| *** openstackgerrit has joined #openstack-valence | 01:18 | |
| openstackgerrit | Anusha Ramineni proposed openstack/valence master: Update podmanager controller to use generic interface https://review.openstack.org/499468 | 03:56 |
|---|---|---|
| openstackgerrit | Anusha Ramineni proposed openstack/valence master: WIP: Ironic generic https://review.openstack.org/505488 | 04:06 |
| openstackgerrit | Akhil jain proposed openstack/valence master: Add new 'device' DB table to valence https://review.openstack.org/498395 | 04:47 |
| *** openstackgerrit has quit IRC | 07:02 | |
| *** openstackgerrit has joined #openstack-valence | 14:00 | |
| *** ntpttr_mobile has joined #openstack-valence | 14:24 | |
| *** ramineni_ has joined #openstack-valence | 14:26 | |
| *** shuquan has joined #openstack-valence | 14:29 | |
| *** akhill_jain has joined #openstack-valence | 14:30 | |
| lin_yang | hello everyone | 14:31 |
| ntpttr_mobile | Hi Lin | 14:31 |
| akhill_jain | Hello | 14:31 |
| ramineni_ | lin_yang: ntpttr_mobile: hi | 14:31 |
| shuquan | hi | 14:32 |
| ntpttr_mobile | So, I saw that Lin had a first email for something he'd like to discuss, Lin would you like to start? | 14:33 |
| *** ntpttr_laptop has joined #openstack-valence | 14:34 | |
| ramineni_ | lin_yang: will you be able to make it to sydney for presentation? | 14:34 |
| lin_yang | sure, ramineni_ is here, so we can discuss more details about our session in summit | 14:34 |
| lin_yang | ramineni_: sorry i will not be there, you know, the budget reason | 14:34 |
| ntpttr_laptop | ah yeah that reminds me, I got word last week that I won't be going to Sydney | 14:35 |
| ramineni_ | lin_yang: oh ok :( | 14:35 |
| ntpttr_laptop | but mkrai I believe will be there | 14:35 |
| shuquan | lin_yang: you will be there, right? | 14:35 |
| lin_yang | ramineni_: Madhuri will be there, so if you need any help we can discuss with her | 14:36 |
| ramineni_ | hopefully myself or akhil would be able to make it, if budget permits | 14:36 |
| lin_yang | shuquan: Bad news, I will not be there. How about you? | 14:36 |
| ramineni_ | lin_yang: ok, sure | 14:36 |
| shuquan | i think i should be there. | 14:37 |
| ntpttr_laptop | glad to hear it shuquan | 14:37 |
| ntpttr_laptop | I guess Sydney is too far/expensive for us this time | 14:38 |
| shuquan | agree | 14:38 |
| lin_yang | that's true | 14:38 |
| *** Mrittika has joined #openstack-valence | 14:38 | |
| ntpttr_laptop | hi Mrittika | 14:38 |
| lin_yang | shuquan: My manager JF will be there, so if you have any topic, you can discuss with him | 14:38 |
| lin_yang | hi Mrittika | 14:39 |
| Mrittika | Hi | 14:39 |
| ramineni_ | Mrittika: Hi | 14:39 |
| lin_yang | ramineni_: for the presentation, do you start to prepare the slides? I am happy to help it | 14:39 |
| Mrittika | Is anyone doing a demo as part of the talks? | 14:39 |
| ntpttr_laptop | so shuquan, I also noticed that since our talk is a lightning talk, we'll only have 10 minutes. I doubt that will really leave time for a demo, unless you want to do the demo while discussing instead of slides? | 14:39 |
| ntpttr_laptop | might make for a more interesting talk I guess if there's something going on like that for a quick lightning talk rather than 4 or 5 slides | 14:40 |
| Mrittika | Demos are generally not visible in the 10 min talks | 14:40 |
| shuquan | you're right. | 14:40 |
| ntpttr_laptop | ah okay, didn't know that Mrittika | 14:40 |
| ramineni_ | lin_yang: sure, thanks.. havent started yet.. planning to start next week, ill send out the initial slides.. we can discuss more and take it from there? | 14:40 |
| Mrittika | They cannot show your screen | 14:40 |
| shuquan | maybe only some slides and a quick demo video | 14:41 |
| lin_yang | ramineni_: sounds great! Thanks. | 14:41 |
| ramineni_ | lin_yang: Mrittika: I think we can plan for demo for our talk, as its 40 mins.. we will have enough time i suppose | 14:41 |
| ntpttr_laptop | Mrittika: if they can't show screen, can we not even have slides for the 10 minute one? | 14:42 |
| Mrittika | Cool! ramineni_ if you prepare and send we can review | 14:42 |
| ramineni_ | hopefully, i would like .. the device orchestration part also to be ready by that time.. | 14:42 |
| Mrittika | ntpttr_laptop: you can have slides | 14:43 |
| ramineni_ | Mrittika: great, thanks | 14:43 |
| Mrittika | You have to send slides to them | 14:43 |
| ntpttr_laptop | gotcha | 14:43 |
| Mrittika | If you embed a video in the slide which auto runs then that will work | 14:43 |
| shuquan | ramineni_: I can aslo help to review it | 14:43 |
| ramineni_ | shuquan: great , thanks .. will send out email once done | 14:44 |
| ramineni_ | ntpttr_laptop: i have 2 more small topics to discuss, if we have time | 14:45 |
| *** Mrittika_ has joined #openstack-valence | 14:45 | |
| ntpttr_laptop | ramineni_: yes, please | 14:45 |
| lin_yang | ramineni_: one concern from our side is the abstract mentioned attach GPGPU and FPGA to composed node | 14:45 |
| ramineni_ | lin_yang: yes | 14:46 |
| Mrittika_ | What is the concern | 14:46 |
| lin_yang | ramineni_: but Intel don't decide when podm will support it | 14:46 |
| *** ntpttr_mobile has quit IRC | 14:46 | |
| shuquan | ramineni_: lin_yang: I only tried nvme before | 14:46 |
| ramineni_ | lin_yang: they can be attached as PCI devices right | 14:47 |
| ramineni_ | if PCI device support is there, it should be possible? | 14:47 |
| lin_yang | ramineni_: So NEC already support it? | 14:47 |
| shuquan | i saw some hardware vendor implement GPGPU | 14:47 |
| ramineni_ | yes, i think so, it can support all kind of PCI devices, thats the reason added it | 14:48 |
| *** Mrittika has quit IRC | 14:48 | |
| ramineni_ | heard they can be dynamically attached after node is composed..according to requirements | 14:48 |
| ramineni_ | workload | 14:49 |
| lin_yang | ramineni_: yeah, it should be possible, but Intel podm don't offically support it, so marketing folks don't want us to deliver wrong message for it | 14:49 |
| lin_yang | but it's fine if OEC support it | 14:49 |
| ntpttr_laptop | I've seen it with NVMe, haven't tried it with others but for device endpoints I've seen entity_type field, for ours is only NVMe drive possible there lin_yang? | 14:49 |
| ramineni_ | lin_yang: oohok | 14:49 |
| Mrittika_ | But it believe ramineni_can present if you can show it working | 14:49 |
| lin_yang | ntpttr_laptop: I think so | 14:50 |
| Mrittika_ | You can specifically talk about ur use case ramineni_ | 14:50 |
| lin_yang | Mrittika_: totally agree | 14:50 |
| ntpttr_laptop | yeah, agree with Mrittika_. Sounds good | 14:50 |
| ramineni_ | Mrittika_: sure , will note that point while preparing slides, and also will double check it, thanks for bringing that up | 14:51 |
| Mrittika_ | Mention that it is possible with the controller and here is the benefit | 14:51 |
| ramineni_ | Mrittika_: ok | 14:52 |
| ntpttr_laptop | ramineni_: what were the other topics you wanted to discuss? | 14:53 |
| ramineni_ | as we are now removing the CONF options for podm credentials, im updating ironic driver to take from podm DB, i found i think we might need to accept some additional parameters while creating podmanager, just need opinion on it | 14:53 |
| ramineni_ | https://github.com/openstack/valence/blob/master/valence/provision/ironic/driver.py#L55 | 14:53 |
| ramineni_ | like in ironic driver info, i see some additional parameters are taken from CONF, CONF.podm.verify_ca | 14:54 |
| ntpttr_laptop | ah yeah that makes sense, is verify_ca not currently accepted? | 14:54 |
| ramineni_ | no | 14:54 |
| ntpttr_laptop | also just so you know, we're looking into submitting an ironic driver for RSD-Lib | 14:54 |
| ntpttr_laptop | so that /Nodes can be manipulated by ironic | 14:55 |
| ramineni_ | ok, nice.. so how the call will be then | 14:55 |
| ramineni_ | directly calls rsd_lib for creating ironic node | 14:55 |
| ramineni_ | ? | 14:55 |
| ramineni_ | from controller? | 14:56 |
| ntpttr_laptop | on the ironic plugin side in valence it'll be almost hte same, only instead of driver='redfish' it'll say driver='rsd' or something similar | 14:56 |
| ntpttr_laptop | the difference will be what ironic itself sees when communicating with RSD hardware | 14:56 |
| ramineni_ | im thinking something like this .. to make it used by multi-podm https://review.openstack.org/#/c/505488/1/valence/provision/ironic/driver.py | 14:57 |
| ramineni_ | does that look ok? | 14:57 |
| ntpttr_laptop | I'll have to take a more detailed look but at first glance it does look ok | 14:57 |
| ramineni_ | ntpttr_laptop: ok, sure, | 14:58 |
| lin_yang | okay this is the driver in valence to register node to ironic, right? It make sense to me | 14:58 |
| ntpttr_laptop | the idea is definitely good | 14:58 |
| ramineni_ | lin_yang: yes | 14:58 |
| ramineni_ | does it makes sense for podm to accept additional parameters while creation like verify_ca etc? | 14:59 |
| ntpttr_laptop | yes, but they should have defaults | 14:59 |
| ramineni_ | or do we need to pass while we run command registering node with ironic | 14:59 |
| ntpttr_laptop | veryfy_ca can default to True maybe | 15:00 |
| ntpttr_laptop | or wait | 15:00 |
| ntpttr_laptop | I'm going off of what Sushy was doing | 15:00 |
| ntpttr_laptop | https://github.com/openstack/sushy/blob/master/sushy/main.py#L51 | 15:01 |
| ramineni_ | ntpttr_laptop: ok | 15:01 |
| lin_yang | maybe we can follow the same way to implement it, it looks good to me | 15:02 |
| ntpttr_laptop | anything else ramineni_ :)? | 15:02 |
| ramineni_ | lin_yang: ntpttr_laptop: thanks , will check | 15:02 |
| ramineni_ | ntpttr_laptop: yes, :) https://review.openstack.org/#/c/505099/1 | 15:03 |
| ramineni_ | small one.. | 15:03 |
| ramineni_ | lin_yang: i remember we added the updating with global requirements before for both valenceclient/valence | 15:03 |
| ramineni_ | but i dont see those patches for valence, but valenceclient works fine | 15:04 |
| ntpttr_laptop | i'll give it a look ramineni_ ! | 15:04 |
| lin_yang | ramineni_: yes, but didn't get any update from global requiements recently, right? | 15:04 |
| lin_yang | ramineni_: okay, I will take a look at it | 15:04 |
| lin_yang | it's wired | 15:04 |
| ramineni_ | like this.. https://review.openstack.org/#/c/495270/ | 15:05 |
| ramineni_ | ntpttr_laptop: lin_yang: great, thanks.. i took a look at infra.. it includes both projects though | 15:05 |
| ramineni_ | ntpttr_laptop: im done from my side | 15:05 |
| ramineni_ | :) | 15:05 |
| ntpttr_laptop | okay, thanks ramineni_ :) I'll review your patches | 15:06 |
| ramineni_ | ntpttr_laptop: thanks, all updated with comments.. @all please have a look :) | 15:07 |
| lin_yang | ramineni_: will do it today, thanks a lot | 15:08 |
| ramineni_ | lin_yang: thanks!! | 15:08 |
| lin_yang | ntpttr_laptop: we are done here today? | 15:09 |
| ntpttr_laptop | yeah I think so | 15:09 |
| ntpttr_laptop | thanks guys | 15:09 |
| ntpttr_laptop | have a good one! | 15:09 |
| shuquan | 88 | 15:09 |
| ramineni_ | Thanks All.. Bye :) | 15:09 |
| lin_yang | thanks all, bye | 15:10 |
| *** shuquan has quit IRC | 15:10 | |
| Mrittika_ | Bye | 15:10 |
| *** Mrittika_ has quit IRC | 15:10 | |
| *** ntpttr_laptop has quit IRC | 15:15 | |
| *** akhil_jainn has joined #openstack-valence | 15:16 | |
| *** ramineni_ has left #openstack-valence | 15:16 | |
| *** akhill_jain has quit IRC | 15:19 | |
| *** akhil_jainn has quit IRC | 15:43 | |
| *** ntpttr_laptop has joined #openstack-valence | 16:21 | |
| openstackgerrit | Merged openstack/valence master: Update Rack and Systems controller to be generic https://review.openstack.org/501973 | 17:26 |
| *** ntpttr_laptop has quit IRC | 18:42 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!