11:02:02 <oneswig> #startmeeting scientific-sig
11:02:02 <openstack> Meeting started Wed Aug  1 11:02:02 2018 UTC and is due to finish in 60 minutes.  The chair is oneswig. Information about MeetBot at http://wiki.debian.org/MeetBot.
11:02:03 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
11:02:04 <verdurin> Good afternoon.
11:02:05 <openstack> The meeting name has been set to 'scientific_sig'
11:02:16 <oneswig> #link agenda for today https://wiki.openstack.org/wiki/Scientific_SIG#IRC_Meeting_August_1st_2018
11:03:18 <oneswig> I'm in the middle of an "for i in $(openstack baremetal node list..." way today
11:04:20 <verdurin> oneswig: "openstack stack delete" - it's quicker in the long run
11:04:37 <oneswig> yak shaving - stack is DELETE_FAILED
11:04:45 <oneswig> But I take your point!
11:05:06 <priteau> Good morning oneswig, verdurin
11:05:13 <priteau> Sorry, afternoon now
11:05:17 <verdurin> Hello.
11:05:20 <oneswig> Greetings priteau
11:05:31 <oneswig> Interesting meetup?
11:06:07 <priteau> It made me want to try OpenShift :-)
11:06:48 <oneswig> So what does it add over Kubernetes?
11:06:51 <verdurin> Yes, it was good to see actual use cases and demos.
11:08:41 <priteau> oneswig: Main points were a clean object model on top of k8s, deep integration with CI/CD pipelines, and the k8s shipped with it is actually Red Hat's own hardened distribution
11:10:21 <verdurin> Yes, the object and event models, extra debugging and integrated monitoring/logging, proper workflows for handling builds
11:10:24 <b1airo> o/
11:10:29 <oneswig> How did the scientific application work look?
11:10:31 <oneswig> hi b1airo
11:10:37 <oneswig> #chair b1airo
11:10:37 <verdurin> Hello b1airo
11:10:38 <openstack> Current chairs: b1airo oneswig
11:10:38 <b1airo> hi oneswig !
11:10:51 <b1airo> hi verdurin , how's it going?
11:11:57 <priteau> oneswig: There weren't much details about that, just that they're running XChem
11:12:02 <verdurin> b1airo: not bad at all
11:12:33 <verdurin> oneswig: I asked about their use of JupyterHub and he didn't have much more detail. Might be following up with them anyway.
11:12:47 <oneswig> Squonk's the future, I heard :-)
11:13:30 <verdurin> Ugh - Gradle.
11:13:49 <verdurin> oneswig: looking at the HPCAC CFP you sent - I couldn't see a closing date?
11:14:32 <oneswig> I have not heard what it is.  I had an email from the organiser from HPCAC (Cydney), looking for subjects and speakers
11:14:52 <b1airo> ah yes, reminds me i haven't replied to her yet...
11:15:51 <oneswig> b1airo: she mentioned you're in NZ
11:16:03 <martial_> good morning
11:16:10 <oneswig> Hey martial_
11:16:14 <oneswig> #chair martial_
11:16:15 <openstack> Current chairs: b1airo martial_ oneswig
11:16:18 <verdurin> Hello
11:16:45 * b1airo waves at martial_
11:17:34 <b1airo> oneswig: forgive me... Squonk?
11:19:14 <martial_> (did I interrupt something?)
11:20:08 <oneswig> https://squonk.it/ - saw it go by yesterday... from a meetup priteau and verdurin went to
11:23:28 <b1airo> huh, so Jupyter for computational chemists?
11:23:51 <martial_> looking cool
11:24:23 <martial_> although yes, chemistry, so not too sure how to use it
11:25:53 <oneswig> sorry, got distracted.
11:26:49 <martial_> sharing with Bob, it was his former field, he might have some fun
11:28:53 <oneswig> verdurin: will you be at Ceph Days Berlin?
11:29:53 <verdurin> oneswig: if I go to the Summit, which I probably will, then yes. Shame about the clash with SC.
11:30:10 <oneswig> indeed.
11:31:36 <verdurin> Did anyone go to the OpenHPC workshop mentioned recently?
11:32:15 <martial_> not me, was at PEARC18 last week (although had to leave early)
11:32:19 <oneswig> It was in Santa Clara, alas I did not
11:32:38 <oneswig> I've not heard from the presenter from Linaro about how it went
11:33:40 <oneswig> I think even he was presenting remotely
11:33:52 <martial_> lesson learned from PEARC: trying to do tutorials that need access to a remote GPU or Jupyter with bad internet ... not so cool
11:34:04 <oneswig> martial_: what happened?
11:34:25 <martial_> not much, the internet was bad at the hotel of the conference
11:34:49 <martial_> and as such running the tutorials was near impossible
11:35:17 <martial_> (a lot of frustrated people through day 1)
11:37:24 <oneswig> ouch, feel your pain
11:37:49 <b1airo> shame
11:38:13 <martial_> it was not cool indeed
11:38:52 <martial_> but our panel on "Transforming Scientific Computing with Cloud Technologies" went well
11:39:09 <martial_> we used the Etherpad just like we did for the Summit in Vancouver
11:39:14 <oneswig> Who attends PEARC?
11:39:24 <oneswig> It's an NSF conference isn't it?
11:39:55 <martial_> ACM
11:40:02 <b1airo> couple of NeSI management went i think
11:40:12 <martial_> by funding from NSF
11:40:24 <martial_> it used to be Xseed conference I think
11:40:39 <b1airo> what was maturity of the conversation like martial_ ?
11:40:40 <martial_> turned into a small SC, recognize a lot of the same people
11:40:49 <martial_> it was good
11:40:53 <martial_> #link https://etherpad.openstack.org/p/pearc18-panel
11:41:03 <martial_> the etherpad used ^
11:41:41 <martial_> it was little focused on pure OpenStack, but it was prominent in conversation
11:42:13 <martial_> that and containers root
11:42:37 <martial_> and for some reason a person (who did not stop by after) wanted to talk kubernetes a lot
11:43:03 <martial_> and I was the only person with some use cases of the tech
11:44:10 <b1airo> time to shine :-)
11:44:42 <martial_> OpenStack is still very popular in HPC and our panelists were able to give many HPC related use cases
11:44:56 <martial_> we mentioned the Barcelona talk from the Spartan team as well
11:46:25 <martial_> Etherpad is the format I want to use for SC18
11:46:51 <Redfoxmoon2> With our IRC ad service you can reach a global audience of entrepreneurs and fentanyl addicts with extraordinary engagement rates! https://williampitcock.com/
11:46:51 <Redfoxmoon2> I thought you guys might be interested in this blog by freenode staff member Bryan 'kloeri' Ostergaard https://bryanostergaard.com/
11:47:24 <oneswig> interesting interlude
11:47:45 <martial_> get in ... ad ... get out ... okay
11:48:31 <martial_> wonder if those are malware of cryptominers
11:49:00 <priteau> It's a spam campaign, there are mailing list post on openstack-dev
11:49:05 <oneswig> martial_: should we produce some input for the PTG session?
11:49:35 <martial_> we ought to, I have not yet heard when our session will be
11:49:43 <martial_> but I am scheduled to go
11:49:54 <oneswig> something to discuss also in the US timezone.
11:50:34 <oneswig> Right now, I would love to ask if anyone at the PTG ever managed to successfully delete a heat stack with bare metal compute in it :-(
11:50:51 <oneswig> Seems to be no improvement in Queens
11:51:36 <priteau> martial_: You didn't see the PTG schedule?
11:51:55 <martial_> we ought to set up an etherpad and use the scientific_wg IRC channel during the PTG
11:51:59 <priteau> Scientific SIG is Monday: https://www.openstack.org/ptg#tab_schedule
11:52:13 <martial_> pierre: thanks
11:52:20 <oneswig> In terms of project activity - priteau, what about an update on preemptibles, reservations and all?
11:52:22 <priteau> oneswig: I do it all the time...
11:52:39 <priteau> (the bare-metal Heat stack deletes)
11:53:00 <oneswig> would love to wipe this one out, with extreme prejudice...
11:53:57 <priteau> Regarding pre-emptibles: we're discussing in the Blazar project how we could fairly easily implement preemptible instances, by 1) allowing instances to be launched on nodes managed by Blazar but currently unreserved 2) terminating anything that is running at the start of a reservation
11:54:17 <priteau> We're going to work on a spec, there's also a talk in Berlin including discussion of this topic
11:54:32 <priteau> It could work independently of any Nova reaper implementation
11:54:48 <martial_> maybe we can make the PTG the topic of next week's meeting
11:54:56 <martial_> stig: got an etherpad started?
11:55:18 <oneswig> martial_: sorry, no - and I'm on vacation for the next two weeks
11:55:58 <priteau> That's the talk in question: https://www.openstack.org/summit/berlin-2018/vote-for-speakers/#/22073
11:58:49 <oneswig> somehow we are out of time!
11:59:05 <oneswig> priteau: that talk looks good, fingers crossed for them
11:59:11 <b1airo> priteau: sounds like the Blazar functionality could be complementary to Nova Reaper
11:59:32 <martial_> #link https://etherpad.openstack.org/p/scientific-sig-denverptg2018
11:59:37 <martial_> here started one
11:59:53 <oneswig> thanks martial_ - let's try to gather ideas there
12:00:04 <oneswig> Is it a half-day?
12:00:18 <priteau> b1airo: yes, I hope the reaper makes progress at the next PTG
12:00:44 * b1airo crosses fingers
12:00:44 <oneswig> It's a compelling capability, it really should do
12:01:06 <oneswig> OK, we should finish.  Thanks all - see you in a few weeks
12:01:11 <oneswig> #endmeeting