Adaptive self-maintenance scheduler

1. A method, comprising:

estimating, based on a current state of a distributed storage system comprising a primary storage server and a plurality of secondary storage servers, an amount of computing resources of the primary storage server and the plurality of secondary storage servers required to perform each of a plurality of self-maintenance processes wherein the plurality of secondary storage servers are configured to perform one or more backup processes and the plurality of self-maintenance processes, wherein the plurality of self-maintenance processes include at least one of garbage collection and correcting inconsistencies associated with the distributed file system, wherein the plurality of secondary storage servers form a distributed file system providing backup storage services to the primary storage server;

estimating, based on one or more inputs, an amount of computing resources of the primary storage server and the plurality of secondary storage servers available to perform one or more of the plurality of self-maintenance processes during at least a first time period of a plurality of time periods;

determining which of the plurality of self-maintenance processes to perform during the at least first time period based on the estimated amount of computing resources of the primary storage server and the plurality of secondary storage servers required to perform each of the plurality of self-maintenance processes and the estimated amount of computing resources of the primary storage server and the plurality of secondary storage servers available to perform one or more of the plurality of self-maintenance processes wherein one or more of the self-maintenance processes are determined to be performed during the at least first time period; and

scheduling, based on the estimated amount of computing resources required and on the estimated amount of computing resources available, the determined one or more self-maintenance processes to perform during the first time period.

2. The method of claim 1 , wherein the one or more inputs includes at least one of a plurality of current activities of the primary storage server and the plurality of secondary storage servers, the current state of the primary storage server and the state of the plurality of secondary storage servers, a plurality of external events, and the estimated amount of computing resources of the primary storage server and the plurality of secondary storage servers required to perform each of the self-maintenance processes.

3. The method of claim 1 , further comprising:

performing each of the scheduled self-maintenance processes during the first time period; and

collecting execution statistics of the performance of each of the scheduled self-maintenance processes.

4. The method of claim 3 , wherein the amount of computing resources of the primary storage server and the plurality of secondary storage servers required and the amount of computing resources of the primary storage server and the plurality of secondary storage servers available are further estimated based on the execution statistics.

5. The method of claim 4 , further comprising:

upon determining that the estimated amount of computing resources available is within a specified range of an actual amount of computing resources available based on the execution statistics, reinforcing the one or more inputs using a machine learning algorithm; and

upon determining that the estimated amount of computing resources available is not within the specified range of the actual amount of computing resources available based on the execution statistics, readjusting the one or more inputs using the machine learning algorithm.

6. The method of claim 1 , wherein the amount of computing resources required and amount of computing resources available include at least one of an amount of I/O resources, network resources, storage capacity, and processing resources.

7. A non-transitory computer-readable storage medium having instructions, which, when executed on a processor, perform an operation comprising:

estimating, based on a current state of a distributed storage system comprising a primary storage server and a plurality of secondary storage servers, an amount of computing resources of the primary storage server and the plurality of secondary storage servers required to perform each of a plurality of self-maintenance processes wherein the plurality of secondary storage servers are configured to perform one or more backup processes and the plurality of self-maintenance processes, wherein the plurality of self-maintenance processes include at least one of garbage collection and correcting inconsistencies associated with the distributed file system, wherein the plurality of secondary storage servers form a distributed file system providing backup storage services to the primary storage server;

estimating, based on one or more inputs, an amount of computing resources of the primary storage server and the plurality of secondary storage servers available to perform one or more of the plurality of self-maintenance processes during at least a first time period of a plurality of time periods;

determining which of the plurality of self-maintenance processes to perform during the at least first time period based on the estimated amount of computing resources of the primary storage server and the plurality of secondary storage servers required to perform each of the plurality of self-maintenance processes and the estimated amount of computing resources of the primary storage server and the plurality of secondary storage servers available to perform one or more of the plurality of self-maintenance processes, wherein one or more of the self-maintenance processes are determined to be performed during the at least first time period; and

scheduling, based on the estimated amount of computing resources required and on the estimated amount of computing resources available, the determined one or more self-maintenance processes to perform during the first time period.

8. The computer-readable storage medium of claim 7 , wherein the one or more inputs includes at least one of a plurality of current activities of the primary storage server and the plurality of secondary storage servers, the current state of the primary storage server and the state of the plurality of secondary storage servers, a plurality of external events, and the estimated amount of computing resources of the primary storage server and the plurality of secondary storage servers required to perform each of the self-maintenance processes.

9. The computer-readable storage medium of claim 7 , wherein the operation further comprises:

performing each of the scheduled self-maintenance processes during the first time period; and

collecting execution statistics of the performance of each of the scheduled self-maintenance processes.

10. The computer-readable storage medium of claim 9 , wherein the amount of computing resources of the primary storage server and the plurality of secondary storage servers required and the amount of computing resources of the primary storage server and the plurality of secondary storage servers available are further estimated based on the execution statistics.

11. The computer-readable storage medium of claim 10 , the operation further comprising:

upon determining that the estimated amount of computing resources available is within a specified range of an actual amount of computing resources available based on the execution statistics, reinforcing the one or more inputs using a machine learning algorithm; and

upon determining that the estimated amount of computing resources available is not within the specified range of the actual amount of computing resources available based on the execution statistics, readjusting the one or more inputs using the machine learning algorithm.

12. The computer-readable storage medium of claim 7 , wherein

the amount of computing resources required and amount of computing resources available include at least one of an amount of I/O resources, network resources, storage capacity, and processing resources.

13. A system, comprising:

a processor; and

a memory containing program code, which, when executed on the processor, performs an operation comprising:

estimating, based on a current state of a distributed storage system comprising a primary storage server and a plurality of secondary storage servers, an amount of computing resources of the primary storage server and the plurality of secondary storage servers required to perform each of a plurality of self-maintenance processes wherein the plurality of secondary storage servers are configured to perform one or more backup processes and the plurality of self-maintenance processes, wherein the plurality of self-maintenance processes include at least one of garbage collection and correcting inconsistencies associated with the distributed file system, wherein the plurality of secondary storage servers form a distributed file system providing backup storage services to the primary storage server;

estimating, based on one or more inputs, an amount of computing resources of the primary storage server and the plurality of secondary storage servers available to perform one or more of the plurality of self-maintenance processes during at least a first time period of a plurality of time periods;

determine which of the plurality of self-maintenance processes to perform during the at least first time period based on the estimated amount of computing resources of the primary storage server and the plurality of secondary storage servers required to perform each of the plurality of self-maintenance processes and the estimated amount of computing resources of the primary storage server and the plurality of secondary storage servers available to perform one or more of the plurality of self-maintenance processes, wherein one or more of the self-maintenance processes are determined to be performed during the at least first time period; and

scheduling, based on the estimated amount of computing resources required and on the estimated amount of computing resources available, the determined one or more self-maintenance processes to perform during the first time period.

14. The system of claim 13 , wherein the one or more inputs includes at least one of a plurality of current activities of the primary storage server and the plurality of secondary storage servers, the current state of the primary storage server and the state of the plurality of secondary storage servers, a plurality of external events, and the estimated amount of computing resources of the primary storage server and the plurality of secondary storage servers required to perform each of the self-maintenance processes.

15. The system of claim 13 , wherein the operation further comprises:

performing each of the scheduled self-maintenance processes during the first time period; and

collecting execution statistics of the performance of each of the scheduled self-maintenance processes.

16. The system of claim 15 , wherein the amount of computing resources of the primary storage server and the plurality of secondary storage servers required and the amount of computing resources of the primary storage server and the plurality of secondary storage servers available are further estimated based on the execution statistics.

17. The system of claim 16 , the operation further comprising:

upon determining that the estimated amount of computing resources available is within a specified range of an actual amount of computing resources available based on the execution statistics, reinforcing the one or more inputs using a machine learning algorithm; and

upon determining that the estimated amount of computing resources available is not within the specified range of the actual amount of computing resources available based on the execution statistics, readjusting the one or more inputs using the machine learning algorithm.

18. The system of claim 13 , wherein the amount of computing resources required and amount of computing resources available include at least one of an amount of I/O resources, network resources, storage capacity, and processing resources.