21:03:43 <b1airo> #startmeeting scientific-wg
21:03:44 <openstack> Meeting started Tue Aug  8 21:03:43 2017 UTC and is due to finish in 60 minutes.  The chair is b1airo. Information about MeetBot at http://wiki.debian.org/MeetBot.
21:03:45 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
21:03:47 <openstack> The meeting name has been set to 'scientific_wg'
21:04:03 <b1airo> #chair martial
21:04:03 <openstack> Current chairs: b1airo martial
21:04:05 <armstrong> Hello
21:04:11 <b1airo> hi armstrong
21:04:50 <priteau> Hello
21:05:07 <b1airo> let's give oneswig another minute or two - i'm just in the process of getting some breakfast (7am here)
21:05:10 <b1airo> hi priteau
21:05:32 <martial> blair: stig is not coming tonight I believe (from his email)
21:05:52 <martial> #link https://wiki.openstack.org/wiki/Scientific_working_group#IRC_Meeting_August_8th_2017
21:06:08 <martial> The topics of interest tonight are listed above
21:06:12 <b1airo> ah ok, haven't read email yet
21:06:19 <martial> #topic Supercomputing 2017
21:06:34 <martial> FYI, the BoF application was submitted
21:07:15 <b1airo> yep, fingers crossed
21:08:44 <b1airo> in the process of submitting i realised we had not done a writeup from SC16 (which we said we would)
21:08:47 <rbudden> hello
21:08:51 <b1airo> hi rbudden
21:08:55 <martial> hello Robert
21:09:16 <rbudden> sorry i’m a bit late, working on our Ocata upgrade
21:09:22 <martial> sounds fun :)
21:09:32 <martial> blairo: cool, thanks for that
21:09:46 <martial> blairo: what is the deadline for approval?
21:10:21 <b1airo> i have started this wiki page as a place to distill down some HPC with OpenStack related notes:
21:10:23 <b1airo> #link https://wiki.openstack.org/wiki/HPC
21:10:54 <b1airo> i think they are aiming to have approvals done by early september
21:12:09 <martial> okay crossing our fingers :)
21:12:19 <martial> #topic Update to the Scientific OpenStack book
21:12:52 <martial> so a v2 is in discussion
21:13:09 <hogepodge> o/ (sorry I'm late)
21:13:11 <b1airo> is hogepodge around
21:13:18 <b1airo> ah, hey there!
21:13:20 <rbudden> good timing ;)
21:13:20 <martial> we are calling everybody interested to reach out
21:13:35 <martial> knowing that there is an unmovable non-negotiable hard deadline of content being completed by September 29
21:13:44 <rbudden> did we decide on just an update to the book or are we adding new content?
21:13:47 <hogepodge> So Kathy is back working for the Foundation part time, and she's interested in helping release v2 of the book.
21:13:58 <martial> and Mr Hoge can tell us more about this particular topic
21:14:17 <hogepodge> We can do anything we want, as long as the September 29 deadline for final content (no changes at all) is met.
21:14:24 <rbudden> s/or/and/
21:14:50 <martial> and this book would be available at the Sydney meeting as well as SC17
21:15:14 <hogepodge> Definitely SC 17, since Sydney is so close probably that. I'll double check and make sure.
21:15:23 <hogepodge> It may push the deadline up by about a week.
21:15:44 <hogepodge> Kathy is collecting the raw resources for me right now, and I will share them in an etherpad.
21:15:48 <b1airo> we were loosely thinking: update existing content and add a couple of new chapters
21:16:13 <hogepodge> https://etherpad.openstack.org/p/scientific-computing-book-update-2017
21:16:17 <martial> I am reaching out to a couple people to provide extra content if at all possible
21:16:19 <b1airo> e.g one on federation
21:16:25 <hogepodge> #link Working document to organize resources https://etherpad.openstack.org/p/scientific-computing-book-update-2017
21:18:18 <hogepodge> I've added the latest info from Kathy into the doc. It looks like she's planning for a November release, I'll check back for Summit release since we'll have loads of events there too.
21:18:26 <priteau> We could provide an update from Chameleon. Hopefully by then Blazar might have become an official project!
21:18:45 <martial> sweet, thanks
21:18:54 <martial> pierre, looking forward to that section :)
21:18:56 <rbudden> i’m hoping to have some good updates with our skip level upgrade to Ocata
21:19:12 <rbudden> possibly Ironic w/Neutron integration if all goes well!
21:19:30 <priteau> We're going to look into multi-tenant bare-metal networking
21:20:01 <priteau> rbudden: Is that what you want to try, or just plain flat shared network?
21:20:39 <rbudden> still just flat shared network, but we have some use cases if we can use Neutron to give floating IPs to private computes for the duration of a job
21:20:57 <rbudden> i.e. have Slurm prologue talk to Neutron and give a node both ingress/egress
21:21:08 <rbudden> i’m not sure the state of that though
21:21:19 <rbudden> i can let you know in the next couple of days ;)
21:21:42 <martial> robert/all, please update the etherpad with a chapter proposal if at all possible
21:21:50 <b1airo> sounds good!
21:22:02 <martial> anything else to know on this topic Mr Hoge?
21:22:49 <martial> People, feel free to continue extending the etherpad
21:23:07 <martial> #topic Request for a SuperUser guest post - volunteers please
21:23:22 <martial> I think the subject says it all :)
21:24:13 <StefanPaetowJisc> Hello?
21:24:18 <hogepodge> I'll be managing the updates, so expect lots of pushing from me. :-D Thanks everybody!
21:24:45 <martial> Mr Hoge, I expect people ought to add contact information on the Etherpad then?
21:25:00 <hogepodge> yes please
21:25:07 <hogepodge> martial: +1
21:26:00 <b1airo> hi StefanPaetowJisc - long time no see!
21:26:13 <b1airo> your nick appears to be too long for hexchat to auto-complete :-)
21:27:01 <b1airo> or could be because he disconnected - need to increase my font size...
21:27:15 <martial> #link https://etherpad.openstack.org/p/scientific-computing-superuserguestpost-2017
21:27:24 <martial> I created an etherpad for that guest post
21:27:36 <b1airo> martial, do we know what we are looking for?
21:27:51 <b1airo> just anything Science+OpenStack ?
21:28:05 <martial> I do not have more data on this topic I confess
21:28:26 <martial> which is why I created the etherpad to allow people to write a small abstract
21:28:41 <martial> and we can work with the UC to decide what they want/need out of it?
21:28:47 <priteau> Is it one story about one use cases, or one story about many use cases?
21:29:11 <StefanPaetowJi-1> Hello?
21:29:30 <martial> priteau: I apologize, I do not know
21:29:34 <martial> StefanPaetowJi-1: welcome back
21:30:23 <StefanPaetowJi-1> Apologies folks... Connection issues. :-/
21:31:10 <rockyg> boy, do I hear ya
21:32:29 <martial> as I wrote on the etherpad for the proposal: "Please add contact information, the topic (related to Scientific + OpenStack) and abstract that are covered in your potential post"
21:33:15 <martial> hopefully this ought to allow people to come up with --good-- crazy ideas :)
21:33:45 <martial> and last topic for the day
21:34:33 <armstrong> priteau: Asked a question above ==> Is it one story about one use cases, or one story about many use cases?
21:34:51 <martial> #topic User Committee nominations
21:35:37 <martial> pierre: I do not have this information, we will try to get some clarification and update the Etherpad if at all possible when we have it
21:36:29 <martial> is there anyone in our community who might be interested in serving on the User Committee?
21:37:23 <armstrong> I might be interested but want to be sure of how I will be useful
21:37:29 <rbudden> martial: any link to what the job entails?
21:37:32 <armstrong> what are they working on?
21:37:57 <rbudden> +1 i’d be interested in details. I have no idea if I’d even be a fit for helping
21:38:02 <martial> there was a post on the UC ML, Nominations are open till the 11th
21:38:12 <martial> let me find the link
21:38:17 <b1airo> #link https://www.openstack.org/foundation/user-committee/
21:38:26 <rbudden> thx b1airo
21:38:31 <martial> or rely on blair :)
21:38:50 <b1airo> this working group falls under the UC in terms of governance
21:39:31 <armstrong> OK thanks for that information
21:39:55 <b1airo> so the UC members are there to look after the governance side of things for OpenStack User representation and to ensure working groups / teams / SIGs etc are empowered and healthy
21:40:25 <b1airo> the UC also has a seat on the board I think
21:41:03 <b1airo> more detail - #link https://wiki.openstack.org/wiki/Governance/Foundation/UserCommittee
21:41:30 <martial> I am trying to find the original call for candidacy in the ML archive
21:42:11 <martial> (ah here it is, August 4th: "[User-committee] UC candidacy")
21:42:22 <rockyg> look for mrhillsman
21:42:33 <rockyg> he posted it.  And/or Shilla
21:43:03 <b1airo> i think given the high numbers of scientific openstack deployments and broad base of this WG it'd be nice to have continuous representation for the community at that level
21:43:26 <rockyg> So, you gonna run, b1airo ?
21:43:44 <b1airo> zioproto has nominated, so that's one active scientific-wg participant who has put their hand up
21:43:50 <StefanPaetowJisc> Interesting, given that EOSC and things like eMedLab may be large OS deployments
21:44:09 <StefanPaetowJisc> I'll happily second zioproto. He's good with this
21:44:12 <b1airo> rockyg, i would be interested, but i need an overcommitment reality-check
21:44:29 <mrhillsman> heard my name
21:44:44 <b1airo> hi mrhillsman
21:44:51 <rockyg> Original posting of the call for nominations for UC, mrhillsman
21:47:01 <b1airo> i think we can probably move on to AOB...
21:47:13 <b1airo> #topic Other Business
21:47:23 <StefanPaetowJisc> Ahh
21:47:28 <StefanPaetowJisc> I have some other busines
21:47:33 <b1airo> shoot
21:47:37 <StefanPaetowJisc> Some GoodNews(tm)
21:47:55 <b1airo> perhaps we are about to find our SuperUser post...
21:48:08 <StefanPaetowJisc> For El Capitan (and possibly earlier) versions of macOS, we have a barebones Moonshot Mac client
21:48:40 <b1airo> cool! what can it do?
21:48:54 <martial> #topic AOB
21:49:04 <StefanPaetowJisc> It's the *really* basic version and it only works with non-domain-linked Mac systems, but this may be of use for those who want to play with OpenStack and 'Moonshot' authn.
21:49:59 <StefanPaetowJisc> It's embarrassing to say that it's taken me 5 months to get to this point (zioproto and others may recall me not being very 'present' at the Milan mid-cycle ops meetup because I was fighting with Mac's version of Heimdal)
21:50:17 <StefanPaetowJisc> but I'm not a dev, so for me this has been a journey to hell and back.
21:50:37 <b1airo> character building
21:51:25 <StefanPaetowJisc> The hope is that once all the little bits and pieces are resolved, you should be able to do non-web (i.e. CLI) SSO into the OS tools, and possibly OS instances, provided they've been set up for Moonshot AuthN.
21:51:42 <rockyg> a once in a lifetime experience, you hope;-)
21:51:49 <b1airo> that sounds neat for sure
21:51:58 <b1airo> is there already a Keystone extension for this?
21:52:01 <StefanPaetowJisc> By extension, this will help projects like the Fedushare project in the US, that is working on a SAML-ECP version of SSO, which will make a *lot* of people happy
21:52:09 <StefanPaetowJisc> Keystone apparently supports GSSAPI
21:52:36 <b1airo> who knew!?
21:52:39 <StefanPaetowJisc> but whether it supports multitrip GSSAPI authn is another question. That was what I *wanted* to work on the last 5 months after chatting to Adam and others in Barcelona
21:52:55 <rockyg> so, this would make a great forum topic....
21:53:45 <StefanPaetowJisc> But then... Heimdal intervened. And then OpenSSL (macOS only does OSSL 0.9.8 as a backward compatibility thing). They may use LibreSSL... so this will be a featurette in our future stream (to build against LibreSSL as well).
21:54:02 <b1airo> agreed rockyg
21:54:39 <StefanPaetowJisc> Lots of blind alleys. Dark back streets. Bad coding practices. Bad assumptions. And lots of (patient) help from an Australian jazz musician (and macOS aficionado).
21:54:59 <b1airo> i'd have given up and built a container within about 30s i think
21:55:25 <StefanPaetowJisc> Well, the option was to use a VM, yes... (with tunnelling via...)
21:55:41 <b1airo> ha
21:55:43 <StefanPaetowJisc> But that's just playing silly buggers.
21:55:44 <rockyg> If you do a forum topic, with a quick demo to start, you can get more folks to work on it and Keystone folks to help integrate and test, maybe.
21:56:27 <StefanPaetowJisc> Sorry, forum topic? (I'm clueless with some of these things).
21:56:48 <b1airo> #link https://wiki.openstack.org/wiki/Forum
21:56:54 <StefanPaetowJisc> There's an Etherpad *somewhere* (I'll have to dig up the URL) that describes the GSSAPI usecase a bit more...
21:57:08 <StefanPaetowJisc> Ahhhhh
21:57:10 <StefanPaetowJisc> Well...
21:57:12 <rockyg> Forum at the Summit (Sydney).  A good place to provide info, status, and get help
21:57:34 <StefanPaetowJisc> *ahem* I asked my management if I could go to Sydney. The answer was... negative.
21:57:51 <rockyg> If you get a forum session, you can ask for support....
21:57:53 <StefanPaetowJisc> When the next summit/conference comes to Europe, they'll sign off on it :-)
21:58:09 <rockyg> from the foundation, that is.
21:58:12 <hogepodge> With time running out, real-time update from Kathy regarding the book. Deadline is Sept 22. Please put your contact information in the etherpad to help me manage the book production.
21:58:14 <hogepodge> https://etherpad.openstack.org/p/scientific-computing-book-update-2017
21:58:45 <martial> 10-4 hogepodge
21:58:59 <b1airo> hogepodge, we'll make a ML post to try and get attention from folks not here today
21:59:23 <b1airo> let's close this up
21:59:27 <hogepodge> (sorry to topic hijack)
21:59:33 <b1airo> #endmeeting