Managing processor overcommit for virtual machines
Systems and methods for managing processor overcommit for virtual machines are disclosed. In one implementation, a processing device may identify, by a hypervisor running on a host computer system, an over-committed physical processor, the physical processor is executing a thread implementing a first virtual processor (vCPU) of a virtual machine (VM). The processing device may further queue, by the hypervisor, a message notifying the VM of hot-unplugging of the first vCPU. Responsive to receiving, by the hypervisor, a notification of a second vCPU of the VM transitioning into an idle state, the processing device may also transmitting the message to the VM.
1. A method comprising:
identifying, by a hypervisor running on a host computer system, an over-committed physical processor, wherein the physical processor is executing a thread implementing a first virtual processor (vCPU) of a virtual machine (VM);
queuing, by the hypervisor, a message notifying the VM of hot-unplugging of the first vCPU; and
responsive to receiving, by the hypervisor, a notification of a second vCPU of the VM transitioning into an idle state, transmitting the message to the VM.
2. The method of claim 1 , wherein identifying the over-committed physical processor comprises:
determining that a task is waiting to be executed by the physical processor while the physical processor is executing the thread implementing the first vCPU.
3. The method of claim 2 , wherein the task is a second thread implementing a second vCPU of a second VM.
4. The method of claim 1 , wherein the first vCPU and the second vCPU are the same vCPU.
5. The method of claim 1 , wherein the message to the VM further comprises an inter-processor interrupt (IPI) to unplug the first vCPU.
6. The method of claim 1 , wherein hot-unplugging of the first vCPU comprises removing the first vCPU from a set of vCPUs associated with the VM without shutting down the VM.
7. The method of claim 1 , wherein transmitting the message to the VM is performed via a shared memory region that is accessible to the vCPUs of the VM.
8. A non-transitory computer-readable storage medium comprising executable instructions that, when executed by a processing device, cause the processing device to:
detect that a first virtual processor (vCPU) of a virtual machine (VM) running on a host computer system is transitioning into an idle state;
responsive to the detection, determine whether a configuration change request of hot-unplugging a second vCPU of the VM is pending;
responsive to determining that the configuration change request of hot-unplugging the second vCPU of the VM is pending, transmit a message notifying the VM of hot-unplugging of the second vCPU; and
mark the configuration change request as complete.
9. The non-transitory computer-readable storage medium of claim 8 , wherein the configuration change request corresponds to a determination that at least one of a plurality of vCPUs of the VM is being executed by an over-committed physical processor of the host computer system.
10. The non-transitory computer-readable storage medium of claim 9 , wherein to identify the over-committed physical processor, the processing logic is further to detect that a first task is waiting to be executed by the physical processor while the physical processor is executing a second task.
11. The non-transitory computer-readable storage medium of claim 8 , wherein the message to the VM further comprises an IPI to hot-unplug the second vCPU.
12. The non-transitory computer-readable storage medium of claim 8 , wherein hot-unplugging of the second vCPU further comprises removing the second vCPU from a set of vCPUs associated with the VM without shutting down the VM.
13. The non-transitory computer-readable storage medium of claim 8 , wherein to detect that the first vCPU is transitioning into the idle state, the processing logic is further to detect that the first vCPU is executing a VM exit instruction.
14. The non-transitory computer-readable storage medium of claim 8 , wherein the processing device is further to:
responsive to waiting for a predetermined period of time, send the message to the VM using a shared memory region that is accessible to the vCPUs of the VM.
15. A system comprising:
a memory; and
a processing device operatively coupled to the memory, wherein the processing device is to:
identify an over-committed physical processor, wherein the physical processor is executing a thread implementing a first virtual processor (vCPU) of a virtual machine (VM);
queuing a message notifying the VM of hot-unplugging of the first vCPU; and
responsive to receiving a notification of a second vCPU of the VM transitioning into an idle state, transmitting the message to the VM.
16. The system of claim 15 , wherein identifying the over-committed physical processor comprises:
determining that a task is waiting to be executed by the physical processor while the physical processor is executing the thread implementing the first vCPU.
17. The system of claim 16 , wherein the task is a second thread implementing a second vCPU of a second VM.
18. The system of claim 15 , wherein the first vCPU and the second vCPU are the same vCPU.
19. The system of claim 15 , wherein the message to the VM further comprises an inter-processor interrupt (WI) to unplug the first vCPU.
20. The system of claim 15 , wherein hot-unplugging of the first vCPU comprises removing the first vCPU from a set of vCPUs associated with the VM without shutting down the VM.