cmdlib: only use index=N for supermin root#1398
cmdlib: only use index=N for supermin root#1398openshift-merge-robot merged 2 commits intocoreos:masterfrom
Conversation
|
@Prashanth684 Could you sanity-check that a |
Sure we can with an initramfs which is what we have now. It's just the custom supermin initramfs which is limited. |
jlebon
force-pushed
the
pr/more-osmet-prep
branch
from
ef247f1 to
3da5c18
Compare
|
@jlebon: PR needs rebase. DetailsInstructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository. |
For coreos#1244, I want to use `serial=osmet` on the disk to mark which one is the source disk containing the metal image instead of relying on `index=N`. However, to use `serial=osmet`, I have to use a separate `-device` arg instead of the implied one from `if=virtio`, and it seems like doing that causes qemu to give those disks higher priority such that e.g. it gets `/dev/vda`, which throws off supermin. Let's switch over all our disks to use `-drive ... -device ...`. While we're here, only rely on `index=N` for the supermin root itself, since we can't use udev symlinks in the kernel cmdline. The cache disk too doesn't need an index, we use the LABEL instead already.
Instead of using SCSI, use virtio-blk's `serial=` option so allow us to identify the disk inside `create_disk.sh`. And since we already have udev running there now, we don't have to walk sysfs anymore looking for our block device.
jlebon
force-pushed
the
pr/more-osmet-prep
branch
from
3da5c18 to
e24956a
Compare
|
Rebased! Was hoping for a sanity-check for multi-arch, though maybe let's just get this in? I'd like to get #1244 in for coreos/coreos-installer#187. |
|
/lgtm |
|
[APPROVALNOTIFIER] This PR is APPROVED This pull-request has been approved by: cgwalters, darkmuggle, jlebon The full list of commands accepted by this bot can be found here. The pull request process is described here DetailsNeeds approval from an approver in each of these files:
Approvers can indicate their approval by writing |
|
/refresh |
|
/refresh |
5 participants
For #1244, I want to use
serial=osmeton the disk to mark which one isthe source disk containing the metal image instead of relying on
index=N. However, to useserial=osmet, I have to use a separate-devicearg instead of the implied one fromif=virtio, and it seemslike doing that causes qemu to give those disks higher priority such
that e.g. it gets
/dev/vda, which throws off supermin.Let's switch over all our disks to use
-drive ... -device .... Whilewe're here, only rely on
index=Nfor the supermin root itself, sincewe can't use udev symlinks in the kernel cmdline. The cache disk too
doesn't need an index, we use the LABEL instead already.