@edison : "That comes from libvirt, this error means: there is already a disk with "vdb" in libvirt's internal meta data, then the agent Attaches the device with the same id "vdb" to the VM, then failed."
Do you have any idea for the solution? does this mean that data of the volume somehow still there, although the volume have been detached from the vm? Actually, i try to create a volume and then attach it to vm immeadiately, without waiting for the volume to be ready (it actually works). After that i detach the volume through the management server, and then try to attach it again. Is this a bug? Or am i doing something wrong? ________________________________________ From: Edison Su [edison...@citrix.com] Sent: Thursday, August 02, 2012 1:37 AM To: cloudstack-dev@incubator.apache.org Subject: RE: Fail attaching volume to a virtual machine > -----Original Message----- > From: Wido den Hollander [mailto:w...@widodh.nl] > Sent: Wednesday, August 01, 2012 4:38 AM > To: cloudstack-dev@incubator.apache.org > Subject: Re: Fail attaching volume to a virtual machine > > On 08/01/2012 05:54 AM, Tommy Aditya Lawanto wrote: > > @edison : > > > > Okay i have follow the instructions from your link. Does this mean i > have to submit bug to bugs.cloudstack.org also? (i upload agent.log on > pastebin.com for now) > > > > - I use KVM Host Cent OS 6.2 > > - my libvirt version is 0.9.4 > > - Here is my agent.log : http://pastebin.com/Lm5GZymR > > Reading this log I'm not sure if this is a libvirt problem or a Agent > problem. > > The XML: > > <disk device='disk' type='file'> > <driver name='qemu' type='qcow2' cache='none' /> > <source > file='/mnt/4a6d1d00-facc-360a-bf04-2eb0be9dbde0/41a96b7a-da21-4a85- > 97f1-ebd58668e531'/> > <target dev='vdb' bus='virtio'/> > </disk> > > And suddenly Qemu says: > > Failed to attach device to i-14-90-VM: internal error unable to execute > QEMU command '__com.redhat_drive_add': Duplicate ID 'drive-virtio- > disk1' > for drive > > Where did the 'drive-virtio-disk1' come from? Didn't libvirt come up > with this name? Since I don't see this in the XML generated by the > CloudStack agent. That comes from libvirt, this error means: there is already a disk with "vdb" in libvirt's internal meta data, then the agent Attaches the device with the same id "vdb" to the VM, then failed. > > > > ________________________________________ > >> From: Edison Su [edison...@citrix.com] > >> Sent: Wednesday, August 01, 2012 3:45 AM > > To: cloudstack-dev@incubator.apache.org > > Subject: RE: Fail attaching volume to a virtual machine > > > > > > Could you follow > http://wiki.cloudstack.org/display/COMM/KVM+agent+debug, then upload > agent.log to somewhere? > > > >> -----Original Message----- > >> From: Tommy Aditya Lawanto [mailto:to...@wowrack.com] > >> Sent: Tuesday, July 31, 2012 1:26 AM > >> To: cloudstack-dev@incubator.apache.org > >> Subject: RE: Fail attaching volume to a virtual machine > >> > >> @edison : > >> > >> i use CentOS 6.2. > >> > >> I tried stop the vm, start it, then attach the volume => no good, it > >> fail > >> > >> Also try reboot, and then attach the volume => also fail > >> > >> Then i try to just stop the vm, attach the volume, and then start > the > >> vm => surprisingly it works > >> > >> After it works, i try to detach the volume => successful. > >> > >> And then i attach it again => fail like usual. > >> > >> Does this mean the virtual machine must be in stopped state in order > to > >> attach volume into it? > >> ________________________________________ > >> From: Edison Su [edison...@citrix.com] > >> Sent: Monday, July 30, 2012 10:59 PM > >> To: cloudstack-dev@incubator.apache.org > >> Subject: RE: Fail attaching volume to a virtual machine > >> > >> What's the kvm host os? Centos/RHEL 6.2? can you stop/start the vm, > >> then attach/detach the volume, will it work? > >> > >>> -----Original Message----- > >>> From: Tommy Aditya Lawanto [mailto:to...@wowrack.com] > >>> Sent: Sunday, July 29, 2012 11:55 PM > >>> To: cloudstack-dev@incubator.apache.org > >>> Subject: RE: Fail attaching volume to a virtual machine > >>> > >>> @david nalley: Okay, here is the xml i get by running "virsh > dumpxml > >>> <instance>" : > >>> > >>> http://pastebin.com/cTSrHwpV > >>> > >>> Thank you for your help :) > >>> ________________________________________ > >>> From: Tommy Aditya Lawanto [to...@wowrack.com] > >>> Sent: Monday, July 30, 2012 9:48 AM > >>> To: cloudstack-dev@incubator.apache.org > >>> Subject: RE: Fail attaching volume to a virtual machine > >>> > >>> @david nalley : I'm running cloudstack version 3, on basic > networking > >>> mode. > >>> > >>> Sorry, but how do i "virsh dumpxml <instance>" on the HyperVisor? > (do > >>> you mean that i should type that command when i ssh to my > management > >>> server? > >>> ________________________________________ > >>> From: Tommy Aditya Lawanto [to...@wowrack.com] > >>> Sent: Friday, July 27, 2012 1:46 PM > >>> To: cloudstack-us...@incubator.apache.org > >>> Subject: RE: Fail attaching volume to a virtual machine > >>> > >>> @david nalley : I'm running cloudstack version 3, on basic > networking > >>> mode. > >>> > >>> Sorry, but how do i "virsh dumpxml <instance>" on the HyperVisor? > (do > >>> you mean that i should type that command when i ssh to my > management > >>> server? > >>> ________________________________________ > >>> From: Tommy Aditya Lawanto [to...@wowrack.com] > >>> Sent: Friday, July 27, 2012 1:32 PM > >>> To: cloudstack-us...@incubator.apache.org > >>> Subject: Fail attaching volume to a virtual machine > >>> > >>> Im getting this error after I removed a Volume from an CentOS 5.8 > >>> Instance and tried to attach another volume to it: > >>> > >>> Failed to attach volume: try volume to VM: f0f33816-0b2d-477d-b673- > >>> 7789ecd9b738; org.libvirt.LibvirtException: internal error unable > to > >>> execute QEMU command '__com.redhat_drive_add': Duplicate ID 'drive- > >>> virtio-disk1' for drive > >>> > >>> How can I fix it? > >>> > >>> Cheers!