| === tds9 is now known as tds | ||
| meena | Odd_Bloke: i see you've started using types! yay! … uhm, what do we get from that? do we have type checkers in the test system? | 11:33 |
|---|---|---|
| === hjensas is now known as hjensas|afk | ||
| Odd_Bloke | meena: For now, they serve as documentation. Per https://cloudinit.readthedocs.io/en/latest/topics/hacking.html#type-annotations, we can't use the typing module yet, so we won't really be able to annotate much of the codebase. | 13:38 |
| rick_h_ | type hints vs type enforcement? | 13:39 |
| Odd_Bloke | It's really all type hinting, because Python doesn't use the type annotations at runtime at all. | 13:47 |
| Odd_Bloke | But yeah, we won't have a CI check that enforces them for now. | 13:47 |
| Odd_Bloke | meena: I would expect @ing a person and commenting on a PR they opened to produce roughly the same notifications in GH. | 14:09 |
| === hjensas|afk is now known as hjensas | ||
| shibumi | hi, is dhclient really necessary? | 16:16 |
| shibumi | why is cloud-init not using systemd-networkd for dhcp? | 16:16 |
| shibumi | I got a bug report for it on arch: https://bugs.archlinux.org/task/66035?dev=16600&type%5B0%5D=&sev%5B0%5D=&due%5B0%5D=&cat%5B0%5D=&status%5B0%5D=open&percent%5B0%5D=&reported%5B0%5D= | 16:16 |
| rharper | shibumi: dhclient is used in early boot only to bring up nic to talk to IMDS; after crawling metadata service, if the distro has a renderer cloud-init writes out network config in the requested format; | 16:18 |
| rharper | dhclient is required, on AWS, for example to enable interaction with the IMDS; | 16:18 |
| blackboxsw | aaaaaand... it's about that time again :) | 16:19 |
| rharper | the system is not configured to use dhclient and we do not keep the dhclient around | 16:19 |
| blackboxsw | #startmeeting cloud-init status meeting | 16:19 |
| meetingology | Meeting started Tue Mar 31 16:19:14 2020 UTC. The chair is blackboxsw. Information about MeetBot at http://wiki.ubuntu.com/meetingology. | 16:19 |
| meetingology | Available commands: action commands idea info link nick | 16:19 |
| blackboxsw | Hello and welcome to another cloud-init community status meeting folks. please feel free to continue any current conversations | 16:19 |
| blackboxsw | I'll interleave status meeting notes with existing conversations. | 16:20 |
| blackboxsw | #chair Odd_Bloke smoser rharper | 16:20 |
| meetingology | Current chairs: Odd_Bloke blackboxsw rharper smoser | 16:20 |
| blackboxsw | our IRC channel topic carries the next planned status meeting for those that wish to participate. All are welcome to interject or drive converstation topics here | 16:20 |
| blackboxsw | let's set that now. to +2 weeks from now | 16:21 |
| === blackboxsw changed the topic of #cloud-init to: pull-requests https://git.io/JeVed | Meeting minutes: https://goo.gl/mrHdaj | Next status meeting April 14 16:15 UTC | 19.4 (Dec 17) drops Py2.7 : origin/stable-19.4 | 20.1 (Feb 18) | https://bugs.launchpad.net/cloud-init/+filebug | ||
| blackboxsw | April 14th, same time | 16:21 |
| shibumi | rharper: thanks! | 16:21 |
| blackboxsw | Previous meeting notes are here | 16:22 |
| blackboxsw | #link https://cloud-init.github.io/status-2020-03-10.html#status-2020-03-10 | 16:22 |
| blackboxsw | The topics we generally cover in this meeting are the following: Previous Actions, Recent Changes, In-progress Development, Community Charter, Upcoming Meetings, Office Hours (~30 mins). | 16:22 |
| blackboxsw | #topic Previous Actions | 16:22 |
| blackboxsw | and from last meeting, no previous actions were unaccounted for. | 16:23 |
| blackboxsw | #topic Recent Changes | 16:23 |
| blackboxsw | recent changes landed in tip of master via git log --since 2020-03-10 | 16:24 |
| blackboxsw | #link https://paste.ubuntu.com/p/55hqVCfnpV/ | 16:24 |
| rharper | shibumi: also, the issue with openstack metadata service not being found is likely related to classless statci route support in the EphemeralDHCP class in cloud-init, we fixed an issue there last fall, https://github.com/canonical/cloud-init/commit/07b17236be5665bb552c7460102bcd07bf8f2be8 | 16:24 |
| rharper | shibumi: long term, we would like to replace dhclient with a python-based one; we've just not had a change to implement a minimal dhclient in python only | 16:24 |
| blackboxsw | in that paste or recent changes, some big items have landed since last we 'met': both NetBSD and OpenBSD distro support is now added to cloud-init. vmware support of guest info gc status, SAP Converged cloud gets identified as OpenStack and Ubuntu Focal prioritizing netplan over ifupdown if both are present | 16:27 |
| blackboxsw | Also Odd_Bloke has been landing and improving cloud-init automated process with near daily branches. github actions/workflows and docs about review and coding style expectations are landing to make it a lot easier for upstream to help speed contributions and reviews | 16:28 |
| blackboxsw | Also, you'll note a lot of dropping python six and other py2-related artifacts from our codebase. Since upstream support is py3.4 or later we can simplify and prune a lot of the vestigial py2 functionality. | 16:29 |
| blackboxsw | Ahh I also forgot, Ec2 now by default (Ubuntu Focal or tip of cloud-init) renders full networking, including secondary IPv4/IPv6 addresses, for all interfaces attached to a VM based on network config supplied by IMDS. Old releases of cloud-init used to only render basic networking on the primary (eth0) nic. | 16:31 |
| blackboxsw | #topic In-progress Development | 16:32 |
| blackboxsw | thanks Goneri and meena for all your the BSD ongoing development and support work there BTW | 16:33 |
| * Goneri waves | 16:33 | |
| blackboxsw | :) upstream is focused a bit on continuing to clean up py2 remnants from tip, adding support for reading netplan configuration from initramfs and continuing to add automation to the github development and release process/tooling to speed reviews. | 16:38 |
| blackboxsw | we are also hammering the review queue a bit better than in the past with daily PR assignments to ensure the system remains more efficient | 16:39 |
| blackboxsw | thanks again for all the code submits folks! | 16:40 |
| blackboxsw | waiting In the wings we will eventually get around to handling the network hotplug solution for cloud-init (if configured) | 16:41 |
| blackboxsw | #topic Community Charter | 16:42 |
| blackboxsw | This section is generally reserved to discuss any general community goals for cloud-init. Per last cloud-init summit we discussed prioritizing the following: | 16:43 |
| blackboxsw | * json schema validation for each cloudinit/config/cc_*py module | 16:43 |
| blackboxsw | * correcting, extending stale datasource documentation under doc/rtd/topics/datasources | 16:43 |
| blackboxsw | those tasks are easy to split up and so we set a goal to try to chunk through it this year | 16:44 |
| blackboxsw | they are also categorized as bugs for easy pickup/assignment for anyone interested. | 16:44 |
| blackboxsw | #link https://bugs.launchpad.net/cloud-init/+bugs?field.tag=bitesize | 16:45 |
| blackboxsw | #topic Office hours (next ~30 mins) | 16:45 |
| blackboxsw | thanks for listening to the above...... In this section a couple upstream devs should be available with eyes on the channel for the next 30 minutes for and bug, feature, PR review questions or concerns. | 16:46 |
| blackboxsw | again thanks for joining. We'll have this meeting again in 2 weeks | 16:46 |
| blackboxsw | I'm going to spend this time working on our cherry-pick script for publishing the Netplan -> ENI work into Ubuntu Focal today. | 16:47 |
| blackboxsw | Thanks for tuning in. | 17:17 |
| blackboxsw | #endmeeting | 17:17 |
| meetingology | Meeting ended Tue Mar 31 17:17:39 2020 UTC. | 17:17 |
| meetingology | Minutes: http://ubottu.com/meetingology/logs/cloud-init/2020/cloud-init.2020-03-31-16.19.moin.txt | 17:17 |
| * meena blames baby for *not* doing BSD work or support lately. | 17:35 | |
| blackboxsw | ohhhh sssssuuuuure meena. blame it on the baby ;) that's why I have kids too ;) | 17:47 |
| blackboxsw | that, and I needed someone in the house that also likes board games | 17:48 |
| meena | currently, she loves screaming, a lot. And honestly, as the youth say: mood. | 17:49 |
| blackboxsw | :) | 17:50 |
| rick_h_ | that's the problem, you think "a kid, yea we're going to do all these fun things together!" | 17:56 |
| rick_h_ | and then you realize none of that starts until you put in a LOT of foundation time (like 3+ years) | 17:56 |
| blackboxsw | true story | 17:57 |
| sarnold | "have you considered a puppy instead?" | 17:58 |
| meena | sarnold: we already have 2 cats (they're not fans) | 18:27 |
| sarnold | haha :) | 18:29 |
| Gaffel | Do I have to resort to raw commands to create a system user and group where a home dir is created and its ownership and security label is set properly? (UID and GID below 1000, but create home) | 20:13 |
| Odd_Bloke | Gaffel: It does look like not creating a homedir for a system user is hardcoded, unfortunately: https://github.com/canonical/cloud-init/blob/master/cloudinit/distros/__init__.py#L471-L478 | 20:22 |
| Gaffel | I was worried about that. That's not how useradd works though, which makes me disappointed. :[ | 20:23 |
| Gaffel | I want to create a user+group with a UID & GID between 500 and 999 that also creates a home directory with the right ownership and SELinux labels. But I guess I'll have to do it differently. | 20:24 |
| Odd_Bloke | Yeah, I'm not sure what the reasoning there is, but it looks like that behaviour has been in place since 2012. | 20:26 |
| Odd_Bloke | It certainly seems like overriding that default behaviour would be desirable. | 20:26 |
| Odd_Bloke | This looks like https://bugs.launchpad.net/cloud-init/+bug/1864728 to em. | 20:26 |
| ubot5 | Ubuntu bug 1864728 in cloud-init "Unable to create interactive "system" user" [Undecided,Incomplete] | 20:26 |
| Odd_Bloke | *me | 20:26 |
| Odd_Bloke | Gaffel: Do you have any interest in addressing the issue in cloud-init? ^_^ | 20:27 |
| Gaffel | Changing the behaviour will be tricky, won't it? | 20:27 |
| Gaffel | I guess no_create_home could be checked and the current behaviour would remain unless you set that to false. | 20:28 |
| Gaffel | I assume that leaving it blank results in none though. | 20:29 |
| Odd_Bloke | Yeah, I think you're along the right path with using no_create_home. | 20:30 |
| Odd_Bloke | The defaulting would be a little tricky, but not super hard once you've reasoned out the matrix of possible combinations. | 20:32 |
| Gaffel | I have the same usecase and OS family as that person. I'm trying to introduce more security hardening at work. I'll look into this. :] | 20:39 |
| Gaffel | I'm new to cloud-init so I'll do my best. :) | 20:43 |
| Odd_Bloke | Gaffel: \o/ You'll want to read through https://cloudinit.readthedocs.io/en/latest/topics/hacking.html to get started. | 20:46 |
| Odd_Bloke | Gaffel: Though actually the CLA info there is out-of-date, you can just fill in your GH username and that's sufficient. | 20:46 |
| Odd_Bloke | (The wording here predates that field being in the CLA form. :p) | 20:46 |
| Gaffel | You mean the field where it wants my Launchpad ID? | 20:53 |
| Gaffel | Or just leave that blank? | 20:53 |
| Gaffel | I can't even read the text in the verification image. :/ | 20:56 |
| rharper | Odd_Bloke: I guess we should update the docs and mention the field in the form | 20:58 |
| Odd_Bloke | rharper: Yep, addressing that now. | 20:58 |
| rharper | awesome! | 20:58 |
| Gaffel | And do I really need to enter my physical address? :| | 20:59 |
| Odd_Bloke | AFAIK, you do. | 21:00 |
| Gaffel | :/ | 21:01 |
| Odd_Bloke | rharper: https://github.com/canonical/cloud-init/pull/297 | 21:06 |
| rharper | reviewing | 21:06 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!