Description of problem: 1.destination vm quits automatically while ping-pong migration hit ENOSPAC 2.have the two hosts connect to the same iscsi server For example - iscsiadm --mode discovery --type sendtargets --portal 10.66.11.239 --discover iscsiadm --mode node --targetname AMD_migration --portal 10.66.11.239 --login Version-Release number of selected component (if applicable): kernel&kvm 2.6.18-260.el5 kvm-83-232.el5 How reproducible: always Steps to Reproduce: 1.please make sure block device has a very small initial size for example- ACTIVE '/dev/tt/deng1' [152.00 MB] inherit 2.qemu-img create -f qcow2 /dev/tt/deng1 20G 3.boot source(A) and destination(B) guest together before doing migration A's command, /usr/libexec/qemu-kvm -rtc-td-hack -usbdevice tablet -no-hpet -drive file=/dev/tt/deng1,if=virtio,boot=on,werror=stop,cache=none,format=qcow2,media=disk -smp 2 -m 2G -net nic,macaddr=00:12:24:58:a1:2f,model=virtio,vlan=0 -net tap,script=/etc/qemu-ifup,vlan=0 -uuid `uuidgen` -boot dc -balloon virtio -notify all -monitor stdio -vnc :3 -soundhw ac97 -drive file=/opt/en_windows_server_2008_r2_standard_enterprise_datacenter_and_web_with_sp1_x64_dvd_617601.iso,if=ide,cache=none,media=cdrom,format=raw -drive file=/opt/virtio-drivers-1.0.1-52454.vfd,if=floppy,cache=none,format=raw B's command /usr/libexec/qemu-kvm -rtc-td-hack -usbdevice tablet -no-hpet -drive file=/dev/tt/deng1,if=virtio,boot=on,werror=stop,cache=none,format=qcow2,media=disk -smp 2 -m 2G -net nic,macaddr=00:12:24:58:a1:2f,model=virtio,vlan=0 -net tap,script=/etc/qemu-ifup,vlan=0 -uuid `uuidgen` -boot dc -balloon virtio -notify all -monitor stdio -vnc :3 -soundhw ac97 -drive file=/opt/en_windows_server_2008_r2_standard_enterprise_datacenter_and_web_with_sp1_x64_dvd_617601.iso,if=ide,cache=none,media=cdrom,format=raw -drive file=/opt/virtio-drivers-1.0.1-52454.vfd,if=floppy,cache=none,format=raw -incoming tcp:0:1111 4.do ping-pong live migration and make sure hitting ENOSPAC every time,if you cannot reproduce the bz please try more than once.In order to make sure vm status is running before each migration,please also extend block device size if necessary. A side - (qemu) migrate -d tcp:10.66.9.103:1111# VM is stopped due to disk write error: virtio0: No space left on device (qemu) migration_notify_done B side - check the vm status,if the VM is paused,please increase the lvm size and then continue vm for the next migration Actual results: After src side finished the migration,dst side vm quitted automatically with the following message - (qemu) # migration: migration process finished raw_pread: read beyond end of file Expected results:The whole migration process is smoothly and vm works correctly. Additional info: Host AMD model name : AMD Phenom(tm) 9600B Quad-Core Processor stepping : 3 cpu MHz : 1200.000