Bug 969767
| Summary: | engine: unexpected exception error when trying to hotplug a disk when its domain is in maintenance | ||||||
|---|---|---|---|---|---|---|---|
| Product: | Red Hat Enterprise Virtualization Manager | Reporter: | Dafna Ron <dron> | ||||
| Component: | ovirt-engine | Assignee: | Maor <mlipchuk> | ||||
| Status: | CLOSED CURRENTRELEASE | QA Contact: | Aharon Canan <acanan> | ||||
| Severity: | medium | Docs Contact: | |||||
| Priority: | unspecified | ||||||
| Version: | 3.2.0 | CC: | abaron, acanan, acathrow, amureini, iheim, jkt, lpeer, Rhev-m-bugs, scohen, yeylon | ||||
| Target Milestone: | --- | Flags: | scohen:
needinfo+
scohen: needinfo+ |
||||
| Target Release: | 3.3.0 | ||||||
| Hardware: | x86_64 | ||||||
| OS: | Linux | ||||||
| Whiteboard: | storage | ||||||
| Fixed In Version: | is9 | Doc Type: | Bug Fix | ||||
| Doc Text: | Story Points: | --- | |||||
| Clone Of: | Environment: | ||||||
| Last Closed: | Type: | Bug | |||||
| Regression: | --- | Mount Type: | --- | ||||
| Documentation: | --- | CRM: | |||||
| Verified Versions: | Category: | --- | |||||
| oVirt Team: | Storage | RHEL 7.3 requirements from Atomic Host: | |||||
| Cloudforms Team: | --- | Target Upstream Version: | |||||
| Embargoed: | |||||||
| Attachments: |
|
||||||
Both operations plugged disk and attach disk should be fixed here. on attach disk, the dialog box, has also a checkbox in the bottom of it, which indicate if the user want to also plug the disk immediately after it was attached. if the storage domain is inactive the plug operation will fail. My question is, should we also fail the attach operation as well (since we can know that the plug will fail), or we should let the user at least attach the disk to the VM but keep it unplugged and add a warning audit log that the storage domain is inactive? (In reply to Maor from comment #2) We should let the user at least attach the disk to the VM but keep it unplugged and add a warning audit log that the storage domain is inactive. Sean, Maor, this is a disgusting user experience. We can have this validation BEFORE the user submits the request - instead of giving only half the functionality, we should FAIL it with an error message that plugging would fail, but the user can choose to attach without plugging. Why is this not preferable? As long as the user can choose to attach without plugging, I am fine with failing it with an error message. (In reply to Sean Cohen from comment #5) > As long as the user can choose to attach without plugging, I am fine with > failing it with an error message. Of course there is such a functionality. Thanks. Verified, tested on RHEVM 3.3 - IS15 environment:
Host OS: RHEL 6.5
RHEVM: rhevm-3.3.0-0.22.master.el6ev.noarch
PythonSDK: rhevm-sdk-python-3.3.0.14-1.el6ev.noarch
VDSM: vdsm-4.12.0-138.gitab256be.el6ev.x86_64
LIBVIRT: libvirt-0.10.2-24.el6.x86_64
QEMU & KVM: qemu-kvm-rhev-0.12.1.2-2.402.el6.x86_64
SANLOCK: sanlock-2.8-1.el6.x86_64
------------------------------------------------------------------------
Error while executing action:
vm_0001:
Cannot attach Virtual Machine Disk. The relevant Storage Domain is inaccessible.
-Please handle Storage Domain issues and retry the operation.
------------------------------------------------------------------------
Closing - RHEV 3.3 Released Closing - RHEV 3.3 Released Closing - RHEV 3.3 Released |
Created attachment 755790 [details] logs Description of problem: I have a floating disk on a vm which is in maintenance. I ran a vm and tried to hotPlug the disk. the engine sends the command to vdsm which returns with Bad volume specification error and engine shows the user the following error: Error while executing action Attach Disk to VM: Unexpected exception engine log error: 2013-06-02 12:22:39,868 ERROR [org.ovirt.engine.core.bll.AttachDiskToVmCommand] (ajp-/127.0.0.1:8702-9) [3db9b428] Command org.ovirt.engine.core.bll.AttachDiskToVmCommand throw Vdc Bll exception. With error message VdcBLLException: org.ovirt.engine.core.vdsbroker.vdsbroker.VDSErrorException: VDSGenericException: VDSErrorException: Failed to HotPlugDiskVDS, error = Unexpected exception vdsm error: Thread-94007::ERROR::2013-06-02 12:22:43,953::dispatcher::66::Storage.Dispatcher.Protect::(run) {'status': {'message': "Volume does not exist: ('8e893a47-ec05-438d-a774-93f6e8a9da9a',)", 'code': 201}} Thread-94007::ERROR::2013-06-02 12:22:43,953::BindingXMLRPC::932::vds::(wrapper) unexpected error Traceback (most recent call last): File "/usr/share/vdsm/BindingXMLRPC.py", line 918, in wrapper res = f(*args, **kwargs) File "/usr/share/vdsm/BindingXMLRPC.py", line 251, in vmHotplugDisk return vm.hotplugDisk(params) File "/usr/share/vdsm/API.py", line 424, in hotplugDisk return curVm.hotplugDisk(params) File "/usr/share/vdsm/libvirtvm.py", line 1781, in hotplugDisk diskParams['path'] = self.cif.prepareVolumePath(diskParams) File "/usr/share/vdsm/clientIF.py", line 275, in prepareVolumePath raise vm.VolumeError(drive) VolumeError: Bad volume specification {'iface': 'virtio', 'format': 'cow', 'optional': 'false', 'volumeID': '8e893a47-ec05-438d-a774-93f6e8a9da9a', 'imageID': '201cf246-bf3e-49cb-b8df-18b81c27229b', 'readonly': 'false', 'domainID': '741 Version-Release number of selected component (if applicable): sf17.2 How reproducible: 100% Steps to Reproduce: 1. in iscsi storage with two hosts and two domains, create a vm with one disk and run it 2. create a floating disk 3. when the vm is up, put the domain in which the floating disk is on in maintenance. 4. when the domain is in maintenance try to attach the floating disk to the vm Actual results: user gets the following error: Error while executing action Attach Disk to VM: Unexpected exception Expected results: engine should check status of domain before sending the hotPlug command to vdsm and if we decide not to do that we should at least create a better error for the user. Additional info: logs