Saturday, 2018-10-27

TheJuliaSkyraven: reading, sorry got distracted by a spec document00:01
TheJuliaSkyraven: yeah, the ipmi driver's "boot to network" is an explicit "boot bios mode to network" as uefi is not in the ipmispec00:01
TheJuliaipmi spec, it is done by a raw command sequence to the bmc... or as a different boot device00:02
TheJuliayeah, that makes sense then because bios mode. ilo inspection is also out of band so you won't have as much information, but it should be correct00:02
*** rcernin has joined #openstack-ironic00:03
TheJuliaSkyraven: If I'm remembering correctly, it would just be the ipmi driver doing it with how ipmi works :\00:03
* TheJulia will one day figure out what time zone Moshe live sin00:05
TheJuliafellow ironic cores: regarding smartnics: the spec at https://review.openstack.org/#/c/582767/11 has been updated. I think it accurately describes the problems and provides a really novel/ingenious solution.00:06
patchbotpatch 582767 - ironic-specs - Add Support for Smart NIC - 11 patch sets00:06
*** jesusaur has joined #openstack-ironic00:18
*** rcernin has quit IRC00:39
SkyravenTheJulia: all clear then :) to be honest I also left part of the story out..I'm installing Contrail which comes prepackaged with Queens and it has a bunch of ansible templates that trigger the Director. Those are also setting the ipmi driver but of course, I can change it to iLO01:13
SkyravenI just was not aware of the limitations and got a bit confused by reading as to where this behavior is coming from and if normal01:13
Skyraventhank you for the clarification :)01:13
SkyravenI have a dumb question though :) if introspect does everything in this case in bios legacy boot mode, then I set it via capabilities for deploy to uefi, is there any risk that the drives/disks are seen differently numbered (sda, sdb, etc) between introspect and deploy ?01:17
*** MattMan_ has quit IRC01:48
*** MattMan_ has joined #openstack-ironic01:48
TheJuliaSkyraven: yeah, it is a bit... I can't think of the right word... odd knowledge to have02:06
TheJuliaSkyraven: I'm not aware of any such issues because the drive lettering typically falls back to driver initialization in the kernel02:07
TheJuliaat least, once the kernel is loaded02:07
TheJuliadifferent between the two... unlikely as long as the kernels are similar or the same02:07
*** dsneddon has quit IRC03:44
*** dsneddon has joined #openstack-ironic03:46
*** dsneddon has quit IRC03:51
*** dsneddon has joined #openstack-ironic03:57
*** dsneddon has quit IRC04:05
*** dsneddon has joined #openstack-ironic04:42
*** dsneddon has quit IRC04:48
*** skazi has joined #openstack-ironic04:50
*** dsneddon has joined #openstack-ironic05:11
*** skazi has quit IRC05:33
*** dsneddon has quit IRC05:36
*** dsneddon has joined #openstack-ironic05:36
*** mcarden has joined #openstack-ironic06:02
*** sthussey has quit IRC06:04
*** gkadam has joined #openstack-ironic06:10
*** toabctl has joined #openstack-ironic06:15
*** pcaruana has joined #openstack-ironic08:30
*** e0ne has joined #openstack-ironic08:52
*** e0ne_ has joined #openstack-ironic08:56
*** e0ne has quit IRC08:58
SkyravenTheJulia: thank you very much for all the help and clarifications:)09:14
*** Skyraven has quit IRC09:15
*** e0ne_ has quit IRC09:39
*** rcernin has joined #openstack-ironic09:54
*** rcernin has quit IRC11:13
*** skazi has joined #openstack-ironic11:27
*** skazi has quit IRC11:35
*** gkadam has quit IRC14:19
*** dsneddon has quit IRC14:58
*** e0ne has joined #openstack-ironic16:39
*** skazi has joined #openstack-ironic16:46
*** e0ne has quit IRC16:53
*** bfournie has quit IRC17:11
*** jezogwza has quit IRC17:41
*** dsneddon has joined #openstack-ironic18:03
*** bfournie has joined #openstack-ironic19:50
*** jesusaur has quit IRC22:27
*** pcaruana has quit IRC22:33
*** rnoriega has quit IRC23:38
*** dsneddon has quit IRC23:47
*** dsneddon has joined #openstack-ironic23:55

Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!