| === waxfire7 is now known as waxfire | ||
| === waxfire4 is now known as waxfire | ||
| === waxfire3 is now known as waxfire | ||
| === waxfire9 is now known as waxfire | ||
| === waxfire7 is now known as waxfire | ||
| === waxfire0 is now known as waxfire | ||
| beantaxi | Kind of a non-cloud-init question ... I'm beginning to get semi-comfortable testing cloud-init userdata in lxd containers. But there are steps that are a bit slow, like when I want to test full-lifecycle and create a completely fresh image, publish it, then launch an instance from my published image | 17:28 |
|---|---|---|
| beantaxi | I'm guessing there some tricks I that can make this faster, anything from how I set up my storage-pool(s), to what sort of cloud instance I run on for testing. I'm currently using a baby OVHcloud instance, but if it speeds things up to Live Large and pay for a faster instance for an hour or two I'm willing to get in that lane | 17:30 |
| beantaxi | Thanks! | 17:30 |
| beantaxi | In particular, it's pretty fast to launch a new image from a 'Ubuntu 20 plus 95% of what I need" base image, and then execute userdata to create a whole new instance. But then it takes minutes to publish this new image, and then several more minutes to launch a new instance from the new image, even though it's all local. | 18:00 |
| beantaxi | I could make things easier on myself during testing, by breaking after instance creation and checking out my new image at that point. But I'd rather just make stuff fast if there's a way to make it fast. Thanks! | 18:01 |
| === waxfire7 is now known as waxfire | ||
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!