[12:01] <hsi43> Hi
[12:01] <hsi43> I've ran into a small glitch with deploying Ubuntu 20 on AWS. when secondary private IP address is in use.
[12:01] <hsi43> Perhaps someone from here has ran into similar or can comment on the behaviour change between Ubuntu 18 LTS and 20 LTS.
[12:01] <hsi43> Ubuntu 18 does not spin up the secondary private IP by default. It does not appear in /etc/netplan/50-cloud-init.yaml.
[12:01] <hsi43> However in Ubuntu 20, 50-cloud-init.yaml contains the secondary address and result is that this IP is brought up on the instance as primary. And primary address is attached also, but appears as secondary.
[12:01] <hsi43> Can be worked around by disabling cloud-init's network config but a more desired outcome would be to have the instance spin up with correct config without additional steps.
[12:01] <hsi43> Thanks for help :)
[13:43] <chiluk> so falcojr and holmanb thanks for the help the day before yesterday.  I figured out what was going on.  I had assumed that the cloud-init was actually processing the network section and actually making changes.  I was unaware that it was just copy and pasting it into /etc/netplan/50-* ... Once I figured that out it was simply a matter of copy and pasting the section myself and then running netplan try.  Netplan was erroring out on 
[13:43] <chiluk> having a set-name  section without a match key.
[15:26] <amessyroom> is there a way with layout to specify the size of partitions in GB or MB rather than percentage of disk ? i.e. layout: [[66, 83], [33, 82]]
[18:24] <blackboxsw> amessyroom: not to date. currently just percentages as you mentioning. That might be an interesting feature to add. cc_disk_setup would have to to a bunch more validation on total size allocated, but we do have a cloudinit.util human2bytes function that could parse and handle converting typical definitions on human-readable strings and units. 
[19:17] <ice10001> Good morning all.
[19:18] <ice10001> blackboxsw I made some great advancement yesterday, with my code. I have run into a snag now with adding and updating repositories
[19:19] <ice10001> Where do I past my code again?
[19:19] <ice10001> *paste
[19:28] <holmanb> ice10001: paste.ubuntu.com, GitHub gist, pastebin, etc are reasonable ways to share
[19:29] <ice10001> Thnx holmanb
[19:29] <holmanb> Anytime :)
[19:34] <holmanb> Newbie: can't say I've used gzip encoded userdata _or_ alicloud, have you tried isolating the issue (getting a success with either 1. unencoded data on alicloud or 2. gzipped data on a different datasource or in a countainer/vm?) 
[19:35] <ice10001> Here is my code, I want to add repositories for postgresql & docker. Does anyone know if this yaml would work?
[19:35] <ice10001> https://pastebin.com/a58bWXrp
[19:35] <holmanb> Newbie: also, were you able to check the logs for warnings/errors?
[19:35] <ice10001> Everything up to apt:    works
[19:35] <ice10001> I just need to add the remaining code for repositories and install
[19:36] <ice10001> ** Add repositories, update them and install the latest
[19:37] <holmanb> ice10001: I don't think the dpkg --printarchitecture will work there
[19:37] <ice10001> Should I be specific? AMD64 ?
[19:38] <ice10001> We will only ever always run amd64
[19:39] <ice10001> x86_64
[19:39] <holmanb> same with $(lsb_release -cs)
[19:41] <ice10001> How would I script that to use the current version?
[19:42] <holmanb> there are a few different replacement variables that the apt module automatically does replacements for: https://cloudinit.readthedocs.io/en/latest/topics/modules.html#apt-configure
[19:42] <holmanb> $RELEASE, $MIRROR, $KEY_FILE, etc
[19:43] <ice10001> :)
[19:46] <holmanb> ice10001: so $(lsb_release -cs) => $RELEASE, and I suspect you could probably drop the arch= altogether or just hardcode that one, since you said it will only run on amd64
[19:47] <ice10001> Ok, let me try that :)
[19:47] <ice10001> Any idea on updating those repos's before installing with the same yaml?
[19:48] <holmanb> not sure what you're asking
[19:50] <ice10001> If I add those repositories, during my ubuntu server install (this is a user-data file for my custom iso) how do I install the latest postgresql from my added repo. Do I have the put the command 'package_upgrade: true' anywhere before calling for the package installs at the end?
[19:52] <holmanb> oh I see, I assume you'd want package_update: true so apt learns of the new repo
[19:53] <ice10001> Yep
[19:54] <ice10001> I have added it here:
[19:54] <ice10001> https://pastebin.com/V7MCA3Yn
[19:55] <ice10001> vscode has not complained...
[19:57] <ice10001> My test Vbox machine with the custom iso did not succeed, there was an error
[19:57] <ice10001> On the current yaml
[20:40] <holmanb> ice10001: did you check the yaml using cloud-init devel schema? what error?
[20:42] <blackboxsw> also "packages:" key is incorrectly nested under package_update. due to that leading whitespace before "   packages:" that's invalid yaml
[20:46] <blackboxsw> "yaml.scanner.ScannerError: mapping values are not allowed here.. line 2, column 11"  which we can also see from cloudinit devel schema --annotate -c <yourfile> https://paste.ubuntu.com/p/mXzPsGMZTh/