17:03:20 #startmeeting training-guides 17:03:22 Meeting started Mon Sep 8 17:03:20 2014 UTC and is due to finish in 60 minutes. The chair is sarob. Information about MeetBot at http://wiki.debian.org/MeetBot. 17:03:23 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 17:03:25 The meeting name has been set to 'training_guides' 17:03:38 #topic docs 17:03:56 docs is under slow transition for Icehouse 17:04:14 partly its my mistake, I did not get enough time to update them due to my busy schedule 17:04:27 sarob: can you help me out with the docs? 17:04:35 to update/modify them for Icehouse 17:04:52 I'm slammed this week, but yes I 17:04:56 Can help 17:05:11 sure, thanks :) 17:05:22 I think sayali has taken up the bug for associate guide 17:05:31 yep 17:05:39 I am almost done fixing the basic-install-guide for Icehouse 17:05:42 needs another brushup 17:05:56 operator guide is left, which needs serious refactoring for now 17:06:04 sayali_: thanks 17:06:22 dguitarbite: what's the status of bringing back any changes you've made to install guide? 17:06:31 dguitarbite: are they icehouse only? 17:06:54 annegent_: we are thinking of releasing Icehouse first and then focus on Juno 17:07:41 That's my pref 17:07:50 I am still not sure if training guides needs should publish content one release later or with the current release 17:08:00 what do you guys think? 17:08:34 I'd like to stay current 17:08:45 We depend on distros to make the packages available in time, or the scripts won't work. 17:08:52 there is a small lag for the training content, but we should be current 17:08:53 Stay with the rest of the docs program 17:09:23 i.e. we can install what's not in the repo, we don't do devstack. 17:09:24 hmm, ok 17:09:58 * rluethi wonders if there's some form of chair inheritence 17:10:13 * dguitarbite problem solved 17:10:20 phew :) 17:10:25 :D 17:11:03 Hey colinmcnamara 17:11:06 I think we should not have to worry about the package preferences 17:11:13 hello colinmcnamara 17:12:39 Rluethi: why would we? 17:12:53 sarob: do what? 17:12:58 sarob: we need the distro packages for our labs section and install section 17:13:14 we cannot use the repositories like devstack except may be for the developer guides 17:13:33 Dguitarbite: ah right 17:13:39 sarob: we can't install juno if ubuntu repos don't have them. 17:14:00 hey guys 17:14:10 hey colinmcnamara 17:14:11 rluethi: whats the usual lag for repos? 17:14:18 hey colinmcnamara 17:14:21 Rluethi: so canonical is about 1 month behind stable 17:14:27 We should ask the install guides people and the distros maintainers for this issue, Im pretty sure they have their unstable (not exact term) repos 17:14:39 sarob: I agree 17:14:59 Afternoon colinmcnamara 17:15:13 lets move on, we could continue this discussion on the docs channel later on. Any questions for the docs sub team? 17:15:17 sarob: that means our scripts will lag somewhat, too; but I certainly wouldn't like us to always be a full release behind. 17:15:29 afternoon Sarob: 17:15:46 afternoon colinmcnamara good to see you 17:15:54 :) 17:16:06 #topic infra 17:16:37 The audio visual content is in progress. Need everyone's inputs and opinions on it 17:17:08 https://etherpad.openstack.org/p/openstck-training-guides%28Audio_Visual_Content%29 17:17:09 sayali_: I missed the memo on this. is there a link? 17:17:20 nm. 17:17:28 #link https://etherpad.openstack.org/p/openstck-training-guides%28Audio_Visual_Content%29 17:18:30 The etherpad link was sent out on the mailing list and it is on the wiki 17:18:35 Sayali_: so these are slides or topics to be covered on video? 17:19:23 Well initially I was thinking we could have them as seperate slides and or combine some 17:19:53 for instance for controller node we could combine it in one video 17:19:54 sayali_: which mailing list? 17:20:14 where as in the basic install guides we maintain the modularity 17:20:34 Sayali_: so we would just cover the training guide books? 17:20:39 rluethi, 17:20:55 Sayali_: I'm confused 17:21:02 sayali_: ah. that explains a lot :). 17:21:02 don't we use Docs ML? 17:21:05 sarob, yes we could start with that 17:21:09 sayali_: we use openstack-docs mailing list 17:22:16 sarob, I am open to not following the training guide structure and building a new structure covering all the content 17:22:38 dguitarbite, oops, my bad. Will send it on that 17:22:40 Sayali: I'd start by just recording teaching each book to a user group and/or class 17:23:01 sayali_: hmm.. that will complicate things 17:23:03 sayali_: following current content initially would be a good start 17:23:31 sarob, I was also considering having the videos as a screenshare of what is being deployed 17:23:56 Sayali_: that sounds good too 17:24:02 instead of just an audio book 17:24:40 Sayali_: keep it simple. Just help deliver the training guide content 17:24:53 matjazp, Initially it would, but we might cover more content without repitition in that case 17:25:12 Sayali_: we may do more someday, but let's do that as a team 17:25:22 I agree 17:25:26 should we vote on this? 17:25:37 sarob, alright so we keep the structure similar to the training-guides? 17:25:48 sure sarob 17:25:54 Sayali_: yes please 17:26:07 sayali_: not the structure per se, but content should augment training guides 17:26:21 Sayali_: making good video quality is hard enuf on its own 17:26:44 What he said 17:27:06 As the place to start 17:27:08 sarob, yes we need to decide on what we would be using for the recording too 17:27:12 * dguitarbite has a question 17:27:16 sarob: that's why we should use clips already recorded (summits, etc.) 17:27:35 Matjazp I like that 17:27:44 who will record the videos (if required)? 17:28:01 The next upstream training at the summit would be awesome 17:28:29 Sayali_: who records? 17:28:52 sarob, I can do it 17:29:10 sarob: we could also use the videos from hackathons 17:29:16 at summit, there's already a professional recordings 17:29:33 Sayali_: go ahead and write down what your plans are into your section on the wiki 17:29:34 and may be try to keep 20-30 mins in the hackathon for creating audio visual content 17:29:39 sarob, anybody else who wants can help too, once we decide on what would render good video recordings 17:30:16 Sayali_: focus on the operations of making video to start 17:30:40 sarob, sure thing 17:30:48 Sayali_: augmenting the existing training materials 17:31:23 sarob, yep 17:31:26 Sayali_: once we get that down, we can move onto new ideas 17:31:47 Sayali_: thx for picking this up 17:32:16 sarob, glad to help :) 17:32:28 #action sayali_ updates av wiki section with plans 17:32:38 next thing, about the labs section 17:32:47 the openstack cluster builds, VMs are running 17:33:03 the current issue is that the VM's (openstack instances) do not get DHCP and metadata 17:33:09 hence cannot be accessed from the host machine 17:33:36 Dguitarbite virtualbox bug? 17:33:51 sarob: I am not sure if its virtualbox bug, or its something to do with ML2 17:34:03 I had complete cluster running for Grizzly 17:34:06 wireshark? 17:34:10 with DHCP and metadata 17:34:17 and see where packets die? 17:34:23 matjazp: I use tcpdump, ip, etc. to do this 17:34:30 Neutron seems to work fine 17:34:44 it may be something to do with libvirt/qemu/kvm/ml2 17:34:50 the OVS tunnels work fine 17:35:01 I am trying to use VLAN tagging instead of GRE to see if that works 17:35:28 the packets seems to reach the compute node, after that I am lost (for now) 17:35:40 I plan to start looking into this, too. I have spent the past week tracking down a particularly elusive filesystem corruption bug. 17:36:03 rluethi: thanks 17:36:27 I hope its not due to hardware limitations and I wish this issue to be resolved without having to tweak any code 17:36:28 We can put it out on ask 17:36:36 yes, I think we should 17:36:51 And link it back to a bug 17:37:28 sarob: I'm not sure if it should be logged as a known bug yet 17:37:41 are we sure its a bug? or is it a misconfiguration 17:37:42 I need some more people to test it and confirm if its the same issue for them too 17:37:48 Dguitarbite your call 17:38:01 matjazp: even if its a misconfiguration, its a bug for training guides labs 17:38:02 sarob: agree with dguitarbite, we don't really know what the problem is yet. 17:38:23 rluethi: I tried it, and yes, no DHCP 17:38:23 we wont be logging a bug under neutron or nova for sure at present till we are extremely sure 17:38:30 or metadata 17:38:35 As soon as the scripts are going, I want to put the call out to the operators ML 17:38:53 To start using as part of training 17:39:11 sarob: this can be used for associate guides training to certain extent 17:39:16 if hardware is the limitation 17:39:27 for hackathos etc. for now 17:39:43 KVM backend would be helpful for debugging that kind of issues. redundancy helps. 17:39:47 More people using good thing 17:39:55 Right? 17:40:08 sarob: yes 17:40:15 we are about 99% there with the scripts 17:40:18 sarob: as soon as it is usable. 17:40:30 we may put it under BETA tag and start spreading the word around 17:40:36 should we vote on this? 17:40:42 dguitarbite: it should work first 17:40:49 dguitarbite: everybody knows that the last % takes half the time. 17:41:06 agree with matjazp. 17:41:07 these scripts will be used by beginners and they will be confused 17:41:12 Sure, what are voting on? 17:41:19 doesn't have to be perfect, but no known major issues. 17:41:24 matjazp, rluethi: ok, then I will log this as a known bug under training guides labs section and fix this issue 17:41:29 no IPs are major issue 17:41:40 matjazp: indeed :). 17:41:47 sarob: I think I got my answer, I wanted to know if we should advertise training labs 17:42:02 Ah right 17:42:04 matjazp: I agree 17:42:25 #action dguitarbite will log a bug about the DHCP issue in openstack training labs 17:43:45 And I will notify the sfbay user group 17:43:59 about 5k people 17:44:20 Should provide good feedback when you are ready 17:44:32 Moving on 17:44:36 sarob: sure, that would be awesome 17:44:54 #topic upstream training patches 17:45:18 there are two issues in my mind: 1. This review https://review.openstack.org/#/c/119264/ 17:45:21 #link https://review.openstack.org/#/c/119264/ 17:45:38 this review adds odp presentation which is a binary, should we let this through? 17:45:59 look at review comments 17:46:26 we should merge this 17:46:57 matjazp: ok, I saw the comments but I dont see a +1 or +2 :) 17:47:12 I'll add my +2 17:47:16 my bad 17:47:21 fixed 17:47:23 :) 17:47:26 so I got confused 17:47:30 I need get on the horse 17:47:32 sure, lets merge them 17:48:22 ahh.. to quick.. There's still Rogers comments to fix 17:49:46 Oops, I guess I can fix and patch 17:50:26 sarob: yes, do that 17:50:58 So what's this about pushing landslide 17:51:08 From the last meeting 17:51:54 You want to merge a stable version to training guides? 17:52:01 Or something else? 17:52:52 Dguitarbite .... 17:52:57 sarob: I dont follow about pushing landslide 17:53:26 There was an action from the last meeting 17:53:43 sarob: ah, yes 17:53:49 Dbite will push python tool 17:53:56 I was trying to figure out where the python tool goes into 17:54:18 ajaeger and I decided to push the python scripts to openstack-doc-tool 17:54:28 so that it runs landslide, builds the html and pdf files 17:54:43 Dguitarbite ah okay. Nice 17:54:46 similar to the maven part it handels as of now 17:54:48 dguitarbite: nice 17:54:54 and it could also be a gate test in jenkins 17:55:00 * sarob likes 17:55:02 for future slides 17:55:13 but it will take some time, 2-3 weeks 17:55:23 * sarob automakic good 17:55:32 dguitarbite: in the mean time, we do it manually 17:55:40 Dguitarbite thanks for getting that going 17:55:42 yes, I will push it on wiki 17:55:52 the steps to do it manually its not that difficule 17:56:05 *difficult and also provide the barebone python scripts for the same 17:56:06 Good wiki fooder 17:56:09 yes 17:56:19 #topic incubation 17:56:41 #action dguitarbite will continue working on python scripts to provide PDF for HTML as a part of openstack-doc-tools 17:56:55 So how does the team feel about asking the TC for this? 17:57:00 #action dguitarbite will update the wiki for manual steps to build the presentations to PDF format 17:57:42 sarob: should we keep this as the priority discussion for next weeks meeting? 17:57:56 Let's do that 17:58:01 sarob: may be a bit early. I'd give it another week or two. 17:58:03 we are kind of out of time and I think we need to prepare a bit more on this 17:58:27 I'm not sure what incubation gets us at this point 17:58:38 Let's talk about it next week 17:58:58 #action speak of incubation at next meeting 17:59:13 #topic any other business 17:59:29 Real quick like 17:59:34 I think we are out of time now :| 17:59:36 im good 17:59:44 I got nothing. 17:59:44 Me too 17:59:50 lets talk on docs channel if requried 17:59:51 I'm good 17:59:52 Cheers ! 17:59:53 next meeting 17:59:58 Bye 18:00:02 thanks a lot, bye 18:00:08 #endmeeting