| 29 Feb 2024 |
 raitobezarius | and fill them in the environment of the disko script at runtime? | 23:58:30 |
| raitobezarius | (mostly, I have divergence due to test environment under virtio do /dev/vdx, SCSI/SATA do /dev/sdx, NVMe do /dev/nvme* | 23:59:32 |
| raitobezarius | * (mostly, I have divergence due to test environment under virtio do /dev/vdx, SCSI/SATA do /dev/sdx, NVMe do /dev/nvme*) | 23:59:33 |
| raitobezarius | while I could emulate NVMe devices to have the exact same thing in QEMU, I wonder about supporting those usecases | 23:59:49 |
| 1 Mar 2024 |
 lassulus | in theory it should work with the gpt type. since we use labels there. in practice this was never tested and somethng will probably break but should be easily fixable | 00:31:12 |
| raitobezarius | funnily, I am debugging a case where I use the .raw generation | 00:31:52 |
| raitobezarius | and mount it using QEMU and the disk doesn't show up at boot | 00:31:58 |
| raitobezarius | which seems to hint at missing drivers or something | 00:32:03 |
| raitobezarius | In reply to @lassulus:lassul.us
in theory it should work with the gpt type. since we use labels there. in practice this was never tested and somethng will probably break but should be easily fixable but how does that work with label | 00:34:57 |
| raitobezarius | I have disk without label | 00:35:00 |
| raitobezarius | * I have disks without label | 00:35:02 |
| raitobezarius | so I need to use sdx/vdx/nvmexyz | 00:35:09 |
| raitobezarius | or I need to use things like wwn | 00:35:14 |
| raitobezarius | but it's hard to get WWN beforehand in Netbox or something | 00:35:22 |
| raitobezarius | this would require a new inventory discipline that I don't have right now | 00:35:29 |
| lassulus | no gpt type will take /dev/sda or /dev/nvmexxx and will format it and then use /dev/disk/by-partlabel for all operations afterwards | 00:37:46 |
| lassulus | so the device should be used at just one location | 00:38:04 |
| raitobezarius | yeah | 00:40:34 |
| raitobezarius | so the thing i am talking about is the initial formatting | 00:40:40 |
| lassulus | ok, but then you should be able to use an environment variable as the initial device | 00:41:40 |
| lassulus | and fill it dynamically when you execute the diskoScript | 00:41:50 |
| raitobezarius | In reply to @lassulus:lassul.us
ok, but then you should be able to use an environment variable as the initial device yeah, just trying to confirm if it makes sense to you | 00:41:50 |
| raitobezarius | like having this weird device = "$PRIMARY_DISK"; ? | 00:42:04 |
| lassulus | yeah, that should work | 00:42:14 |
| lassulus | but not sure that it actually does :) | 00:42:20 |
| raitobezarius | ack | 00:42:23 |
| raitobezarius | will try | 00:42:27 |
| lassulus | but if it doesn't that is a bug worth fixing | 00:42:28 |
| raitobezarius | if i find things | 00:42:28 |
| raitobezarius | i will send you PRs | 00:42:30 |
