Decoupling compute and storage resources in cloud-based HCI (hyper-converged infrastructure)

1. A method for decoupling compute and storage resources in a hyper-converged infrastructure (HCI) deployment, the method comprising:

receiving, by a control plane of the HCI deployment, a request to add a host to a host cluster of the HCI deployment;

in response to the request, determining, by the control plane, a storage capacity to provision for the host;

provisioning, by the control plane, the host from a host platform of an infrastructure on which the HCI deployment is implemented;

provisioning, by the control plane, one or more storage volumes from a storage platform of the infrastructure, the storage platform running on physical server resources in the infrastructure that are separate from the host platform, the one or more storage volumes having the determined storage capacity; and

causing, by the control plane, the one or more storage volumes to be network-attached to the host, wherein upon being network-attached to the host, firmware on the host causes the one or more storage volumes to be exposed as directly-attached storage devices to a hypervisor of the host, and

wherein the hypervisor of the host is configured to make the one or more storage volumes available, as part of a virtual storage pool, to one or more virtual machines in the HCI deployment for data storage.

2. The method of claim 1 wherein the host is a diskless host that does not use any storage resources local to the host for storing data.

3. The method of claim 1 wherein the virtual storage pool also comprises storage volumes provisioned from the storage platform that are network-attached to other hosts in the HCI deployment.

4. The method of claim 1 further comprising, at a time the host is taken offline for maintenance or experiences a failure:

detaching the one or more storage volumes from the host;

provisioning a second host from the host platform of the infrastructure; and

attaching the one or more storage volumes to the second host.

5. The method of claim 1 further comprising, when a timeout is detected with respect to a storage volume in the one or more storage volumes:

waiting for a period of greater than one hour before concluding that the storage volume has failed.

6. The method of claim 1 wherein the hypervisor is configured to avoid performing data deduplication with respect to data written to the one or more storage volumes.

7. The method of claim 1 wherein the virtual storage pool does not include any local storage resources of the host or other hosts in the HCI deployment.

8. A non-transitory computer readable storage medium having stored thereon program code executable by a control plane of a hyper-converged infrastructure (HCI) deployment, the program code embodying a method for decoupling compute and storage resources in the HCI deployment, the method comprising:

receiving a request to add a host to a host cluster of the HCI deployment;

in response to the request, determining a storage capacity to provision for the host;

provisioning the host from a host platform of an infrastructure on which the HCI deployment is implemented;

provisioning one or more storage volumes from a storage platform of the infrastructure, the storage platform running on physical server resources in the infrastructure that are separate from the host platform, the one or more storage volumes having the determined storage capacity; and

causing the one or more storage volumes to be network-attached to the host, wherein upon being network-attached to the host, firmware on the host causes the one or more storage volumes to be exposed as directly-attached storage devices to a hypervisor of the host, and

wherein the hypervisor of the host is configured to make the one or more storage volumes available, as part of a virtual storage pool, to one or more virtual machines in the HCI deployment for data storage.

9. The non-transitory computer readable storage medium of claim 8 wherein the host is a diskless host that does not use any storage resources local to the host for storing data.

10. The non-transitory computer readable storage medium of claim 8 wherein the virtual storage pool also comprises storage volumes provisioned from the storage platform that are network-attached to other hosts in the HCI deployment.

11. The non-transitory computer readable storage medium of claim 8 wherein the method further comprises, at a time the host is taken offline for maintenance or experiences a failure:

detaching the one or more storage volumes from the host;

provisioning a second host from the host platform of the infrastructure; and

attaching the one or more storage volumes to the second host.

12. The non-transitory computer readable storage medium of claim 8 wherein the method further comprises, when a timeout is detected with respect to a storage volume in the one or more storage volumes:

waiting for a period of greater than one hour before concluding that the storage volume has failed.

13. The non-transitory computer readable storage medium of claim 8 wherein the hypervisor is configured to avoid performing data deduplication with respect to data written to the one or more storage volumes.

14. The non-transitory computer readable storage medium of claim 8 wherein the virtual storage pool does not include any local storage resources of the host or other hosts in the HCI deployment.

15. A computer system implemented as part of a control plane of a hyper-converged infrastructure (HCI) deployment, the HCI deployment, the computer system comprising:

a processor; and

a non-transitory computer readable medium having stored thereon program code that, when executed, causes the processor to:

receive a request to add a host to a host cluster of the HCI deployment;

in response to the request, determine a storage capacity to provision for the host;

provision the host from a host platform of an infrastructure on which the HCI deployment is implemented;

provision one or more storage volumes from a storage platform of the infrastructure, the storage platform running on physical server resources in the infrastructure that are separate from the host platform, the one or more storage volumes having the determined storage capacity; and

cause the one or more storage volumes to be network-attached to the host, wherein upon being network-attached to the host, firmware on the host causes the one or more storage volumes to be exposed as directly-attached storage devices to a hypervisor of the host, and

wherein the hypervisor of the host is configured to make the one or more storage volumes available, as part of a virtual storage pool, to one or more virtual machines in the HCI deployment for data storage.

16. The computer system of claim 15 wherein the host is a diskless host that does not use any storage resources local to the host for storing data.

17. The computer system of claim 15 wherein the virtual storage pool also comprises storage volumes provisioned from the storage platform that are network-attached to other hosts in the HCI deployment.

18. The computer system of claim 15 wherein the program code further comprises code that causes the processor to, at a time the host is taken offline for maintenance or experiences a failure:

detach the one or more storage volumes from the host;

provision a second host from the host platform of the infrastructure; and

attach the one or more storage volumes to the second host.

19. The computer system of claim 15 wherein the program code further comprises code that causes the processor to, when a timeout is detected with respect to a storage volume in the one or more storage volumes:

wait for a period of greater than one hour before concluding that the storage volume has failed.

20. The computer system of claim 15 wherein the hypervisor is configured to avoid performing data deduplication with respect to data written to the one or more storage volumes.

21. The computer system of claim 15 wherein the virtual storage pool does not include any local storage resources of the host or other hosts in the HCI deployment.