17:12:19 <ccneill> #startmeeting OpenStack Security Project 17:12:20 <openstack> Meeting started Thu Mar 10 17:12:19 2016 UTC and is due to finish in 60 minutes. The chair is ccneill. Information about MeetBot at http://wiki.debian.org/MeetBot. 17:12:21 <redrobot> ccneill #startmeeting security 17:12:22 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 17:12:25 <openstack> The meeting name has been set to 'openstack_security_project' 17:12:31 <ccneill> derp.. 17:12:34 <lhinds> thanks ccneill 17:12:41 <michaelxin> mvaldes: +1 17:12:58 <ccneill> so.. Anchor :D 17:13:16 <ccneill> sorry, getting ahead of myself. summit planning. 17:13:19 <diazjf> Summit Planning :) :) 17:13:26 <ccneill> #topic Summit Planning 17:13:26 <mvaldes> #topic Summit Planning 17:13:29 <redrobot> ccneill for future reference the name ends up being used for the directory where the logs are stored. For consistency this meeting should use "security" 17:13:35 <krotscheck> o/ 17:13:36 <redrobot> ccneill http://eavesdrop.openstack.org/meetings/security/2016/ 17:13:52 <ccneill> redrobot: dang, knew I got that wrong. thanks for the heads-up >< 17:14:01 <ccneill> anyone have anything on summit planning? 17:14:15 <michaelxin> Did everyone already register for the summit? 17:14:18 <redrobot> yes, has anyone submitted the BYOK spec for the cross-project track? 17:14:24 <michaelxin> The early bird price will end tomorrow. 17:14:26 <diazjf> ccneill, I wanted to talk about BYOK 17:14:43 <ccneill> well, I think you have the floor :) 17:14:44 <redrobot> if not picked for cross-project, I'd like to see BYOK be a fishbowl 17:14:50 <krotscheck> All I have is that a threat analysis session on CORS was proposed. 17:15:06 <diazjf> I have a spec up for BYOK 17:15:07 <diazjf> https://review.openstack.org/#/c/271517/ 17:15:56 <ccneill> krotscheck: I've been meaning to take a look at the CORS stuff some more. is there a CR or something up for that re: security? 17:16:04 <mvaldes> #link https://review.openstack.org/#/c/271517/ 17:17:15 <krotscheck> ccneill: Not certain what you mean. There's patches in all of the various API's? 17:17:38 <michaelxin> From last time, the agreement is to have a threat modeling session for CORs 17:17:40 <ccneill> krotscheck: sorry, I was just wondering if there was an explicit action item for OSSP to review CORS 17:17:43 <ccneill> all I've seen is the spec 17:17:45 <ccneill> #link http://specs.openstack.org/openstack/openstack-specs/specs/cors-support.html 17:18:18 <ccneill> is your concern individual implementations for each project? or the implementation in general 17:18:24 <krotscheck> ccneill: All i know is the discussion from last week about "Hey, we should do a Thread Analysis at the summit, that'd be a neat exercise" 17:18:36 <ccneill> krotscheck: gotcha. agreed! 17:19:22 * krotscheck won't really have much time to organize it, but is willing to be there as an expert on the spec. 17:19:29 <krotscheck> (The W3C spec that is) 17:20:00 <michaelxin> There might not be too much to talk besides implementation 17:20:09 <michaelxin> The spec is solid. 17:20:10 <ccneill> cool, well I think I'll leave the duty of officially setting up that session to our friend hyakuhei since I'm not even doing a great job running this meeting ;) 17:20:15 <michaelxin> The implementation is key 17:20:30 <michaelxin> Bad implementation will have big impact. 17:21:03 <ccneill> ok, anything else for summit? 17:21:14 <ccneill> going once.. 17:21:21 <michaelxin> Not sure 17:21:28 <ccneill> cool, Anchor time 17:21:31 <ccneill> #topic Anchor 17:21:55 <ccneill> redrobot? anyone else? 17:22:38 <michaelxin> let's move on 17:22:42 <ccneill> cool 17:22:44 <ccneill> #topic Bandit 17:22:49 <michaelxin> Tim? 17:22:49 <ccneill> hmm 17:22:53 <ccneill> no tmcpeak it appears 17:23:03 <lhinds> I have something quickly.. 17:23:07 <ccneill> anyone else have Bandit news? 17:23:10 <ccneill> cool 17:23:31 <browne> so on bandit 17:23:40 <lhinds> request was made to put some release notes together, so I digged out a format I have used before, and apache F releases typically go for 17:23:56 <lhinds> #link http://paste.openstack.org/show/489522/ 17:23:58 <browne> i've been trying to fix projects to get rid of bandit.yaml 17:24:11 <lhinds> not sure if this is what we want, here is a live example 17:24:18 <lhinds> #link https://db.apache.org/ojb/release-notes.txt 17:24:24 <lhinds> so its a rolling release note 17:24:33 <ccneill> browne: has YAML-free config landed in the version PyPI is shipping now? 17:24:35 <michaelxin> browne: What's your take about lhinds's release note? 17:24:37 <lhinds> you add each new release and it stays in the project root 17:24:51 <ccneill> lhinds: is this related to / in the same format as "reno"? 17:25:04 <ccneill> this thing 17:25:05 <browne> ccneill: not totally until 1.0, but most projects can get away without it. so been changing that 17:25:08 <ccneill> #link http://docs.openstack.org/developer/reno/ 17:25:10 <lhinds> I don't know what reno is yet :S 17:25:12 <browne> and putting bandit in the pep8 testenv 17:25:28 <ccneill> lhinds: it came up last week when we were talking about release notes, but I'm not familiar with it either 17:25:35 <browne> the integration test is still failing so, i'm still working to fix projects and bandit 17:25:41 <lhinds> ahh I see, so looks like we have a format already 17:25:48 <michaelxin> browne: +1 17:26:11 <cjschaef> browne: the problem I have with bandit in the pep8 testenv, is when bandit checks for dirty repo's 17:26:15 <ccneill> browne: cool cool. what projects are you working with at this point? 17:26:51 <browne> ccneill: mainly focused on the ones in our integration test that are broken 17:27:00 <browne> so magnum, magnum client, etc 17:27:02 <ccneill> Designate was interested in implementing it, but I told them to hold off til 1.0 to land for the YAML-free config 17:27:37 <browne> cjschaef: yeah, not sure i like it in pep8 either, but think infra wants to cut down on the number of jenkins jobs 17:28:03 <cjschaef> browne: that makes sense, and I wasn't thinking of gates initially too 17:28:09 <ccneill> browne / cjschaef: do you think it might be worthwhile/feasible to propose a "security" gate job? 17:28:17 <ccneill> including bandit, syntribos, other tools to come? 17:29:14 <browne> ccneill: we'd have to run it by infra, but sounds good to me once we have these other sec scans (syntribos) running 17:29:33 <cjschaef> ccneill: I like that idea, and agree with browne on getting those tools ready first 17:29:47 <ccneill> mdong and I will be working on Syntribos more starting next week 17:30:14 <mvaldes> nice idea 17:30:17 <ccneill> so we'll definitely keep everyone posted as we go 17:30:45 <browne> ccneill: cool. is there a guinea pig project you'll start with? keystone? 17:31:13 <ccneill> browne: so we've been testing it internally on Solum 17:31:46 <ccneill> I think we're kind of using that as our guinea pig, and then we'll go from there 17:31:51 <michaelxin> After Solum, it will be keystone. 17:31:55 <browne> xool 17:31:57 <browne> cool 17:32:00 <ccneill> what he said ^ :) 17:32:33 <ccneill> cool, anything else on bandit? 17:32:45 <ccneill> going once.. 17:33:05 <ccneill> cool 17:33:08 <ccneill> #topic Docs 17:33:48 <ccneill> so I know I said I'd try to get a changelog going for bandit last week... unfortunately, I didn't have time this week 17:34:12 <ccneill> but I will have a lot more time for OSSP-related projects starting next week 17:34:19 <michaelxin> eelmiko? 17:34:19 <ccneill> for the foreseeable future 17:34:25 <ccneill> _elmiko ? 17:34:26 <michaelxin> elmiko? 17:34:31 <hyakuhei> Hey. Sorry. Crazy morning - accident in seattle 17:34:38 <hyakuhei> elmiko can’t make it today. 17:34:40 <ccneill> oh no 17:34:48 <lhinds> ccneill I can help with anything you might have that needs doing 17:34:54 <lhinds> inbetween jobs just now 17:34:56 <hyakuhei> unrelated (accident and elmiko) 17:35:17 <ccneill> hyakuhei: well, hope everyone's okay 17:35:23 <lhinds> + 1 17:35:28 <michaelxin> sorry to hear it 17:35:31 <hyakuhei> ccneill: Not sure. Traffic was closed for a long time. 17:35:35 <ccneill> hyakuhei: you want to take over? I'm not great at this whole chairing thing 17:35:37 <michaelxin> hope everyone is fine. 17:35:42 <hyakuhei> +1 17:35:52 <hyakuhei> Who’s chairing the meeting today? 17:35:56 <ccneill> o/ 17:36:00 <michaelxin> hyakuhei: ccneill 17:36:12 <hyakuhei> excellent! Thank you ccneill 17:36:13 <michaelxin> he is doing great so far. 17:36:17 <mvaldes> doing a stand-up job 17:36:23 <ccneill> no problem. we're on Docs 17:36:28 <hyakuhei> Please keep on rolling :) 17:36:40 <michaelxin> ccneill: move on next topic 17:36:58 <ccneill> lhinds: I'll look into this reno thing next week and I'll post it in os-sec if I come up with anything 17:36:59 <ccneill> okay 17:37:01 <ccneill> moving on 17:37:09 <ccneill> #topic Syntribos 17:37:37 <michaelxin> mdong: Can give detailed updates. 17:37:40 <hyakuhei> btw most of the HP people are in meeting this morning 17:37:54 <hyakuhei> (annual onsite this week) 17:37:55 <mdong> So as ccneill said earlier we’re gonna dedicate more time to Syntribos 17:38:07 <michaelxin> For resrouces, ccneill and mdong will work almost full time on syntribos 17:38:17 <hyakuhei> That’s awesome! 17:38:19 <ccneill> hyakuhei: gotcha. we realized how much we miss you all when you're not around this morning! :) 17:38:25 <michaelxin> We are also expecting two more developers from Intel joining the project. 17:38:28 <hyakuhei> awwww 17:38:37 <mvaldes> +1 17:38:42 <michaelxin> Still working on it. 17:38:53 <hyakuhei> That’s very cool 17:38:55 <ccneill> lots of big ideas for Syntribos :) 17:39:13 <michaelxin> Now, mdong is testing Solum using Syntribos 17:39:17 <ccneill> off the top of my head, we'll probably try to rip out the CAFE pieces to make it more palatable to OS projects 17:39:32 <michaelxin> He is working on some payload generation stuff. 17:40:00 <mdong> we’ll be testing Solum as a guinea pig, which hopefully will be a learning experience about how the tool can be improved 17:40:01 <michaelxin> There is a blueprint about payload automatic generation. 17:40:46 <nsun1> wiki for Syntribos? 17:41:07 <ccneill> michaelxin: do you have the link for that BP? 17:41:22 <mdong> there’s no syntribos wiki yet 17:41:36 <mdong> but all our blueprints can be found at https://blueprints.launchpad.net/syntribos 17:41:39 <mvaldes> where do we put a wiki? 17:41:51 <ccneill> mdong / michaelxin: I see the "create-payloads-cinder/glance/keystone/etc" BPs 17:42:04 <mvaldes> nsun1: just this so far https://github.com/openstack/syntribos 17:42:10 <michaelxin> That will be a good action item for ccneill and mdong 17:42:14 <ccneill> ah 17:42:15 <ccneill> this? 17:42:17 <ccneill> #link https://blueprints.launchpad.net/syntribos/+spec/research-ways-to-generate-payload 17:42:27 <michaelxin> ccneill: yes 17:42:32 <hyakuhei> It’s really great to see this moving forward, I especially like that Intel will be getting involved too. Making it more than just “that Rackspace thing” will be great for addoption. Not that I don’t appreciate all the work Rack is doing. 17:42:45 <ccneill> hyakuhei: +1 17:42:46 <mvaldes> +1 17:43:19 <mvaldes> also thanks to browne for helping review some of the CR's 17:43:29 <michaelxin> ccneill: That's all for syntribos 17:43:31 <michaelxin> move on please 17:43:41 <ccneill> hyakuhei: brought this up earlier, but would like your thoughts 17:43:50 <nsun1> sure 17:43:58 <ccneill> do you think it makes sense for us to propose a "security" gate, running bandit + syntribos + future tools(?) 17:44:31 <ccneill> browne mentioned infra was trying to cut down on Jenkins jobs, so I figured it might make sense to consolidate at some point 17:44:45 <hyakuhei> I think I’d prefer it to be compositional _but_ we can certainly create white papers on best practice combinations and perhaps even offer that as a bundle. So long as that bundle isn’t the _only_ way to use Syntribos/Bandit etc 17:45:49 <ccneill> hyakuhei: compositional? so jobs for each project? 17:46:15 <ccneill> hyakuhei: I agree, shouldn't be the only way to implement it, but would set bar for total adoption pretty low 17:46:45 <ccneill> in any case, we can probably hold off until Syntribos is in a more finished state 17:46:50 <ccneill> since we only have Bandit at the moment 17:47:04 <ccneill> okay, moving on unless anyone has more on Syntribos? 17:47:15 <michaelxin> hyakuhei: https://review.openstack.org/#/c/288150/ still shows Needs Workflow 17:47:44 <michaelxin> What do we need to get the CR move on? Thanks. 17:48:50 <ccneill> did we lose everyone? 17:48:54 <hyakuhei> michaelxin: I’ll check now 17:49:04 <michaelxin> hyakuhei: Thanks. 17:49:07 <hyakuhei> ok I’ll poke $people 17:49:07 <ccneill> #topic Publicity 17:49:09 <hyakuhei> I can’t +2 17:49:24 <ccneill> anything for Publicity? 17:49:50 <michaelxin> hyakuhei: who should we include for +2? Thanks. 17:50:01 <hyakuhei> michaelxin: that ’s what I need to find out 17:50:09 <hyakuhei> ttx: ^ Any idea? 17:50:40 <michaelxin> hyakuhei: cool, Thanks. 17:51:44 <ccneill> anything on publicity? got ~8 minutes 17:51:52 <hyakuhei> I don’t think that’s moved anywhere since last week 17:52:05 <michaelxin> ccneill: Do you still have friends in UT-Austin? 17:52:05 <hyakuhei> I did ask for budget for cool swag at the summit 17:52:07 <mvaldes> we did an internal tech talk and gave OSSP another plug 17:52:12 <ccneill> cool, yeah I guess we're waiting on bandit 1.0 17:52:14 <ccneill> hyakuhei: nice 17:52:19 <hyakuhei> mvaldes: cool! 17:52:24 <ccneill> michaelxin: not really? :\ I'm an old man 17:52:34 <michaelxin> haha 17:52:46 <michaelxin> Maybe, we can ask professor Matt T 17:52:53 <ccneill> haha yes! 17:53:12 <ccneill> ok, moving on 17:53:14 <ccneill> #topic OSSN 17:53:23 <ccneill> #link https://review.openstack.org/#/q/status:open+project:openstack/security-doc,n,z 17:53:32 <ccneill> any movement here? 17:53:33 <hyakuhei> https://bugs.launchpad.net/ossn 17:53:55 <hyakuhei> Seems to have stalled out a little. I’ll see if I can get a spring on that in the next few days. 17:54:46 <michaelxin> we only have three 17:54:51 <michaelxin> it is good 17:55:08 <ccneill> ok cool 17:55:13 <ccneill> sounds like we have that on lock 17:55:16 <ccneill> #topic Blog 17:55:17 <hyakuhei> Yeah, we just need to clear them out, some are a little bit old. 17:55:40 <hyakuhei> It’s still going forward, we need to get some more content in there though :) 17:55:42 <ccneill> how's the blog? any posts you want eyes on, hyakuhei ? 17:56:11 <ccneill> we can probably get a blog post together about Syntribos in the coming weeks 17:56:13 <hyakuhei> The TA one needs some polish 17:56:25 <ccneill> e.g. a guide on how to use it, or what is different about it 17:56:25 <michaelxin> ccneill: +1 17:56:27 <hyakuhei> https://openstack-security.github.io/threatanalysis/2016/02/07/anchorTA.html 17:56:32 <hyakuhei> ccneill: great idea 17:57:05 <ccneill> hyakuhei: yeah, I liked this post. where do you do review? 17:57:11 <hyakuhei> github 17:57:25 <hyakuhei> https://github.com/openstack-security/openstack-security.github.io 17:57:31 <ccneill> ah so just do a PR on the post? gotcha\ 17:57:52 <ccneill> I'll try to add some thoughts, though I think this is a pretty solid start already 17:57:57 <hyakuhei> Yeah, standard Github workflow I guess :) 17:57:58 <ccneill> ok 2 minutes 17:58:00 <hyakuhei> Thanks ccneill 17:58:06 <ccneill> #topic CORS 17:58:14 <ccneill> sorry we didn't save much time for this diazjf 17:58:48 <ccneill> oops, I mean krotscheck 17:58:56 <krotscheck> Now orries :) 17:58:59 <krotscheck> yay 2 minutes! 17:59:13 <ccneill> yay! good news is, I won't be chair next week :) 17:59:20 <ccneill> so we'll probably actually get to it 17:59:22 <krotscheck> So, what'll it take to get a fishbowl at the summit? 17:59:31 <krotscheck> (tick tock tick tock) 17:59:41 <ccneill> let's take it to #openstack-security 17:59:46 <ccneill> think we gotta wrap up here 17:59:58 <ccneill> thanks, all! 18:00:00 <ccneill> #endmeeting