Revert "cmdlib: Use -hda/-hdb instead of virtio-scsi incantations" #1292
Conversation
This reverts commit ad728ab because it apparently doesn't work on s390x/ppc64le: coreos#1288 (comment)
|
I'll try to lower these into Ideally we use |
|
/lgtm |
|
[APPROVALNOTIFIER] This PR is APPROVED This pull-request has been approved by: cgwalters, 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 |
| cachedisk=() | ||
| if [ -f "${workdir}/cache/cache.qcow2" ]; then | ||
| cachedisk=("-hdb" "${workdir}/cache/cache.qcow2") | ||
| cachedisk=("-drive" "if=none,id=drive-scsi0-0-0-1,discard=unmap,file=${workdir}/cache/cache.qcow2" \ |
There was a problem hiding this comment.
Random thought...do we still need the cache disk versus relying on a 9p mount? I may investigate this.
Would simplify things a lot if we had just one disk attached to the VM versus two.
There was a problem hiding this comment.
That was the original design, but we moved away from that because composing directly into 9p caused issues. And composing on the side and moving cache/data back and forth was a lot of overhead.
|
FWIW, the issue on ppc64le was due to the driver ibmvscsi missing from the supermin initrd. I filed a bug against supermin at https://bugzilla.redhat.com/show_bug.cgi?id=1819019 I have not tested on s390x, so I don't know if they have the same issue. |
5 participants
This reverts commit ad728ab because
it apparently doesn't work on s390x/ppc64le:
#1288 (comment)