Cache allocation for controller boards based on prior input-output operations
A metadata structure of a storage array stores metadata associated with a plurality of prior input-output operations. The metadata comprises an indication of which of the controller boards was utilized to service a prior input-output operation and an input-output size for the prior input-output operation. A given input-output operation is obtained and a target controller board and a target portion of the storage array are identified based at least in part on the given input-output operation. A given controller board is determined to have a higher likelihood of receiving a future input-output operation than at least one other controller board based at least in part on the metadata and a portion of the cache is allocated to the given controller board for storing target data associated with the given input-output operation. The target controller board utilizes the portion of the cache allocated to the given controller board to service the input-output operation.
1. An apparatus comprising:
a storage array comprising a plurality of controller boards and a cache, the cache comprising a metadata structure, the metadata structure storing metadata associated with a plurality of prior input-output operations, each of the prior input-output operations being directed to a corresponding portion of the storage array, the metadata associated with each prior input-output operation comprising an indication of which of the plurality of controller boards was utilized to service that prior input-output operation and an input-output size for that prior input-output operation, at least one processing device of the storage array being configured:
to obtain a given input-output operation from a host device;
to identify a target controller board of the plurality of controller boards and a target portion of the storage array based at least in part on the obtained given input-output operation;
to determine, based at least in part on the metadata stored in the metadata structure, a likelihood that a given controller board of the plurality of controller boards will receive a future input-output operation;
to determine, based at least in part on the likelihood, that the given controller board has a higher likelihood of receiving a future input-output operation than at least one other controller board of the plurality of controller boards;
to allocate a portion of the cache to the given controller board as a buffer for storing target data associated with the given input-output operation; and
to service the given input-output operation by the target controller board, the target controller board utilizing the buffer to service the input-output operation.
2. The apparatus of claim 1 wherein the at least one processing device is further configured to determine, based at least in part on the metadata stored in the metadata structure, a most likely input-output size for the future input-output operation and wherein the at least one processing device is configured to allocate the portion of the cache as the buffer with a size corresponding to the determined most likely input-output size.
3. The apparatus of claim 1 wherein determining, based at least in part on the metadata structure, the likelihood that the given controller board of the plurality of controller boards will receive the future input-output operation comprises performing a principal component analysis on the metadata.
4. The apparatus of claim 1 wherein the metadata structure comprises a plurality of tracks, each track comprising metadata corresponding to one or more of the plurality of prior input-output operations, each track corresponding to a particular portion of the storage array.
5. The apparatus of claim 1 wherein the cache comprises a plurality of banks, each of the banks being designated as a bank for buffers of a corresponding input-output size, the at least one processing device being further configured to dynamically adjust the size of one or more of the banks based at least in part on the metadata stored in the metadata structure.
6. The apparatus of claim 1 wherein the at least one processing device is further configured:
to obtain one or more utilization rates from the host device;
to determine, based at least in part on the obtained one or more utilization rates and the metadata stored in the metadata structure, whether the target controller board should perform a remote read of the target data from the buffer of the given controller board or a remote push should be performed to copy the target data from the buffer of the given controller board to a buffer of the target controller board for servicing the given input-output operation; and
to cause performance of one of the remote read operation and the remote push operation based at least in part on the determination of whether the target controller board should perform the remote read or the remote push should be performed.
7. The apparatus of claim 1 wherein the at least one processing device is further configured to perform asynchronous pre-caching of data on the plurality of controller boards based at least in part on the metadata stored in the metadata structure.
8. A method comprising:
obtaining a given input-output operation from a host device;
identifying a target controller board of a plurality of controller boards of a storage array and a target portion of the storage array based at least in part on the obtained given input-output operation;
determining, based at least in part on metadata stored in a metadata structure of the storage array, a likelihood that a given controller board of the plurality of controller boards will receive a future input-output operation, the metadata being associated with a plurality of prior input-output operations, each of the prior input-output operations being directed to a corresponding portion of the storage array, the metadata associated with each prior input-output operation comprising an indication of which of the plurality of controller boards was utilized to service that prior input-output operation and an input-output size for that prior input-output operation;
determining, based at least in part on the likelihood, that the given controller board has a higher likelihood of receiving a future input-output operation than at least one other controller board of the plurality of controller boards;
allocating a portion of the cache to the given controller board as a buffer for storing target data associated with the given input-output operation; and
servicing the given input-output operation by the target controller board, the target controller board utilizing the buffer to service the input-output operation;
wherein the method is performed by at least one processing device of the storage array.
9. The method of claim 8 wherein the method further comprises determining, based at least in part on the metadata stored in the metadata structure, a most likely input-output size for the future input-output operation and wherein the at least one processing device is configured to allocate the portion of the cache as the buffer with a size corresponding to the determined most likely input-output size.
10. The method of claim 8 wherein determining, based at least in part on the metadata structure, the likelihood that the given controller board of the plurality of controller boards will receive the future input-output operation comprises performing a principal component analysis on the metadata.
11. The method of claim 8 wherein the metadata structure comprises a plurality of tracks, each track comprising metadata corresponding to one or more of the plurality of prior input-output operations, each track corresponding to a particular portion of the storage array.
12. The method of claim 8 wherein the cache comprises a plurality of banks, each of the banks being designated as a bank for buffers of a corresponding input-output size, the method further comprising dynamically adjusting the size of one or more of the banks based at least in part on the metadata stored in the metadata structure.
13. The method of claim 8 wherein the method further comprises:
obtaining one or more utilization rates from the host device;
determining, based at least in part on the obtained one or more utilization rates and the metadata stored in the metadata structure, whether the target controller board should perform a remote read of the target data from the buffer of the given controller board or a remote push should be performed to copy the target data from the buffer of the given controller board to a buffer of the target controller board for servicing the given input-output operation; and
causing performance of one of the remote read operation and the remote push operation based at least in part on the determination of whether the target controller board should perform the remote read or the remote push should be performed.
14. The method of claim 8 wherein the method further comprises performing asynchronous pre-caching of data on the plurality of controller boards based at least in part on the metadata stored in the metadata structure.
15. A computer program product comprising a non-transitory processor-readable storage medium having stored therein program code of one or more software programs, wherein the program code, when executed by at least one processing device of a storage array, the storage array comprising a plurality of controller boards and a cache, the cache comprising a metadata structure, the metadata structure storing metadata associated with a plurality of prior input-output operations, each of the prior input-output operations being directed to a corresponding portion of the storage array, the metadata associated with each prior input-output operation comprising an indication of which of the plurality of controller boards was utilized to service that prior input-output operation and an input-output size for that prior input-output operation, causes the at least one processing device:
to obtain a given input-output operation from a host device;
to identify a target controller board of the plurality of controller boards and a target portion of the storage array based at least in part on the obtained given input-output operation;
to determine, based at least in part on the metadata stored in the metadata structure, a likelihood that a given controller board of the plurality of controller boards will receive a future input-output operation;
to determine, based at least in part on the likelihood, that the given controller board has a higher likelihood of receiving a future input-output operation than at least one other controller board of the plurality of controller boards;
to allocate a portion of the cache to the given controller board as the buffer for storing target data associated with the given input-output operation; and
to service the given input-output operation by the target controller board, the target controller board utilizing the buffer allocated to the given controller board to service the input-output operation.
16. The computer program product of claim 15 wherein the program code further causes the at least one processing device to determine, based at least in part on the metadata stored in the metadata structure, a most likely input-output size for the future input-output operation and wherein the at least one processing device is configured to allocate the portion of the cache as the buffer with a size corresponding to the determined most likely input-output size.
17. The computer program product of claim 15 wherein determining, based at least in part on the metadata structure, the likelihood that the given controller board of the plurality of controller boards will receive the future input-output operation comprises performing a principal component analysis on the metadata.
18. The computer program product of claim 15 wherein the metadata structure comprises a plurality of tracks, each track comprising metadata corresponding to one or more of the plurality of prior input-output operations, each track corresponding to a particular portion of the storage array.
19. The computer program product of claim 15 wherein the cache comprises a plurality of banks, each of the banks being designated as a bank for buffers of a corresponding input-output size, the program code further causing the at least one processing device to dynamically adjust the size of one or more of the banks based at least in part on the metadata stored in the metadata structure.
20. The computer program product of claim 15 wherein the program code further causes the at least one processing device:
to obtain one or more utilization rates from the host device;
to determine, based at least in part on the obtained one or more utilization rates and the metadata stored in the metadata structure, whether the target controller board should perform a remote read of the target data from the buffer of the given controller board or a remote push should be performed to copy the target data from the buffer of the given controller board to a buffer of the target controller board for servicing the given input-output operation; and
to cause the performance of one of the remote read operation and the remote push operation based at least in part on the determination of whether the target controller board should perform the remote read or the remote push should be performed.