[sheepdog-users] Whether there are other ways to make other nodes taking over the virtual machine when one node is lost.

Liu Yuan namei.unix at gmail.com
Tue May 28 10:35:16 CEST 2013


On 05/28/2013 04:30 PM, xiao li wrote:
> But I  I start the Virtual Machine on a remote host such as 172.6.6.6.
> When the host node(10.2.7.12) crashes, the memory state should exist on
> the 172.6.6.6. Why can not my host (172.6.6.6)  take over the VM through
> the data on the other nodes, such as the data on the node(10.2.7.13).

This needs multi-connection and connection restart support from QEMU
sheepdog block dirver. Those two features are planned, but not yet started.

Thanks,
Yuan


More information about the sheepdog-users mailing list