Note: This bug is displayed in read-only format because the product is no longer active in Red Hat Bugzilla.

Bug 1685777

Summary: Block node is read-only when start a qcow2 image on https server and with auto-read-only=on set on its format node
Product: Red Hat Enterprise Linux Advanced Virtualization Reporter: aihua liang <aliang>
Component: qemu-kvmAssignee: Virtualization Maintenance <virt-maint>
qemu-kvm sub component: Storage QA Contact: aihua liang <aliang>
Status: CLOSED WONTFIX Docs Contact:
Severity: low    
Priority: low CC: chayang, coli, juzhang, kwolf, ngu, qzhang, rbalakri, virt-maint
Version: 8.1Keywords: Triaged
Target Milestone: rcFlags: pm-rhel: mirror+
Target Release: ---   
Hardware: Unspecified   
OS: Unspecified   
Whiteboard:
Fixed In Version: Doc Type: If docs needed, set a value
Doc Text:
Story Points: ---
Clone Of: 1685772 Environment:
Last Closed: 2021-03-15 07:33:55 UTC Type: Bug
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:
Bug Depends On: 1685772    
Bug Blocks:    

Comment 1 aihua liang 2019-03-06 02:53:11 UTC
In qemu-kvm-3.1.0-18.module+el8+2834+fa8bb6e2.x86_64, also hit this issue.

Comment 2 Hanna Czenczek 2019-05-16 12:03:36 UTC
Kevin, what do you think of this?  (The fact that a qcow2 node with auto-read-only on top of a read-only node isn’t switched to read-only automatically.)

I suppose we can make auto-read-only switch nodes to being read-only if we cannot give them the write permission they wish to take.  But I’m a bit afraid of the consequences.  Won’t this be too automagical?  (Like, shouldn’t we then also undo that whenever possible; do we have to switch a node to read-only because another node unshares the write permission; etc.)  Or is that maybe what you had in mind all along?

Max

Comment 3 Kevin Wolf 2019-05-16 12:55:43 UTC
I think the proper way to do this would be what we do in file-posix: Make it dependent on the users of the node. bdrv_open() would open the image read-only and only attaching a writer changes it to read-write; detaching changes it back.

Depending on the use case, just falling back to read-only on startup and leaving it there might be enough and would be much easier. Though this BZ looks like we don't even have an actual use case attached to it? Maybe failure isn't a problem in practice?

Comment 4 Hanna Czenczek 2019-11-19 13:27:21 UTC
A solution to this problem (as proposed by Kevin in comment 3) wouldn’t be trivial, so I won’t work on resolving this issue until there is an impact on an actual use case.

(Note that in the meantime you simply have to specify read-only=on for the qcow2 node instead of auto-read-only, just as you always did before auto-read-only was introduced.)

Max

Comment 5 Ademar Reis 2020-02-05 22:54:59 UTC
QEMU has been recently split into sub-components and as a one-time operation to avoid breakage of tools, we are setting the QEMU sub-component of this BZ to "General". Please review and change the sub-component if necessary the next time you review this BZ. Thanks

Comment 8 RHEL Program Management 2021-03-15 07:33:55 UTC
After evaluating this issue, there are no plans to address it further or fix it in an upcoming release.  Therefore, it is being closed.  If plans change such that this issue will be fixed in an upcoming release, then the bug can be reopened.