18:01:25 #startmeeting training-labs 18:01:25 Meeting started Mon Jul 18 18:01:25 2016 UTC and is due to finish in 60 minutes. The chair is dbite. Information about MeetBot at http://wiki.debian.org/MeetBot. 18:01:26 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 18:01:29 The meeting name has been set to 'training_labs' 18:01:50 roll call 18:02:20 hi dbite 18:02:25 julen: hi 18:02:49 rluethi: hellop 18:03:14 hi rluethi 18:03:31 julen: May be he needs a few more minutes 18:03:43 julen: So what is the news with PXE feature 18:03:47 #topic PXE 18:04:51 well... I haven't done much, I just checked roger's patch 18:04:59 ok 18:05:03 anything interesting? 18:05:21 which is quite good by the way... he fixed the pxecluster file 18:06:32 uff.. Roger just wrote saying he has issues with IRC 18:06:46 julen: sounds good. I havent yet seen any email from his side so I assume he has some technical issues 18:08:40 anything else on this topic? 18:08:47 let's wait some minutes, maybe he can restart or something.. 18:09:11 ping? 18:09:17 not from my side.. I didn't have time to write anything 18:09:21 hi rluethi! :) 18:09:35 rluethi_: pong 18:09:36 rluethi_ 18:09:40 hey, sorry, had to switch to the web client. 18:09:46 rluethi_: no worries 18:09:51 issues with my mac IRC client. 18:09:51 we are discussing the PXE feature 18:09:59 reading up on it. 18:10:04 * dbite wants to keep todays meeting short. 18:10:34 I have one general suggestion... 18:10:46 shoot 18:11:17 it really does take a lot of time to build the whole thing, and a big share of that time waste is the fact that we are using the us repositories 18:12:09 we should put some "sed" in somewhere to remove the "us." part of the repository addresses, so that it takes the closest one to the current location 18:12:45 or catch the locale from the local machine and use it somehow 18:13:31 julen: for regular builds it's not much of a problem, because it happens only once (basedisk build). 18:14:05 julen: for PXE build, it would make sense to have the required packages on pxeserver, then we only need to download them once. 18:14:37 but if you have a good way of picking the appropriate server, that would be very welcome :). 18:14:45 rluethi_: sure.. or even copy them directly from the base disk 18:15:40 it is easy on the /etc/apt/sources.list, but I still don't know how to do it on the preseed 18:15:58 julen: we should think about this after the current patch is merged 18:16:20 julen: We could think about catching the packages on the PXE server node. 18:16:34 yeah... I just wanted to put it on the "todo" list :) 18:16:36 yes, it is an improvement, it is not required for the initial version. 18:17:14 julen: yes, certainly a worthwhile improvement. 18:17:36 I'll try to figure out how to do it on the preseed 18:18:35 in any case.. back to the topic... 18:18:38 julen: maybe you can append to the boot command like I did for fixing the PXE server. 18:18:40 rluethi_: great job! 18:18:54 julen: thx 18:19:26 next topic? 18:19:35 so, what is still missing? 18:19:54 for pxe booting? 18:20:00 attaching the machine back to the OS scripts, right? 18:20:02 yes 18:20:20 it should be all working. for VMs, anyway. 18:20:30 baremetal will need some additional changes. 18:20:55 oh! no.. that's also done! :) 18:21:11 julen: have you tested it with baremetal? 18:22:00 rluethi_: well... I am a little restricted with hardware, but I can try it with "virtual" baremetal... 18:22:34 julen: I will need to test with real hardware. there are some things that you can't emulate. 18:23:06 initially the machines should be started by hand anyway, until we write some WOL part or something 18:23:15 rluethi_: for example? 18:24:40 julen: rluethi_ lets discuss this on ML rather than here. 18:24:42 julen: MTU comes to mind. But the point is more that you don't know until you tried. There's always something you didn't consider. 18:24:56 dbite: right. next up? 18:25:08 #topic Parser 18:25:28 I did not get enough quality time for finishing the parser 18:25:41 most of my time went in adding the custom code to RST 18:25:57 you guys should get an email when I finish the initial bits very soon 18:26:28 looking forward to testing it ;) 18:26:35 same here. 18:26:51 * dbite Is working very hard :) 18:26:58 #topic python port 18:27:06 rluethi_: Pushed his VirtualBox code on github 18:27:23 I am again trying to make sense of how to get things together but I am not stuck :) 18:27:35 rluethi_: When do you wish to see the python port in action? 18:27:47 dbite: I am working on adding PXE functionality to the Python port. 18:28:01 dbite: what do you mean? 18:28:15 I see it in action every time I test it. 18:28:30 rluethi_: I was thinking that we should be releasing Newton very soon (1.5 months~( 18:28:49 so we should aim to get the python version running before we work on the O release 18:28:57 dbite: I'll do my best to get it into a mergeable state. 18:29:30 dbite: we need to port the scripts to newton as well. 18:29:36 and, is it so, that the official Newtone release will be just in python, or we still keep the bash one too? 18:29:43 rluethi_: Ill take that over 18:29:58 julen: Do you want to contribute to Newton port too? 18:29:59 dbite: makes sense. 18:30:14 sure.. if I can somehow... 18:30:23 I can always test... 18:30:45 julen: I want to drop the bash version as soon as possible, maintaining two code bases is consuming too much of our resources. 18:30:47 rluethi_: I have planned some time for the Netwon port. This time though I want to get alternative method as compared to one single big patch 18:31:04 #topic Netwon Port 18:31:20 julen: I will give you some more details soon, free this weekend? 18:31:20 is it just about reading the docs and updating the scripts? 18:31:39 julen: Almost but making it work reilably takes a lot more effort 18:31:40 hmm... not fully, but we can arrange something 18:32:11 rluethi_: does it make sense to release Newton first? Also I guess I will have to additionally work on install-guides again. 18:32:31 Then immediately get the python port out. 18:32:43 rluethi_: I plan to get the playground for the python port in asap. 18:32:44 dbite: Newton first, before what? Python port? 18:32:49 is it so, that the bash scripts are 100% compatible with the python port? 18:32:56 Yes before python port. 18:33:12 julen: We start with eliminating the host side script first 18:33:23 julen: the same scripts on the client side. 18:33:27 julen: So they should definitely be 100% compatible. 18:33:44 good 18:34:10 dbite: the order shouldn't matter, because they should be independent of each other. 18:34:27 rluethi_: I think its important for our sanity IMHO 18:34:42 (unless Newton requires architectural changes like Mitaka did) 18:35:13 rluethi_: May be but let me check. Also I have some plans to do the porting in a way that I could kind of learn upgrading the OpenStack version specifics 18:35:17 by doing it hands on 18:35:46 rluethi_: This should allow us to add this feature later this year or early next year. Should be a cool challenge to take over. 18:36:06 dbite: live upgrade? 18:36:21 julen: Not exactly 18:36:45 well... "cold" upgrade... 18:36:45 julen: But we could try live upgrade, I definitely do not have time for live upgrade for the next 8 months. 18:37:16 ask Dirk ;) 18:37:51 (because of the rule_the_stack... ) 18:38:19 well.. but first things first: priority list for me? 18:38:31 julen: Lets sync up this weekend for the same 18:38:39 lets move on 18:38:47 #topic release_cycle 18:38:51 dbite: ok 18:39:11 Last time we decided to have one major release and one optional minor release. 18:39:22 rluethi_: Do you also want to keep some milestones for the same? 18:40:19 I don't really know what this means in the context of this project. That might be something that needs some discussion. 18:40:36 Maybe just go with whatever is easy and discuss the rest in Barcelona. 18:40:59 rluethi_: Alright, I think I got what I understand right now. We just stick to one major release and one mile stone if required between the major releases. 18:41:46 dbite: please cc us in those discussions when possible so we have a chance to learn what you are learning. 18:41:49 rluethi_: Dont worry about it, I will do the required work for the same with dhellmann, I think I remember what we discussed last time but I forgot to push the governance patch. I will push it soon, worst case tomorrow morning 18:42:10 rluethi_: yes, the other chat was on IRC so it was probably not possible to share it :) 18:42:29 dbite: fair enough. 18:43:23 #topic any other beezneez? 18:43:33 reviews. 18:44:00 did anyone get a chance to look at the two patches I posted Saturday? 18:44:04 rluethi_: Sitting right after the meeting. I will finish them reviews ;D 18:44:35 yes, I already voted... 18:44:49 julen: my apologies, I just noticed that you did. 18:44:57 julen: thank you. 18:45:37 ok, so all set, anything else? 18:45:49 julen: have you been able to test them, or did you just read the code? 18:46:39 I am asking because the proxy patch does need more testing. I built a test setup, but I would like somebody else to confirm that it works. 18:46:44 rluethi_: copy pasted the code 18:47:32 rluethi_: I will test it. 18:47:51 So guys, lets finish this meeting for today. We could additionally continue our discussion over IRC 18:47:55 I set up a test machine and killed its default routes. 18:48:10 It had to go through a squid server, only way out. 18:48:19 I hope. 18:48:58 Okay, let's end this then. 18:49:32 bye guys 18:49:35 #endmeeting