed on the system). | 15:13 |
| piken | I would then have it use internals of python to first pull down a manifest from the nova repo that contains the deps for the system noted. | 15:14 |
| piken | MANIFEST.centos for example | 15:14 |
| ranjib | piken: good idea .. | 15:14 |
| piken | It would then loop through that to install rpms via yum, and use easy_install and such as needed. | 15:14 |
| piken | ranjib: I think it would make it clean to manage as each os would have a function set in the main on how to install for them, but the package naming and such would be in the manifest | 15:15 |
| ranjib | piken: how do we do the manual installations.. stuff like gflags binding? system commans? | 15:15 |
| *** Cybodog has quit IRC | 15:15 | |
| piken | ranjib: I would have a funciton for each as most will be the same on all systems. | 15:16 |
| piken | As for system commands, I would have a map in the script that associates commands with the system they are on, and use that map to determine what commands to run. | 15:16 |
| piken | I got the idea for this from the fusion-forge installer. https://fusionforge.org/scm/viewvc.php/*checkout*/trunk/src/fusionforge-install-1-deps.php?root=fusionforge&revision=10070&content-type=text%2Fplain&pathrev=10070 | 15:21 |
| mdiver_ | ranjib: For Centos how did you get the python-nova , nova-api, nova-objectstore, nova-compute ... etc ?? | 15:21 |
| piken | It does similar but in php except it has all the dependency names for each system internal. | 15:21 |
| ranjib | mdiver_: branched it from launchpad only .. | 15:22 |
| piken | bzr clone https://launchpad.net/nova or use the dl link on the side of https://launchpad.net/nova | 15:23 |
| piken | sry bzr branch https://launchpad.net/nova | 15:23 |
| ranjib | had to replace the shebang line.. i used sed for that, dont know if theres a sain way.. wish python has something like rvm | 15:23 |
| ranjib | *sane | 15:25 |
| *** gaveen has quit IRC | 15:25 | |
| piken | That is why I prefer writing installers in standards that use the installed version on the system no matter what. | 15:25 |
| piken | if you write in code that would work on 2.4 and 2.6 if it runs on /usr/bin/python or /usr/bin/python26 it doesn't matter. | 15:26 |
| *** gaveen has joined #openstack | 15:26 | |
| grizzletooth | use perl for the install script and you won't have to worry about it <grizzletooth ducks and runs for cover> | 15:27 |
| piken | LOL | 15:28 |
| ranjib | grizzletooth: :-) | 15:28 |
| ranjib | leaving aside readability.. even im more comfortable with perl.. | 15:29 |
| grizzletooth | I aspire to be a python coder, but am still stuck with too much perl experience ;) | 15:29 |
| * piken slaps grizzletooth with a peice of trout! | 15:30 | |
| grizzletooth | ouch - stinky | 15:30 |
| piken | lol | 15:30 |
| piken | I have done too much perl in my life, and too much php. | 15:30 |
| *** rds__ has joined #openstack | 15:30 | |
| piken | I actually have become fond of python. It is a very clean and very easy language to work with. | 15:30 |
| ranjib | i did loads of perl, a bit python, and now ruby .... | 15:31 |
| piken | Talk to me about perl when you have had to help recode a huge part of Bugzilla before they had a style guide. lol | 15:31 |
| ranjib | piken:python is full of boiler plates | 15:31 |
| piken | lol | 15:31 |
| piken | I am not a fan of ruby | 15:31 |
| ranjib | every time i see method(self) ,, i wonder what was wrong with perl then? | 15:32 |
| *** khussein has joined #openstack | 15:33 | |
| *** schisamo has joined #openstack | 15:34 | |
| piken | Only thing that pisses me off with python is != is not standard | 15:34 |
| piken | Stupid <> | 15:34 |
| notmyname | piken: what are you talking about? != is the normal way to do inequality checks in python | 15:35 |
| *** hazmat has joined #openstack | 15:36 | |
| *** comstud has joined #openstack | 15:38 | |
| *** ChanServ sets mode: +v comstud | 15:38 | |
| ranjib | i just like python's readability .. | 15:40 |
| ranjib | else gic, versions inconsistencies, package management.. everything sucks big time | 15:40 |
| piken | notmyname: one of the pep's pushes a standard use of <> over != | 15:40 |
| creiht | piken: http://docs.python.org/library/stdtypes.html#comparisons | 15:41 |
| creiht | <> is obsolete | 15:42 |
| piken | Hey, sue me. I haven't read the peps in a while. lol. Must have obsoleted it with 3.0 | 15:42 |
| redbo | also, I hate how python makes you end statements with semicolons | 15:43 |
| notmyname | just for fun: http://docs.python.org/release/2.0/ref/comparisons.html | 15:44 |
| notmyname | mentions <> is depreciated | 15:45 |
| creiht | hah | 15:45 |
| notmyname | piken: just giving you a hard time :-) | 15:45 |
| eday | vishy: are you guys still seeing any concurrency issues with eventlet? | 15:48 |
| *** spectorclan has joined #openstack | 15:50 | |
| *** metcalfc has joined #openstack | 15:51 | |
| *** metcalfc has joined #openstack | 15:52 | |
| *** al-maisan is now known as almaisan-away | 16:00 | |
| vishy | eday: I haven't been monitoring it recently | 16:01 |
| vishy | eday: we put HAProxy in front of 8 nova-apis | 16:01 |
| *** sebastianstadil has joined #openstack | 16:02 | |
| vishy | eday: last time i saw it, it appeared to be connecting to sql that was giving the error. I'm thinking using the thread local session in db/sqlalchemy/session.py might fix it, but i haven't been able to investigate. | 16:03 |
| eday | vishy: ahh, ok | 16:19 |
| *** mdiver_ has quit IRC | 16:29 | |
| *** mmalgeri has joined #openstack | 16:29 | |
| *** neogenix has joined #openstack | 16:29 | |
| *** eldarnugaev has quit IRC | 16:29 | |
| *** Dweezahr has quit IRC | 16:35 | |
| *** infernix has quit IRC | 16:36 | |
| *** ttx has quit IRC | 16:36 | |
| *** zykes- has quit IRC | 16:36 | |
| *** morfeas has quit IRC | 16:36 | |
| *** Xenith has quit IRC | 16:36 | |
| *** spike has quit IRC | 16:36 | |
| *** ttx has joined #openstack | 16:36 | |
| *** ttx has joined #openstack | 16:36 | |
| *** Xenith has joined #openstack | 16:36 | |
| *** morfeas has joined #openstack | 16:36 | |
| *** infernix has joined #openstack | 16:36 | |
| *** spike has joined #openstack | 16:36 | |
| *** joearnold has joined #openstack | 16:38 | |
| *** mdiver has joined #openstack | 16:40 | |
| *** joearnold has quit IRC | 16:41 | |
| ranjib | mdiver: u there> | 16:43 |
| ranjib | mdiver: u need to remove daeom and python-daemon, u should install python-daemon -1.5.5 | 16:46 |
| *** zykes- has joined #openstack | 16:46 | |
| *** Dweezahr has joined #openstack | 16:46 | |
| piken | ranjib: he did | 16:46 |
| piken | There looks to be a metaclass issue with lockfile currently for him | 16:47 |
| piken | he might need to use an older version of lockfile as it looks like the newest lockfile and python-daemon are fubar in pypi at the moment | 16:47 |
| piken | 0.9.0 and 0.9.1 have issues. | 16:48 |
| *** DubLo7 has joined #openstack | 16:49 | |
| *** dpackard has joined #openstack | 16:50 | |
| dpackard | hi all | 16:50 |
| dpackard | can anyone tell me if installing openstack nova makes changes to /etc/inetd.conf ? | 16:51 |
| ranjib | piken: really? how he installed the python_daemon? | 16:52 |
| ranjib | piken: i just did it using easy_install-2.6 python-daemon==1.5.5 , and now its working fine | 16:52 |
| piken | He installed it from the tarball | 16:53 |
| *** jdmaturen has quit IRC | 16:53 | |
| soren | dpackard: It doesn't. | 16:53 |
| piken | http://paste.openstack.org/show/90/ | 16:53 |
| dpackard | thanks soren | 16:54 |
| piken | ranjlib: there is the error the import gives and I have found some ubuntu bugs on google that point it at a lockfile 0.9.0 issue | 16:54 |
| *** littleidea has quit IRC | 16:58 | |
| *** jdmaturen has joined #openstack | 17:00 | |
| *** sebastianstadil has quit IRC | 17:02 | |
| *** sebastianstadil has joined #openstack | 17:05 | |
| *** ranjib has quit IRC | 17:06 | |
| *** neogenix has quit IRC | 17:08 | |
| elasticdog | is there not a swift.sh script to help set up the saio environment like the nova.sh script? | 17:17 |
| *** jdmaturen has quit IRC | 17:20 | |
| grizzletooth | elasticdog: not that I have seen - just the SAIO and multiserver docs http://swift.openstack.org/development_saio.html http://etherpad.openstack.org/MultiServerSwift | 17:21 |
| elasticdog | grizzletooth: thanks...I've just been going through the development_saio page and thought there might be a quicker way | 17:23 |
| *** DubLo7 has quit IRC | 17:23 | |
| *** joearnold has joined #openstack | 17:23 | |
| *** jdmaturen has joined #openstack | 17:24 | |
| notmyname | elasticdog: I think there is a chef script around somewhere that sets up a SAIO | 17:28 |
| creiht | I'm not sure if the chef script has been updated with the latest version though | 17:30 |
| elasticdog | notmyname: thanks | 17:31 |
| *** littleidea has joined #openstack | 17:31 | |
| *** joearnol_ has joined #openstack | 17:32 | |
| *** joearnold has quit IRC | 17:32 | |
| *** joearnold has joined #openstack | 17:32 | |
| *** sparkycollier has joined #openstack | 17:33 | |
| *** dendrobates is now known as dendro-afk | 17:35 | |
| *** grizzletooth has left #openstack | 17:36 | |
| *** kw1 has joined #openstack | 17:36 | |
| *** dendro-afk is now known as dendrobates | 17:37 | |
| mdiver | piken: I've downgrade the lockfile to 0.8.0 it seem to work now.. | 17:38 |
| *** kw1 has left #openstack | 17:39 | |
| *** sparkycollier has quit IRC | 17:41 | |
| piken | ranjib: There you go. you need python-daemon 1.5.5 and lockfile 0.8.0 as daemon 1.6 and lockfile 0.9.x are fubar. | 17:48 |
| vishy | soren: any idea if libvirt supports serial console output to two places at once? We're trying to have ajaxterm as well as sending the output to a file. | 17:48 |
| mdiver | It is now complaining about ./CA directory, I can see it on the source directory, Should I copy it over, or is there a way of specifying where the nova dir is ? any NOVA_HOME= ? Something like that ? | 17:49 |
| *** ibarrera has quit IRC | 17:50 | |
| *** littleidea has quit IRC | 17:50 | |
| *** joearnol_ has joined #openstack | 17:58 | |
| *** joearnold has quit IRC | 18:00 | |
| *** jc_smith has joined #openstack | 18:01 | |
| *** kevnfx has quit IRC | 18:01 | |
| vishy | mdiver: --ca_path=/path/to/ca | 18:02 |
| vishy | try running the binaries from inside the source tree | 18:02 |
| vishy | and it should find the existing one | 18:02 |
| vishy | cd /path/to/nova/ | 18:03 |
| vishy | ./bin/nova-xxxx | 18:03 |
| mdiver | vishy: got it. Thanks. | 18:03 |
| *** littleidea has joined #openstack | 18:04 | |
| jk0 | vishy: you ever see 'connection refused' when using sqlite and starting nova-compute with your install script? | 18:05 |
| jk0 | it's almost like the sqlite db is locked | 18:06 |
| *** joearnol_ has quit IRC | 18:07 | |
| *** joearnold has joined #openstack | 18:08 | |
| vishy | sounds likely | 18:08 |
| vishy | sqlite locking is pretty bad for multiple binaries | 18:08 |
| vishy | i usually use mysql | 18:08 |
| jk0 | that's what I figured | 18:09 |
| vishy | but i've usually only had to just rerun the command if it is locked | 18:09 |
| piken | have the openstack mailing lists been active? I did the signup for them and haven't seen anything come accross. I want to make sure work spamfilter is not eating it. | 18:09 |
| jk0 | I | 18:09 |
| spectorclan | piken which lists? | 18:09 |
| vishy | it spews errors on trying to update locked db | 18:09 |
| jk0 | I'm getting conn refused for compute, network and volume. | 18:09 |
| vishy | piken: there hasn't been anything recently | 18:09 |
| vishy | jk0 o really? | 18:10 |
| jk0 | there's no indication of locking but that's what I'm assuming | 18:10 |
| jk0 | yeah | 18:10 |
| vishy | jk0: if you restart them, do they work? | 18:10 |
| jk0 | nah, no go | 18:10 |
| vishy | jk0 sounds like a rabbit issue | 18:10 |
| jk0 | API and objectstore fire right up | 18:10 |
| vishy | connection refused is usually because rabbit isn't running | 18:10 |
| vishy | API and objectstore don't connect to rabbit so that would make sense | 18:11 |
| jk0 | good call, I'll look quick | 18:11 |
| *** khussein_ has joined #openstack | 18:13 | |
| jk0 | vishy: good catch - that was exactly it | 18:13 |
| vishy | jk0: np | 18:13 |
| *** dragondm has joined #openstack | 18:15 | |
| *** jed has joined #openstack | 18:15 | |
| *** khussein has quit IRC | 18:15 | |
| *** khussein_ is now known as khussein | 18:15 | |
| *** ptremblett has quit IRC | 18:24 | |
| *** jdmaturen has joined #openstack | 18:25 | |
| spectorclan | Reminder - Register for Design Summit at http://openstack.org/register | 18:25 |
| *** jdmaturen has joined #openstack | 18:25 | |
| *** Cybodog has joined #openstack | 18:28 | |
| *** dpackard has quit IRC | 18:33 | |
| *** jc_smith has quit IRC | 18:33 | |
| *** miclorb_ has joined #openstack | 18:36 | |
| *** jc_smith has joined #openstack | 18:37 | |
| *** dragondm has quit IRC | 18:41 | |
| vishy | dendrobates: are you here? | 18:46 |
| dendrobates | vishy: yep | 18:46 |
| zul | is the schedule up yet for the dev side of the summit yet? | 18:47 |
| vishy | dendrobates: what is the deal on the voting? Someone was saying that you have to be added to Nova or Swift in order to vote? | 18:48 |
| dendrobates | creiht: on this page: code.launchpad.net/nova which do you need sessions for? | 18:48 |
| vishy | in addition to signing the cla | 18:48 |
| creiht | dendrobates: what does the public want? :) | 18:49 |
| dendrobates | vishy: you needed to be on a member of any of the lp teams | 18:49 |
| vishy | dendrobates: and how does one get added to the teams? | 18:49 |
| dendrobates | vishy: it is too late, jonathon already grabbed the list. You need to ping Jonathon if someone needs to be added to the voting | 18:50 |
| creiht | dendrobates: I'm assuming you are asking about blueprints? | 18:50 |
| dendrobates | creiht: yes | 18:50 |
| vishy | dendrobates: ok | 18:50 |
| dendrobates | zul: working on it now | 18:50 |
| zul | dendrobates: cool | 18:51 |
| dendrobates | creiht: I assume you want to discuss all the ones marked future- at least? | 18:51 |
| creiht | It depends on if there is anything to discuss? :) | 18:51 |
| creiht | The only thing that I can think of that people would be really interested in discussing is the s3 api compatibility | 18:52 |
| dendrobates | if there is nothing, I would just use the time to go over your plans and see if anyone has any comments. | 18:52 |
| creiht | That's the problem, we don't have much in the way of "plans" | 18:52 |
| dendrobates | then this is the time to make them. | 18:53 |
| creiht | most everything for at least a while is going to be incremental improvements | 18:53 |
| creiht | but that's the thing, we have no major features to add | 18:53 |
| creiht | for the near term | 18:53 |
| devcamcar | creiht: how goes? i may have such a feature in mind | 18:53 |
| creiht | and nobody else has requested much else :) | 18:53 |
| creiht | devcamcar: !!!! | 18:53 |
| creiht | how's it going? :) | 18:53 |
| devcamcar | creiht: good! | 18:53 |
| devcamcar | i'll be in san antonio all next week | 18:54 |
| creiht | awesome | 18:54 |
| dendrobates | devcamcar: awesome create a blueprint and I'll schedule it. | 18:54 |
| devcamcar | one of the big features we've been discussing internally is multi-region support for swift | 18:54 |
| creiht | ok | 18:54 |
| *** Cybodog has quit IRC | 18:54 | |
| devcamcar | being able to store a replica in an external swift store | 18:54 |
| creiht | right | 18:54 |
| devcamcar | big win for us | 18:54 |
| creiht | add a blueprint! :) | 18:54 |
| devcamcar | i'm going to put a high level blue print up today | 18:54 |
| creiht | cool | 18:55 |
| creiht | dendrobates: the only other possibly interesting thing to discuss is the large file support, but that is mostly implemented | 18:55 |
| devcamcar | creiht: how large is a large file? | 18:55 |
| creiht | as large as you want it to be? | 18:56 |
| creiht | :) | 18:56 |
| devcamcar | one meeeeeeellion TB | 18:56 |
| creiht | If you can keep your connection open long enough to upload it, sure :) | 18:56 |
| devcamcar | creiht: how are you guys handling that under the cover? | 18:56 |
| devcamcar | are you chunking on client side? | 18:56 |
| creiht | the current implementation involves creating a manifest, and then chunking across the cluster | 18:57 |
| creiht | and it is done server side | 18:57 |
| *** dysinger has joined #openstack | 18:57 | |
| devcamcar | one of the questions we were asked recently is how easy it would be to saturate a 10G connection with a single large file being uploaded | 18:57 |
| creiht | heh | 18:57 |
| devcamcar | a lot of our potential users have a small number of huge files | 18:57 |
| creiht | hrm | 18:57 |
| devcamcar | swift is basically bound by spindle speed more than network speed | 18:58 |
| creiht | correct | 18:58 |
| devcamcar | s3 is same way | 18:58 |
| devcamcar | only way around that is chunking client side, it gets messy | 18:58 |
| creiht | There is an idea down the road to expose a similar chunking to the client side | 18:58 |
| devcamcar | anyway, just curious if you guys are thinking about it from that point of view | 18:58 |
| creiht | where the client would upload the manifest, and chunks | 18:59 |
| devcamcar | that would have huge value for our customers | 18:59 |
| creiht | that way you could upload those in parallel | 18:59 |
| creiht | It has its own set of issues :) | 18:59 |
| devcamcar | we have 10G everywhere now, though we're already talking about 40G or 100G pipes | 18:59 |
| devcamcar | creiht: indeed it does! | 18:59 |
| creiht | devcamcar: we can talk about it then next week | 18:59 |
| devcamcar | creiht: worth a blueprint? | 19:00 |
| creiht | sure, why not? :) | 19:00 |
| *** sebastianstadil has quit IRC | 19:01 | |
| *** khussein has quit IRC | 19:05 | |
| spectorclan | If anyone needs help with hotels next week, let me know and I can try and find you a roomate | 19:06 |
| *** littleidea has quit IRC | 19:10 | |
| devcamcar | creiht: high level rough of multi region: https://blueprints.launchpad.net/swift/+spec/multi-region | 19:15 |
| _0x44 | jaypipes: Are you about? | 19:15 |
| jaypipes | _0x44: sure am. | 19:16 |
| _0x44 | Is there a way to make nose report on cyclomatic complexity the way it can report on code coverage? | 19:18 |
| jaypipes | _0x44: not that I know of. | 19:18 |
| jaypipes | _0x44: but isn't there a way to ask pylint to do that? eday? | 19:19 |
| _0x44 | Great, thanks :) | 19:19 |
| creiht | devcamcar: so what you are talking about in the proposal can already be done client side today | 19:20 |
| alekibango | dendrobates: what do you think about my blueprint? https://blueprints.launchpad.net/nova/+spec/reserve-and-limit-resources... | 19:20 |
| creiht | a small change to the auth would allow it to return several endpoint urls, to which the client could send data to either (or both) | 19:20 |
| devcamcar | creiht: that is a great feature but want to see swift dictate that automatically on the server side if configured to do so | 19:21 |
| creiht | ok | 19:21 |
| creiht | that wasn't entirely clear from the description | 19:21 |
| devcamcar | creiht: ok i will make that more explicit | 19:21 |
| creiht | cool... and should be interesting to discuss next week :) | 19:22 |
| devcamcar | indeed | 19:22 |
| devcamcar | creiht: in short: swift installation A is configured to be aware of B, when data is uploaded to A, N number of copies are also replicated to B | 19:24 |
| devcamcar | and vice versa | 19:24 |
| creiht | yeah | 19:24 |
| devcamcar | creiht: and bonus points for making it configurable by container :) | 19:24 |
| creiht | hah | 19:25 |
| creiht | container stored procedures! :) | 19:25 |
| devcamcar | oo | 19:25 |
| _0x44 | jaypipes: It doesn't look like there's a way to make pylint do it, I'll have to install another module. Thanks for the help | 19:25 |
| jaypipes | _0x44: selenium, then? | 19:26 |
| *** khussein has joined #openstack | 19:28 | |
| jaypipes | _0x44: http://www.traceback.org/2008/03/31/measuring-cyclomatic-complexity-of-python-code/ | 19:28 |
| dendrobates | alekibango: interesting, there is some overlap with blueprints that pvo is working on. | 19:30 |
| alekibango | do you think we can plan that for bxar? | 19:30 |
| _0x44 | jaypipes: I found that earlier, but I hadn't yet installed it because I was hoping for a nose plugin like complexity.py | 19:30 |
| jaypipes | _0x44: ah, gotcha. | 19:31 |
| alekibango | pvo: please what is dendrobates talking about, can you tell me more? | 19:32 |
| alekibango | dendrobates: not everyone wants to use shared resources :) there are customers who want reserved ones... and there are customers who would like to have mix or reserved and mix of shared | 19:33 |
| alekibango | eh, meant mix of reserved and shared :) | 19:34 |
| alekibango | ie, 10MBps reserved bandwidth and sharing rest with other users of the host | 19:35 |
| dendrobates | alekibango: it seems to me that throttling with hard and soft limits can accomplish what you are asking for. | 19:35 |
| alekibango | dendrobates: i know its not exactly easy. bug kvm, lvm or UML can be easily limited by normal means | 19:36 |
| dendrobates | few users are willing to pay for resources they are not using currently. That is one of the advantages of utility computing | 19:36 |
| piken | I would stress that the limits be very flexable as ip over ib with kvm has seen 6.6GBps through put for our vms and we would not want to limit our selves out of the performance. | 19:36 |
| alekibango | s/bug/but/ sorry, myfingers are weak today | 19:37 |
| alekibango | dendrobates: yes... do you think we can target this for bexar? | 19:37 |
| alekibango | i will help with it | 19:37 |
| *** Cybodog has joined #openstack | 19:37 | |
| dendrobates | alekibango: why reserve resources? why do you care, as long as they are there when you try to use them? | 19:38 |
| _0x44 | alekibango: Can it wait three months to whatever the C release is? | 19:38 |
| alekibango | well, to be able to sell virtual host having certain parameters | 19:38 |
| alekibango | certain SLA | 19:38 |
| alekibango | predictable | 19:38 |
| dendrobates | your electric company does not reserve power for you. | 19:38 |
| _0x44 | dendrobates: If you ask Blake and Opie they'll tell you lots of customers want explicitly specified hard limits | 19:38 |
| alekibango | not more, not less | 19:38 |
| dendrobates | alekibango: we can do that without reservations | 19:38 |
| alekibango | dendrobates: yes and suddenly power fails when my neigbour uses his tesla inspired tools | 19:39 |
| dendrobates | _0x44: I agree with the concept of limits, just not reservations | 19:39 |
| *** khussein has quit IRC | 19:39 | |
| _0x44 | dendrobates: pm coming, I'm not sure I'm allowed to talk about what I'm going to say next publicly. | 19:39 |
| piken | dendrobates: a pure qouta system would the preferred as you can limit a user, but not restrict them if needed. | 19:39 |
| alekibango | dendrobates: i understand this will require reserved hosts... | 19:39 |
| alekibango | dendrobates: this comes as request from my customer, a hosting company | 19:40 |
| alekibango | and he has tons of customers who need this :) | 19:40 |
| alekibango | i know its not using the hardware to the max, but it gives some sort of good feeling to customers | 19:41 |
| alekibango | dendrobates: reservation of that is just change in scheduler | 19:42 |
| alekibango | limiting is for nova-manage | 19:42 |
| alekibango | right? | 19:42 |
| alekibango | if someone will guide me a bit, i can work on it | 19:43 |
| pvo | alekibango: whats the difference in reservation and provisioning a vm and not using it? | 19:43 |
| dendrobates | what gives me a good feeling is that when I flip on a light, it comes on. I don't care if the power company reserved power for me. | 19:43 |
| alekibango | reservations means that when other users will abuse resources of the host, you will not be affected | 19:43 |
| pvo | alekibango: thats not reservations | 19:44 |
| pvo | alekibango: those are resource limits | 19:44 |
| alekibango | yes, and planning | 19:44 |
| alekibango | pvo: maybe i have wrong terminology :) | 19:44 |
| dendrobates | I realize that some people may want this, but I think they are missing the point of the cloud | 19:44 |
| piken | alekibango: No hosting company could live at that as the key to the hosting business is over provision. | 19:44 |
| pvo | alekibango: thats just how you configure your hypervisor | 19:44 |
| piken | I worked for 5 years for one of the worlds largest and know that to be the truth | 19:44 |
| alekibango | i want limits and scheduling -- and the scheduler should respect thole limits, so it will not schedule 150% of load on 1 cpu | 19:44 |
| *** guynaor has joined #openstack | 19:45 | |
| pvo | alekibango: scheduler is pluggable. You can write your own to take this into account already | 19:45 |
| *** Cybodog has quit IRC | 19:45 | |
| alekibango | yes | 19:45 |
| *** Cybodog has joined #openstack | 19:46 | |
| piken | alekibango: at that point your not doing any better then buying single core servers that are bare metal. KVM and Xen do a good job of resource swaping for hosts using more resources taking from those that are not for a reason. | 19:46 |
| alekibango | pvo: what do you mean - thats just how you configure hypervisor? | 19:46 |
| *** ptremblett has joined #openstack | 19:46 | |
| pvo | you can configure your hypervisor to burst or have hard caps | 19:46 |
| piken | alekibango: a quota system is a must, but reserving alotments of resources is against the idea of KVM and Xen. | 19:47 |
| piken | pvo: you can set hard caps, but not soft caps. | 19:47 |
| alekibango | piken: i understand you, but there is demand for this :) | 19:47 |
| alekibango | really | 19:47 |
| pvo | piken: hard caps, yes | 19:47 |
| alekibango | piken: maybe only partial reservation | 19:47 |
| dendrobates | we need to throttle network i/o, disk i/o and CPU. pvo: do your blueprints cover throttling? | 19:47 |
| *** Cybodog has quit IRC | 19:47 | |
| pvo | alekibango: reservation is the same as provisioning, no? | 19:48 |
| piken | alekibango: That is a demand that you will kill the performance of the cloud as a whole to meet. | 19:48 |
| alekibango | i want to be able to tell hard limits -- not only max, but also MIN | 19:48 |
| pvo | dendrobates: isn't there already throttling in openstack api? | 19:48 |
| pvo | didn't gundlach put that in? | 19:48 |
| *** Cybodog has joined #openstack | 19:48 | |
| alekibango | piken: i know, but it might give you more money | 19:48 |
| alekibango | its about SLA | 19:48 |
| dendrobates | pvo: on the hosts, using xs? | 19:48 |
| alekibango | and people willing to pay | 19:48 |
| pvo | throttling at the pai | 19:48 |
| pvo | api | 19:48 |
| alekibango | pvo: you are working on throttling? | 19:49 |
| alekibango | on limits? | 19:49 |
| pvo | there was work done on it. | 19:49 |
| piken | alekibango: actually it is not, it is against SLA believe it or not. Ran into this head on at HostWay when doing OpenVZ vps's | 19:49 |
| dendrobates | pvo: I'm talking about backend, not API. | 19:49 |
| pvo | dendrobates: not sure what you mean then. | 19:49 |
| dendrobates | pvo: do your spec cover hard and soft disk i/o limits and network i/o limit to prevent one vm from affecting it's neighbors? | 19:50 |
| *** irahgel1 has left #openstack | 19:51 | |
| pvo | not disk io, but we do cover network io | 19:51 |
| alekibango | piken: i want to be able to tell: you will have 30-50 MBps available | 19:51 |
| _0x44 | piken: OpenVZ lies about virtualization | 19:51 |
| alekibango | not only you can be limited to max 50, min 0 | 19:51 |
| dendrobates | I know, we can do disk i/o with KVM and hopefully soon with xs too. | 19:52 |
| piken | _0x44: at the time it was the best you can find. lol | 19:52 |
| alekibango | or you will get shared 200 | 19:52 |
| *** dragondm has joined #openstack | 19:52 | |
| dendrobates | pvo: but it's in your specs? I'm trying to see if we need another. | 19:52 |
| alekibango | piken: again, this is what this hosting company needs, and its one of largests in my country | 19:52 |
| alekibango | they have companies waiting to pay for this limited hw :) | 19:53 |
| piken | alekibango: I understand what you are saying, but from the idea of a cloud, that would be a step in the wrong direction for the core project. Maybe for an extension for you. But I would highly advise keeping that from the core of nova. | 19:53 |
| alekibango | we have also some cheap hosting companies, who use sharing to the max | 19:53 |
| alekibango | but thats not the way of my customer | 19:53 |
| piken | alekibango: maybe you should focus on vm migration priorities instead of reservations. The best solution would be if a node cannot offer the load a user is using, move them and keep it to hard limits only. | 19:54 |
| alekibango | piken: right | 19:54 |
| alekibango | that would be the one | 19:54 |
| alekibango | sheepdog + live instant migration | 19:55 |
| alekibango | or somthing like this | 19:55 |
| piken | That is live able, but needs a lot as there is no live migration in nova at the moment. | 19:55 |
| piken | That even opens up the option of node load balancing like vsphere. | 19:55 |
| alekibango | when do you think we can make migration real? | 19:55 |
| alekibango | piken: right... | 19:55 |
| *** Cybodog has quit IRC | 19:56 | |
| piken | Don't look at me. lol. I am still working out some blueprinting info and work for the VolumeManger Vishy started so I can use it internal here for VirtFS + Lustre over infiniband. | 19:56 |
| alekibango | my other blueprint might be to move instantions in a way they best use hardware in time (for example shools have different needs than shops or news) | 19:56 |
| alekibango | but thats for far future maybe :) | 19:57 |
| alekibango | and live migration would solve this also | 19:57 |
| alekibango | i think, live migration would be good | 19:58 |
| *** Cybodog has joined #openstack | 19:58 | |
| alekibango | + those limits | 19:58 |
| dendrobates | alekibango: there is a live migration spec | 19:58 |
| *** Orman has joined #openstack | 19:58 | |
| Orman | Hello | 19:58 |
| alekibango | yes ty | 19:58 |
| piken | Hey Orman | 19:59 |
| alekibango | Orman: hello. i am not doing security diagrams yet :) i just have beed drawing some nova arch pictures | 19:59 |
| devcamcar | piken: you're implementing a VolumeManager for VirtFS and Lustre? | 19:59 |
| alekibango | which talk about AMQP centricity | 19:59 |
| piken | devcamcar: I pulled a branch of Vishy's VolumeManager and I am working out a Driver for VirtFS. | 20:00 |
| Orman | Ok | 20:00 |
| Orman | Hey piken. | 20:00 |
| devcamcar | piken: neat, though i personally feel an urge to rub myself down with steel wool whenever i deal with lustre :) | 20:00 |
| piken | Not lustre directly as we are mounting Lustre on the physical node and going to virtfs to the directory as Lustre over 40gbps infiniband is faster then lustre over 6.6gbps max with virtio-net | 20:01 |
| *** neogenix has joined #openstack | 20:01 | |
| alekibango | and who is working on nova limits/throtling? :) | 20:01 |
| devcamcar | piken: interesting | 20:01 |
| alekibango | piken: what do you think about ceph? | 20:02 |
| _0x44 | alekibango: For Austin it was comstud | 20:02 |
| devcamcar | creiht: blue print for client side chunks, mainly just a starting point for discussion | 20:02 |
| devcamcar | https://blueprints.launchpad.net/swift/+spec/client-side-chunking | 20:03 |
| alekibango | _0x44: ty | 20:03 |
| devcamcar | very very high level | 20:03 |
| Orman | alekibango: When will you have your architecture pictures up? | 20:03 |
| piken | alekibango: it is nice, if you like loosing your data every other day and seeing big red warnings that say "WARNING: THIS IS NOT PRODUCTION READY!!!" | 20:03 |
| _0x44 | alekibango: You're welcome | 20:03 |
| alekibango | piken: :)) | 20:03 |
| alekibango | piken: i didnt have the courage to try | 20:03 |
| alekibango | but it looks promising | 20:03 |
| piken | We have tried a lot here, Ceph + BTRFS, GlusterFS, and Lustre. Lustre is the most promising and stable as long as you have the money to through at top of the line hardware. | 20:04 |
| alekibango | and glusterfs? | 20:04 |
| alekibango | its coming to be nice, but i found it slow | 20:04 |
| piken | It is slow and if you have more then 2 storage nodes, the data replication becomes a huge bottle neck. | 20:04 |
| piken | I think a test of Lustre on our SAN and infiniband the other day was showing 50GBps reads. | 20:05 |
| piken | ah, nvm. 50GBps writes at a sustained of 6.3GBps per client. | 20:06 |
| alekibango | so, what do we need to make live migration real (at least for kvm)? | 20:06 |
| piken | Lots of Prayer? | 20:06 |
| piken | lol | 20:06 |
| alekibango | :) | 20:06 |
| alekibango | btw lxc containers might make sense even for kvm/qemu/UML instances --> to limit | 20:07 |
| *** ctennis has quit IRC | 20:07 | |
| alekibango | i agree limits are enough for me if we will get live migration working during winter | 20:08 |
| Orman | devcamcar: Nice blueprint. ;) | 20:08 |
| Orman | I am trying to figure out which blueprint I should do. | 20:09 |
| *** khussein has joined #openstack | 20:09 | |
| Orman | I need to add ore to clarify what I mean in the first paragraph of tje Virtual Guest. | 20:10 |
| Orman | More details would work. ;) | 20:11 |
| alekibango | we need diagrams | 20:11 |
| alekibango | lots of diagrams | 20:11 |
| alekibango | :) | 20:11 |
| alekibango | pictures | 20:11 |
| alekibango | explained | 20:11 |
| Orman | Ok | 20:11 |
| Orman | I'll make a Virtual Guest diagrams when I get the chance | 20:12 |
| Orman | Both the guest and the host shold be on seperate layers. | 20:12 |
| alekibango | piken: btw how do you combat silent data corruption problem ? :) | 20:13 |
| alekibango | meeting is in 45 minutes? | 20:14 |
| Orman | What meeting? | 20:14 |
| alekibango | Orman: #openstack-meeting | 20:14 |
| Orman | Ah I see | 20:14 |
| alekibango | see top of wiki pages | 20:15 |
| Orman | Well I am going to head out for a bit here in aminute to vote. | 20:15 |
| piken | alekibango: Never solved it in 1.8. We are trying 2.0 and it solves it via embedded lvm in the partition manager for lustre now to do striping better. | 20:15 |
| Orman | I might be back in time for the meeting | 20:15 |
| alekibango | piken: interesting :) | 20:15 |
| soren | vishy: Same serial port pointed to two places? | 20:17 |
| soren | vishy: I don't believe it does, no. | 20:18 |
| soren | vishy: Let me check. | 20:19 |
| soren | vishy: It doesn't, no. | 20:21 |
| soren | vishy: It shouldn't be too difficulat to add, though. | 20:21 |
| *** ctennis has joined #openstack | 20:21 | |
| *** ctennis has joined #openstack | 20:21 | |
| soren | vishy: What does ajaxterm need? | 20:21 |
| Orman | see ya guys | 20:22 |
| Orman | later | 20:22 |
| Orman | gtg | 20:22 |
| *** Orman has quit IRC | 20:22 | |
| piken | Who do I have to talk to about joining the Nova and Swift teams on LP? | 20:24 |
| creiht | piken: I think you just click the "join" button :) | 20:25 |
| piken | or the openstack team as a whole. I want to be active on the mailing lists and it turns out the reason I cannot access them is that. lol | 20:25 |
| creiht | swift-core and nova-core are a different thing | 20:25 |
| piken | hmm, don't see join link anywhere on lp | 20:26 |
| creiht | hrm | 20:26 |
| creiht | piken: try https://launchpad.net/~swift | 20:27 |
| creiht | and https://launchpad.net/~nova | 20:27 |
| dendrobates | creiht: the lp teams are locked while we revamp our CLA policy. | 20:27 |
| creiht | oh | 20:28 |
| creiht | heh | 20:28 |
| piken | Bah, make it hard to join the mailing lists. lol | 20:28 |
| creiht | sorry | 20:28 |
| dendrobates | if anyone wants to join and they have signed the CLA, I'll add them | 20:28 |
| dendrobates | they will be unlocked after the summit next week | 20:28 |
| piken | I can sign it for my personal contributions, but they won't include anything from my company as we are waiting on legal for those. | 20:29 |
| _0x44 | dendrobates: I signed the CLA twice and have to addresses in two countries, can my vote count twice? | 20:29 |
| _0x44 | s/to/two/ | 20:29 |
| dendrobates | _0x44: I am not involved in the voting | 20:29 |
| _0x44 | dendrobates: Should I ask jbryce? | 20:30 |
| dendrobates | release meeting in 30 min | 20:30 |
| dendrobates | _0x44: did you get two emails with links? | 20:30 |
| *** littleidea has joined #openstack | 20:30 | |
| soren | vishy: So, afaict, ajaxterm just exec()'s something and uses stdio for communication. It shouldn't be much work to make libvirt export the serial port to a unix socket and have a daemon that writes everything to the console log as well as allow ajaxterm to connect through it. | 20:31 |
| *** metoikos has joined #openstack | 20:32 | |
| *** allsystemsarego has quit IRC | 20:33 | |
| *** Cybodog has quit IRC | 20:34 | |
| *** sparkycollier has joined #openstack | 20:36 | |
| vishy | soren: I thought we might have to do something like that. | 20:36 |
| *** Cybodog has joined #openstack | 20:37 | |
| creiht | mtaylor: would it be possible to set up hudson on the swift 1.1 series? | 20:40 |
| creiht | and tarmac | 20:40 |
| *** sebastianstadil has joined #openstack | 20:42 | |
| *** BK_man has joined #openstack | 20:43 | |
| _0x44 | dendrobates: I didn't get two emails with links, I'm only in launchpad once :) | 20:43 |
| *** devcamcar has quit IRC | 20:45 | |
| *** anotherjesse has quit IRC | 20:45 | |
| *** xtoddx has quit IRC | 20:45 | |
| *** vishy has quit IRC | 20:45 | |
| *** anotherjesse has joined #openstack | 20:46 | |
| *** devcamcar has joined #openstack | 20:46 | |
| *** sleepsonthefloor has joined #openstack | 20:46 | |
| *** vishy has joined #openstack | 20:47 | |
| *** xtoddx has joined #openstack | 20:48 | |
| *** sparkycollier has quit IRC | 20:53 | |
| alekibango | i have reworded my blueprint to be smarter, more generic and compatible: https://blueprints.launchpad.net/nova/+spec/limit-resources-ensure-availibility | 20:54 |
| alekibango | now i believe it will be acceptable by most :) | 20:56 |
| *** kevnfx has joined #openstack | 20:56 | |
| *** Orman has joined #openstack | 20:57 | |
| Orman | I'm back. | 20:57 |
| dendrobates | meeting in 2 min | 20:58 |
| Orman | Pardon me,but I have never been in on the meeting. | 20:58 |
| Orman | Where do Igo? | 20:59 |
| eday | #openstack-meeting | 20:59 |
| alekibango | Orman: /join #openstack-meeting | 20:59 |
| alekibango | ouch i have mistake in title :) | 20:59 |
| alekibango | finally i hope https://blueprints.launchpad.net/nova/+spec/limit-resources-ensure-availability :) | 21:00 |
| *** littleidea has quit IRC | 21:06 | |
| *** littleidea has joined #openstack | 21:07 | |
| *** Cybodog has quit IRC | 21:15 | |
| *** Cybodog has joined #openstack | 21:15 | |
| *** westmaas1 has quit IRC | 21:17 | |
| *** littleidea has quit IRC | 21:23 | |
| *** littleidea has joined #openstack | 21:23 | |
| *** cruciform has joined #openstack | 21:27 | |
| *** spectorclan has quit IRC | 21:30 | |
| *** jdmaturen has left #openstack | 21:31 | |
| *** johnpur has quit IRC | 21:32 | |
| *** Orman has quit IRC | 21:33 | |
| *** mdiver has quit IRC | 21:36 | |
| * ttx goes to bed | 21:38 | |
| *** joearnold has quit IRC | 21:38 | |
| mtaylor | creiht: yes! | 21:41 |
| mtaylor | creiht: but not this instant | 21:41 |
| *** littleidea_ has joined #openstack | 21:41 | |
| *** littleidea has quit IRC | 21:45 | |
| *** littleidea_ is now known as littleidea | 21:45 | |
| *** HouseAway is now known as AimanA | 21:51 | |
| *** joearnold has joined #openstack | 21:52 | |
| *** jkakar has joined #openstack | 21:59 | |
| *** kevnfx has quit IRC | 22:00 | |
| mtaylor | creiht: so remind me and stuff | 22:01 |
| creiht | mtaylor: k, no worries | 22:02 |
| *** sebastianstadil has quit IRC | 22:03 | |
| *** khussein has quit IRC | 22:04 | |
| *** ppetraki has quit IRC | 22:04 | |
| *** pvo has quit IRC | 22:04 | |
| *** sparkycollier has joined #openstack | 22:06 | |
| *** theron has quit IRC | 22:10 | |
| *** pvo has joined #openstack | 22:11 | |
| *** ChanServ sets mode: +v pvo | 22:11 | |
| *** Orman has joined #openstack | 22:13 | |
| *** ddumitriu has joined #openstack | 22:14 | |
| *** Orman_ has joined #openstack | 22:16 | |
| *** Orman has quit IRC | 22:20 | |
| Orman_ | alekibango: That was a great meeting. | 22:22 |
| *** khussein has joined #openstack | 22:27 | |
| *** gondoi has quit IRC | 22:32 | |
| alekibango | heh :) | 22:37 |
| Orman_ | Yep. ;) | 22:42 |
| alekibango | those conferences... i wish we could do them via inet only | 22:42 |
| *** littleidea has quit IRC | 22:43 | |
| *** abecc has quit IRC | 22:46 | |
| *** sparkycollier has quit IRC | 22:46 | |
| *** abecc has joined #openstack | 22:46 | |
| Orman_ | Yeah vidoe chat | 22:52 |
| Orman_ | ;) | 22:52 |
| alekibango | for me audio+ text + screen is ok | 22:52 |
| *** pvo has quit IRC | 22:59 | |
| *** pvo has joined #openstack | 23:00 | |
| *** pvo has quit IRC | 23:00 | |
| *** pvo has joined #openstack | 23:00 | |
| *** ChanServ sets mode: +v pvo | 23:00 | |
| *** littleidea has joined #openstack | 23:00 | |
| *** DubLo7 has joined #openstack | 23:05 | |
| *** mdiver has joined #openstack | 23:05 | |
| *** pvo_ has joined #openstack | 23:07 | |
| *** ChanServ sets mode: +v pvo_ | 23:07 | |
| *** hggdh has quit IRC | 23:08 | |
| *** Cybodog has quit IRC | 23:08 | |
| *** blakeyeager has quit IRC | 23:09 | |
| *** pvo has quit IRC | 23:10 | |
| *** metoikos has quit IRC | 23:10 | |
| *** pvo_ has quit IRC | 23:11 | |
| *** hggdh has joined #openstack | 23:11 | |
| *** perestrelka has quit IRC | 23:12 | |
| *** perestrelka has joined #openstack | 23:15 | |
| *** neogenix has quit IRC | 23:18 | |
| *** jkakar has quit IRC | 23:25 | |
| Orman_ | alekibango: right | 23:27 |
| Orman_ | So where were we in our security discussion | 23:27 |
| Orman_ | ;) | 23:27 |
| *** sebastianstadil has joined #openstack | 23:28 | |
| jc_smith | anybody seen this one in nova-compute before? | 23:35 |
| jc_smith | Traceback (most recent call last): | 23:35 |
| jc_smith | File "/opt/novascript/nova/nova/compute/manager.py", line 92, in run_instance | 23:35 |
| jc_smith | yield self.driver.spawn(instance_ref) | 23:35 |
| jc_smith | AttributeError: virConnect instance has no attribute 'nwfilterDefineXML' | 23:35 |
| *** krish has quit IRC | 23:36 | |
| jc_smith | hmm, maybe wrong version of libvirt | 23:37 |
| *** zaitcev has joined #openstack | 23:42 | |
| *** devx has joined #openstack | 23:43 | |
| *** gaveen has quit IRC | 23:47 | |
| *** pvo has joined #openstack | 23:48 | |
| *** ChanServ sets mode: +v pvo | 23:48 | |
| *** ambo has joined #openstack | 23:52 | |
| *** elasticdog has quit IRC | 23:53 | |
| *** littleidea has quit IRC | 23:54 | |
| *** sebastianstadil has quit IRC | 23:54 | |
| *** abecc has quit IRC | 23:54 | |
| *** dysinger has quit IRC | 23:57 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!