=== meetingology` is now known as meetingology
[14:34] <smoser> mgerdts: so for bug 1763511
[14:34] <ubot5`> bug 1763511 in cloud-init (Ubuntu) "DataSourceSmartOS should default to ext4" [Undecided,New] https://launchpad.net/bugs/1763511
[14:34] <smoser> just curious
[14:35] <smoser> you have disks > 116 TiB
[14:35] <smoser> oops. 16 TiB
[14:35] <smoser>  ?
[14:37] <mgerdts> Yeah, we are starting to need some big disks.
[14:37] <mgerdts> :)
[14:38] <mgerdts> Whether that is the optimal storage layout or not is another question.
[14:38] <mgerdts> On the backend it is zfs, spread across many spindles
[14:40] <smoser> that was my next question.
[14:40] <mgerdts> I guess a TiB isn't what it used to be.
[14:41] <smoser> 16TiB ought to be enough for anyone.
[14:41] <mgerdts> VP of engineering here disagrees.
[14:42]  * smoser was adjusting a quote of a former CEO at microsoft
[14:42] <mgerdts> Yeah, a few words into my next reply, I caught that.  :)
[14:43] <mgerdts> 16 TiB is the new 640 KiB
[14:43] <mgerdts> Time to make a t-shirt.
[14:44] <mgerdts> And refill the coffee cup
[15:25] <mgerdts> smoser: now that I have your approval on a couple merge proposals, what's next for them?
[15:33] <smoser> i'll get them pulled in today.
[15:39] <mgerdts> great, thanks.  Just wasn't sure if there was any process step that I needed to complete first.
[15:57] <paulgrmn> Has anyone been able to get the nocloud data source to work with the OpenStack qcow2 images provided by Debian? Do they not support nocloud?
[16:01] <smoser> paulgrmn: https://asciinema.org/a/132009
[16:03] <paulgrmn> I followed that, but can't get it to work with the latest images. I also don't see cloud-init looking for the NoCloud datasource on the console during virt-install.
[16:05] <smoser> paulgrmn: hm.. well, it worked in 20170725
[16:05] <smoser> i havent tried since. its possible they disabled that... but one woudl hope not
[16:05] <paulgrmn> Yeah, that's what I was wondering --- if it was a regression or if I'm goofing up something.
[16:07] <smoser> paulgrmn: you used https://cdimage.debian.org/cdimage/openstack/current/debian-9.4.3-20180416-openstack-amd64.qcow2
[16:07] <smoser>  ?
[16:07] <paulgrmn> I would assume my error if I saw it try "DatasourceNoCloud" and fail, but it doesn't seem to even look for it.
[16:08] <paulgrmn> I used https://cdimage.debian.org/cdimage/openstack/current-9/debian-9-openstack-arm64.qcow2
[16:09] <smoser> arm ?
[16:09] <paulgrmn> Derp. No. amd64
[16:09] <paulgrmn> https://cdimage.debian.org/cdimage/openstack/current-9/debian-9-openstack-amd64.qcow2
[16:11] <paulgrmn> Which has the same MD5sum as the one you linked.
[16:13] <smoser> alright. i just verified it "worked for me"
[16:13] <paulgrmn> OK. Thanks a lot. Must be something I'm doing wrong.
[16:13] <smoser> here.
[16:14] <smoser> git clone https://gist.github.com/smoser/635897f845f7cb56c0a7ac3018a4f476 boot-test
[16:14] <smoser> cd boot-test
[16:14] <smoser> ./check-dependencies
[16:14] <smoser> cloud-localds -v --network-config=network-config-v1.yaml            seed.img user-data.yaml meta-data.yaml
[16:15] <smoser> ./boot ../debian-9.4.3-20180416-openstack-amd64.qcow2 seed.img -snapshot
[16:15] <smoser> then i logged in as 'debian' and 'passw0rd'
[16:16] <paulgrmn> Thanks for your help. I'll give it a try.
[16:21] <smoser> h
[16:21] <smoser> blackboxsw: https://code.launchpad.net/~chad.smith/cloud-init/+git/cloud-init/+merge/342428 didnt like your code.
[16:23] <smoser> and rharper had mentioned he used timit on the socket.inet_ntoa
[16:23] <smoser> i'd prefer  not to have 2 implementations of that
[16:23] <smoser> (in annother of my coments there)
[16:23] <smoser> regex...
[16:24] <smoser> i'll be back in ~ 90 minutes and review more.
=== r-daneel_ is now known as r-daneel
[17:05] <blackboxsw> hab smoser , I need to rebase
[17:05] <blackboxsw> pushing now
=== r-daneel_ is now known as r-daneel
[17:49] <blackboxsw> <smoser> and rharper had mentioned he used timit on the socket.inet_ntoa.   Yeah I addresssed that comment and pushed the common logic into cloudinit.net.network_state.net_prefix_to_ipv4_mask. So there is no cloudinit.netinfo.netdev_cidr_to_mask anymore
[17:49]  * blackboxsw runs this branch on ec2 and openstack just to see that the output looks good
[18:15] <rharper> nice
[18:26] <smoser> blackboxsw: oh. good. thanks. sorry. i didnt notice.
[18:27] <smoser> err.. read that wrong
[18:51] <blackboxsw> +1, the branch kinda started to get bigger than I'd prefer. ohh well.
[18:55] <smoser> its really not so bad. you did a good job
=== r-daneel_ is now known as r-daneel
[19:07] <smoser> blackboxsw: i posted some feedback there.
=== r-daneel_ is now known as r-daneel
[21:33] <blackboxsw> smoser: rharper wrapping up feedback. will have something up tonight on this thanks
[22:36] <smoser> blackboxsw: i'll be back in to check later.