[00:57] babbageclunk: here's a small gomaasapi https://github.com/juju/gomaasapi/pull/81 [00:57] wallyworld: looking [01:00] wallyworld: approved! [01:00] ty [02:53] babbageclunk: it looks like names for vapps must be globally unique [02:53] ? [02:53] weird [02:54] so when you clone one you give it a uuid for a name? [02:55] it's possible (probable?) that i'm not specifying the path properly [02:57] timClicks: yeah, it would make more sense that it needs to be unique within the resource pool [03:20] babbageclunk: https://github.com/juju/names/pull/92 [03:22] wallyworld: looking [03:37] wallyworld: approved - sorry, got sidetracked [03:39] babbageclunk: ty, no worries [04:22] kelvinliu_: whenever you get a chance https://github.com/juju/juju/pull/10510 [04:22] no rush [04:22] yep [04:34] wallyworld: lgtm thanks! [04:46] timClicks: any luck with cloudinit/vsphere? [04:48] wallyworld: progress, but not completion - I'm still getting naming conflicts [04:57] timClicks: you're not blocked? ie no need to ask for advice on their slack channel? [05:07] wallyworld: kelvinliu_ : add-k8s UX PTAL https://github.com/juju/juju/pull/10511 [05:08] o [05:08] k [05:08] :) [05:08] dunno what to think of that response [05:47] anastasiamac: did you want to discuss the k8s host cloud behaviour? [05:49] wallyworld: yes but m out of time now... [05:49] do u want to do it late-ish or tomorrow [05:50] no worries, tomorrow is fine. i have soccer in about 3 hours [05:50] k [05:51] actually i could call u from the car in about 20/30... does that work [05:51] ? [05:51] wallyworld: ^^ [05:52] sure [05:52] \o/ [14:51] manadart: review pls? https://github.com/juju/juju/pull/10512 [14:51] manadart: there is an add-subnet featuretest left to be resolved from the subnet ID change. [15:02] manadart: have you played with the networking common cache? [15:03] looks like a different idea of a cache [15:04] hml: I have not written code against it, but I saw it while reworking stuff into core/network. [15:04] manadart: yeah.. the code is juju 1.x [15:05] wondering if we should keep it or not [15:06] manadart: do we have plans for BackingSubnetInfo in the long run? [15:06] hml: It looks like a lot of optimisation effort for a not-particularly-hot path. [15:06] manadart: agreed. [15:10] hml: I noted the comments against Backing[Space|Subnet]Info too. [15:10] hml: I think we audit the benefit of the stuff in networking/common. If we can throw it away, the backing types go too. [15:11] manadart: that’d be nice. [15:11] If we keep it, I think we should look at replacing them with the core xInfo types. [15:11] it maybe giving me grief with one of feature tests and the id change [15:11] there’s little overlap between the two SubnetInfos [15:25] hml: Just doing QA on your patch. Added a card to audit networkingcommon. [15:26] manadart: rgr [15:32] hml: Approved. Going EoD too. [15:32] * manadart waves. [15:32] manadart: ty and g'night [21:26] timClicks: just wanted to check vsphere status before release call... how's it going? [21:33] wallyworld: I'm still working my way through incompatibilities between vapps and vms [21:34] wallyworld: debugging is *very* slow because I need to stream the vmdk every time, which adds about 30mins of latency to every code change [21:53] timClicks: you could hardcode the vmdk path to not be based on the controller uuid and then just reuse it each time. [21:53] (and comment out the deletion too) [21:54] that's a good suggestion [21:58] babbageclunk: do you have a few minutes? [21:58] timClicks: yup [21:58] in standup?