Power management for distributed storage systems

1. A method for a computing system in a virtualized computing environment to perform power management for a distributed storage system accessible by a cluster in the virtualized computing environment, the method comprising:

determining, by the computing system, that a power-off requirement is satisfied for a first host from the cluster, wherein the power-off requirement is determined to be satisfied based on multiple second hosts from the cluster complying with a data placement policy configured for the cluster after the first host is powered off, and wherein the data placement policy specifies one or more of the following: a number of failure to tolerate, a stripe width corresponding to a number of disk stripes per virtual machine object, and a fault domain associated with each of the multiple second hosts;

based on the data placement policy, selecting, by the computing system, one or more of the multiple second hosts and placing virtual machine data stored on the first host onto the one or more of the multiple second hosts;

instructing, by the computing system, to migrate one or more virtual machines supported by the first host to the one or more of the multiple second hosts; and

instructing, by the computing system, to power off the first host.

2. The method of claim 1 , wherein the determining that the power-off requirement is satisfied for the first host comprises:

prior to determining that the power-off requirement is satisfied for the first host, determining that data rebalancing is required for in the cluster.

3. The method of claim 2 , wherein the method further comprises:

after powering off the first host, performing data rebalancing to distribute data components substantially evenly among storage devices associated with the multiple second hosts.

4. The method of claim 1 , wherein the determining that the power-off requirement is satisfied for the first host comprises:

determining that a resource utilization level associated with the first host is lower than a predetermined threshold.

5. The method of claim 1 , wherein the determining that the power-off requirement is satisfied for the first host comprises:

determining that one of more of the multiple second hosts have sufficient storage capacity for the virtual machine data.

6. The method of claim 1 , wherein the selecting the one or more of the multiple second hosts is based on an amount of storage capacity available on each of the multiple second hosts.

7. A non-transitory computer-readable storage medium comprising instructions, which when executed in a computing system in a virtualized computing environment, cause the computing system to perform a method to perform power management for a distributed storage system accessible by a cluster in the virtualized computing environment, the method comprising:

determining, by the computing system, that a power-off requirement is satisfied for a first host from the cluster, wherein the power-off requirement is determined to be satisfied based on multiple second hosts from the cluster complying with a data placement policy configured for the cluster after the first host is powered off, and wherein the data placement policy specifies one or more of the following: a number of failure to tolerate, a stripe width corresponding to a number of disk stripes per virtual machine object, and a fault domain associated with each of the multiple second hosts;

based on the data placement policy, selecting, by the computing system, one or more of the multiple second hosts and placing virtual machine data stored on the first host onto the one or more of the multiple second hosts;

instructing, by the computing system, to migrate one or more virtual machines supported by the first host to the one or more of the multiple second hosts; and

instructing, by the computing system, to power off the first host.

8. The non-transitory computer-readable storage medium of claim 7 , wherein the determining that the power-off requirement is satisfied for the first host comprises:

prior to determining that the power-off requirement is satisfied for the first host, determining that data rebalancing is required for in the cluster.

9. The non-transitory computer-readable storage medium of claim 8 , wherein the method further comprises:

after powering off the first host, performing data rebalancing to distribute data components substantially evenly among storage devices associated with the multiple second hosts.

10. The non-transitory computer-readable storage medium of claim 7 , wherein the determining that the power-off requirement is satisfied for the first host comprises:

determining that a resource utilization level associated with the first host is lower than a predetermined threshold.

11. The non-transitory computer-readable storage medium of claim 7 , wherein the determining that the power-off requirement is satisfied for the first host comprises:

determining that one of more of the multiple second hosts have sufficient storage capacity for the virtual machine data.

12. The non-transitory computer-readable storage medium of claim 7 , wherein the selecting the one or more of the multiple second hosts is based on an amount of storage capacity available on each of the multiple second hosts.

13. A computing system in a virtualized computing environment configured to perform power management for a distributed storage system accessible by a cluster in the virtualized computing environment, comprising:

a processor; and

a non-transitory computer-readable medium having stored thereon program code that, upon being executed by the processor, cause the processor to:

determine that a power-off requirement is satisfied for a first host from the cluster, wherein the power-off requirement is determined to be satisfied based on multiple second hosts from the cluster complying with a data placement policy configured for the cluster after the first host is powered off, and wherein the data placement policy specifies one or more of the following: a number of failure to tolerate, a stripe width corresponding to a number of disk stripes per virtual machine object, and a fault domain associated with each of the multiple second hosts;

based on the data placement policy, select one or more of the multiple second hosts and place virtual machine data stored on the first host onto the one or more of the multiple second hosts;

instruct to migrate one or more virtual machines supported by the first host to the one or more of the multiple second hosts; and

instruct to power off the first host.

14. The computing system of claim 13 , wherein program code for determining that the power-off requirement is satisfied for the first host cause the processor to:

prior to determining that the power-off requirement is satisfied for the first host, determine that data rebalancing is required for in the cluster.

15. The computing system of claim 14 , wherein the program code further cause the processor to:

after powering off the first host, perform data rebalancing to distribute data components substantially evenly among storage devices associated with the multiple second hosts.

16. The computing system of claim 13 , wherein program code for determining that the power-off requirement is satisfied for the first host cause the processor to:

determine that a resource utilization level associated with the first host is lower than a predetermined threshold.

17. The computing system of claim 13 , wherein program code for determining that the power-off requirement is satisfied for the first host cause the processor to:

determine that one of more of the multiple second hosts have sufficient storage capacity for the virtual machine data.

18. The computing system of claim 13 , wherein program code for selecting the one or more of the multiple second hosts is based on an amount of storage capacity available on each of the multiple second hosts.