Virtual machine recovery in shared memory architecture

A computer system for virtual machine recovery using heartbeat monitoring in a shared pooled memory architecture wherein the shared pooled memory is accessible to a plurality of hosts, the computer system comprising:

• a memory device providing the shared pooled memory;
• a processor;
• and
• a non-transitory computer readable medium having stored thereon program code, the program code causing the processor to:
• monitor, by an agent running on a first host of the plurality of hosts, heartbeats of one or more hosts of the plurality of hosts, the plurality of hosts providing a plurality of virtual machines access to the shared pooled memory, the monitoring enabling the first host to determine a current status of the one or more hosts;
• upon failing to receive a heartbeat from a second host of the one or more hosts with in a threshold time period since a last heartbeat was received from the second host, identify the second host as an isolated host, the second host running a set of virtual machines of the plurality of virtual machines;
• upon identifying the second host as an isolated host, change a third host from a maintenance mode to an active mode and resume the set of virtual machines on the third host of the one or more hosts using memory page table metadata stored on the shared pooled memory;
• and
• place a fourth host, within the plurality of hosts, in maintenance mode to serve as a new backup host.

A non-transitory computer readable storage medium having stored thereon program code executable by a processor, the program when executed by a processor perform operations for virtual machine recovery using heartbeat monitoring in a shared pooled memory architecture wherein the shared pooled memory is accessible to a plurality of hosts, the operations comprising:

• monitoring, by an agent running on a first host of the plurality of hosts, heartbeats of one or more hosts of the plurality of hosts, the plurality of hosts providing a plurality of virtual machines access to the shared pooled memory, the monitoring enabling the first host to determine a current status of the one or more hosts;
• upon failing to receive a heartbeat from a second host of the one or more hosts with in a threshold time period since a last heartbeat was received from the second host, identifying the second host as an isolated host, the second host running a set of virtual machines of the plurality of virtual machines;
• upon identifying the second host as an isolated host, changing a third host from a maintenance mode to an active mode and resuming the set of virtual machines on the third host of the one or more hosts using memory page table metadata stored on the shared pooled memory;
• and
• placing a fourth host, within the plurality of hosts, in maintenance mode to serve as a new backup host.