The following error shows up in xend.log while doing a local live
migration of a 2.6.32 based PV guest running on a Xen 3.4.0 based
system:[2010-12-28 09:00:51 4826] DEBUG (DevController:144) Waiting for 51728.This message shows up in the domU's dmesg: vbd vbd-51728: 16 Device in use; refusing to closevm.cfg has: disk = ['file:/OVS/seed_pool/OVM_EL5U5_X86_64_PVM_10GB/System.img,xvda,w','phy:/dev/sda1,xvdb,w']The messages do not occur after changing "w" to "w+" (may have been "w!"). Is "w+" required for local migrations? Is it masking an underlying bug? waitForBackend() calls hotplugStatusCallback(): xen/source/tools/python/xen/xend/server/DevController.py waitForBackend() calls hotplugStatusCallback() 556 xswatch(statusPath, hotplugStatusCallback, ev, result) xen/source/tools/python/xen/xend/server/DevController.py hotplugStatusCallback()This treats HOTPLUG_STATUS_BUSY as a one-time error. It seems like it should return 1 (keep watching) on HOTPLUG_STATUS_BUSY, something like: if status is not None: We would time out if it it stays busy for DEVICE_CREATE_TIMEOUT. Not sure if we would need to do an unwatch in that case. Thanks, Chuck |
_______________________________________________ Xen-devel mailing list Xen-devel@xxxxxxxxxxxxxxxxxxx http://lists.xensource.com/xen-devel