| *** jkilpatr has quit IRC | 00:16 | |
| *** crushil has joined #openstack-cyborg | 01:24 | |
| *** crushil has quit IRC | 05:13 | |
| *** crushil has joined #openstack-cyborg | 06:18 | |
| *** crushil has quit IRC | 10:39 | |
| *** jkilpatr has joined #openstack-cyborg | 11:56 | |
| jkilpatr | zhipengh[m], so the agent can't interact with placement well IMO. Each agent is on each compute and knows only about it's own local resources. If we want the agent to go and tell Nova 'hey we have N of these devices' it needs to know global state | 13:27 |
|---|---|---|
| jkilpatr | which is the job of the conductor. | 13:27 |
| jkilpatr | Otherwise we assume that no two computes will have the same type of accelerator, which is fatally flawed IMO | 13:27 |
| zhipengh[m] | Does it have to know the global state ? | 13:41 |
| zhipengh[m] | Currently for Nova, it is the Nova compute you n each node interacting with Placement, as far as I understand | 13:42 |
| *** crushil has joined #openstack-cyborg | 14:49 | |
| *** crushil has quit IRC | 15:14 | |
| *** crushil has joined #openstack-cyborg | 15:24 | |
| *** crushil has quit IRC | 17:08 | |
| *** crushil has joined #openstack-cyborg | 17:09 | |
| *** goldenfri has joined #openstack-cyborg | 17:45 | |
| *** crushil has quit IRC | 19:55 | |
| *** crushil has joined #openstack-cyborg | 20:25 | |
| *** ttk2[m] has quit IRC | 21:51 | |
| *** zhipengh[m] has quit IRC | 21:59 | |
| *** ttk2[m] has joined #openstack-cyborg | 22:11 | |
| *** zhipengh[m] has joined #openstack-cyborg | 22:12 | |
| *** crushil has quit IRC | 23:16 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!