| 28 May 2024 |
 adamcstephens | hmm, i'm not sure where that comes from | 14:39:18 |
| adamcstephens | In reply to @hexa:lossy.network
a bit annoying that it is stuck with a single node what does this mean? | 14:39:26 |
 hexa | so you can set up your local CLI to use remotes | 14:41:38 |
| hexa | * so, you can set up your local CLI to control remote clusters over the HTTP API | 14:42:51 |
| hexa | incus config trust add foo prints a token | 14:43:06 |
| hexa | and you can configure that with incus remote add somecluster <token> | 14:43:19 |
| hexa | and then incus remote switch somecluster | 14:43:26 |
| hexa | but this isn't somecluster, it is somehostofsomecluster only | 14:43:40 |
| adamcstephens | ahh, so the remote isn't cluster aware | 14:43:56 |
| hexa | yeah | 14:44:02 |
| hexa | the keying probably is | 14:44:04 |
| hexa | but the CLI isn't smart enough to failover to other cluster members | 14:44:17 |
| adamcstephens | yeah that's a shame. so you'd have to add all the nodes individually and target them | 14:46:16 |
| hexa | and switch projects as needed 🙂 | 14:46:26 |
| adamcstephens | what kind of environment are you trying to replace? | 14:48:36 |
| adamcstephens | vmware? | 14:48:39 |
| hexa | yes, esxi with vcenter server | 15:02:42 |
| hexa | but nobody at the workplace loved vmware to begin with | 15:02:52 |
| hexa | it is very much consultware with expensive licensing programs | 15:03:13 |
| hexa | I also still have ganeti and proxmox (both with ceph) running elsewhere 😄 | 15:04:20 |
| hexa | and the way things look like right now … we should probably migrate everything to incus | 15:04:36 |
| hexa | * and the way things look right now … we should probably migrate everything to incus | 15:04:42 |
 steveej | how scalable are incus clusters? can it compete with k3s? | 19:46:14 |
| steveej | * how scalable are incus clusters? can it compete with k3s or nomad? | 19:46:19 |
| hexa | not sure, would be best to ask in #lxc on libera | 20:01:28 |
| adamcstephens | i'm not sure it is really competition for those, which are primarily OCI container orchestrators | 20:09:39 |
| steveej | nomad has a ton of exec drivers beyond "containers", including raw_exec which i used to just run nix on the host. i don't like their new license though (BUSL) | 20:11:32 |
| adamcstephens | i'm aware. still not the same space as incus, which is designed for running whole-OS containers and VMs | 20:14:23 |
| 29 May 2024 |
| steveej | In reply to @adam:robins.wtf
i'm aware. still not the same space as incus, which is designed for running whole-OS containers and VMs got it. whole-OS (except kernel/initrd i assume) containers are well suited towards reusing the existing nixpkgs modules so i'm starting to understand your preference to incus :-) | 10:32:27 |
| 30 May 2024 |
| hexa | is there really no intent to backport non-lts release for incus? 🤔 | 23:52:35 |
