| dtantsur | TheJulia: not sure what mercury is here, but no problems with optionally passing a session to client calls. | 08:01 |
|---|---|---|
| opendevreview | Takashi Kajinami proposed openstack/ironic master: doc: Fix wrong example of [ipmi] cipher_suite_versions https://review.opendev.org/c/openstack/ironic/+/918158 | 09:39 |
| TheJulia | dtantsur: the networking service :) | 13:16 |
| dtantsur | ahh, a new name :) | 13:42 |
| dtantsur | TheJulia: the downside of creative names: we probably need to run it by some trademark people (in the foundation?) | 13:43 |
| TheJulia | well, not approved by any sense and we're just in consensus building phase, but it should be fairly safe in the grand scheme at least for us to do some prototyping | 13:44 |
| dtantsur | yeah, thinking mid-term | 13:44 |
| TheJulia | I've already found an awful thing I didn't spot in n-g-s which makes me want to throw things | 13:44 |
| TheJulia | actually, two awful things, one less so | 13:44 |
| dtantsur | sounds intriguing | 13:44 |
| TheJulia | three | 13:44 |
| TheJulia | one is a remotable method over rpc | 13:44 |
| dtantsur | oh no, they're multiplying | 13:44 |
| TheJulia | lol | 13:45 |
| TheJulia | oh, and yes, I totally agree in midterm | 13:45 |
| TheJulia | next is the ml2 plugin does a remotable DB change | 13:45 |
| TheJulia | (like, seriously, why?! at least it is in super limited cases which likely means a "client" ml2 library could do the needful for neutron compatability) | 13:45 |
| dtantsur | ... this all sounds like a really robust design ... | 13:45 |
| TheJulia | (so this all might boil down more into "attach/detach", but have to wrap a brain around exactly what happens across that layer) | 13:46 |
| TheJulia | well, attach/detach/please_validate_x/please_validate_y, etc | 13:46 |
| TheJulia | well, so the are not remotable now, they execute inside of the neutron process | 13:48 |
| TheJulia | anyway, need to eliminate them for this to be viable | 13:48 |
| dtantsur | yeah, the RPC layer will probably be designed from scratch (which is what your spec is doing IIUC) | 13:49 |
| TheJulia | sort of yeah | 13:51 |
| TheJulia | yeah, sort of what I'm doing since I started to design the client :) | 13:51 |
| TheJulia | I figured, start with the client and decompose the basic requirements and interactions out of the known needed methods and sort of go from there | 13:52 |
| dtantsur | That's probably the way to go since Ironic is where the requirements actually begin | 13:53 |
| TheJulia | yup | 13:55 |
| TheJulia | and really, our interactions are really simple for what we *need* | 13:55 |
| TheJulia | it is how to "do the needful" that is the hard part | 13:55 |
| * TheJulia posts a new comment to the spec and gets ready to go get breakfast | 13:56 | |
| TheJulia | ... and the hardware store to buy a tool for the wife's project and then groceries.... | 13:56 |
| TheJulia | (do the needful, and provide a cleanish path a client plugin could pickup and use) | 13:57 |
| * dtantsur needs to pack stuff for camping next week | 14:37 | |
| TheJulia | Enjoy! | 14:37 |
| dtantsur | Thanks :) | 14:41 |
| dtantsur | I hope the weather cooperates: Black Forest in May is a risky bet, I must admit | 14:41 |
| TheJulia | I imagine it could be very beautiful | 14:47 |
| dtantsur | TheJulia: yep, walking distance from https://duckduckgo.com/?q=titisee&t=ffab&iax=images&ia=images | 14:48 |
| TheJulia | Very nice | 14:50 |
| dtantsur | Unfortunately, the water temperature probably doesn't allow kayaking (nor do we have space in our car to take also the kayak lol) | 14:51 |
| dtantsur | but plenty of hiking | 14:51 |
| TheJulia | Well, enjoy! Time for breakfast and then to go get tools to begin assembling the forge. | 14:52 |
| dtantsur | Yep, won't distract you more. Enjoy your meals! | 14:52 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!