Optimized read IO for mix read/write scenario by chunking write IOs
In one embodiment, a storage system receives a number of input/output (IO) request transactions at the storage system having multiple storage devices. For each of the plurality of IO request transactions, the system determines a number of child IO requests required to complete the IO request transaction. The system tags the IO request transaction and/or the associated child IO requests with a tag identifier. For each of the child requests that is a write IO request, the system determines an optimal write IO request size, segments the write IO request into a number of sub-IO write requests, each having an optimal request size, and interleaves sub-IO write requests with read IO requests for servicing to avoid impact in performance to read IO requests for a mixed IO workload.
1. A computer-implemented method to prioritize input/output request transactions for a storage system, the method comprising:
receiving a plurality of input/output (IO) request transactions at the storage system having a plurality of storage devices;
for each of the plurality of IO request transactions, determining a plurality of child IO requests required to complete the IO request transaction;
tagging the IO request transaction and the associated child IO requests with a unique tag identifier;
for each of the child IO requests that is a write IO request, determining an optimal write IO request size by analyzing IO sequences on an IO bus having mixed read and write IO requests at a fixed point in time to determine an optimal write IO request size for that fixed point in time so that the optimal write IO request size can be applied to a subsequent write IO request;
segmenting the write IO request into a plurality of sub-IO write requests, each having an the optimal write IO request size; and
interleaving sub-IO write requests with read IO requests for servicing to avoid impact in performance to the read IO requests for a mixed IO workload.
2. The computer-implemented method of claim 1 , wherein determining an optimal write IO request size comprises dynamically computing an optimal sub-IO write request size during run time to achieve optimal read IO request latency for a mixed IO workload.
3. The computer-implemented method of claim 1 , wherein determining an optimal write IO request size comprises automatically selecting an optimal write IO request size from a predetermined list of write IO request sizes based on a previous read IO request size.
4. The computer-implemented method of claim 1 , wherein interleaving sub-IO write requests with read IO requests comprises interleaving the sub-IO write requests with the read IO requests while maintaining a transaction priority of the IO request transactions.
5. The computer-implemented method of claim 4 , wherein the transaction priority is maintained by servicing sub-IO write requests and pending read IO requests based on an arrival time and a completion rate of the respective IO requests.
6. The computer-implemented method of claim 1 , wherein sub-IO write requests are interleaved by grouping sub-IO write requests with child read IO requests of the associated IO request transaction or child read IO requests of other read IO request transactions to be serviced.
7. A non-transitory machine-readable medium storing instructions which, when executed by one or more processors, cause the one or more processors to perform operations comprising:
receiving a plurality of input/output (IO) request transactions at the storage system having a plurality of storage devices;
for each of the plurality of IO request transactions, determining a plurality of child IO requests required to complete the IO request transaction;
tagging the IO request transaction and the associated child IO requests with a unique tag identifier;
for each of the child IO requests that is a write IO request, determining an optimal write IO request size by analyzing IO sequences on an IO bus having mixed read and write IO requests at a fixed point in time to determine an optimal write IO request size for that fixed point in time so that the optimal write IO request size can be applied to a subsequent write IO request;
segmenting the write IO request into a plurality of sub-IO write requests, each having an the optimal write IO request size; and
interleaving sub-IO write requests with read IO requests for servicing to avoid impact in performance to the read IO requests for a mixed IO workload.
8. The non-transitory machine-readable medium of claim 7 , wherein determining an optimal write IO request size comprises dynamically computing an optimal sub-IO write request size during run time to achieve optimal read IO request latency for a mixed IO workload.
9. The non-transitory machine-readable medium of claim 7 , wherein determining an optimal write IO request size comprises automatically selecting an optimal write IO request size from a predetermined list of write IO request sizes based on a previous read IO request size.
10. The non-transitory machine-readable medium of claim 7 , wherein interleaving sub-IO write requests with read IO requests comprises interleaving the sub-IO write requests with the read IO requests while maintaining a transaction priority of the IO request transactions.
11. The non-transitory machine-readable medium of claim 10 , wherein the transaction priority is maintained by servicing sub-IO write requests and pending read IO requests based on an arrival time and a completion rate of the respective IO requests.
12. The non-transitory machine-readable medium of claim 7 , wherein sub-IO write requests are interleaved by grouping sub-IO write requests with child read IO requests of the associated IO request transaction or child read IO requests of other read IO request transactions to be serviced.
13. A data processing system, comprising:
a processor; and
a memory couple to the process to store instructions, which when executed by the processor, cause the processor to perform operations, the operations include receiving a plurality of input/output (IO) request transactions at the storage system having a plurality of storage devices;
for each of the plurality of IO request transactions, determining a plurality of child IO requests required to complete the IO request transaction;
tagging the IO request transaction and the associated child IO requests with a unique tag identifier;
for each of the child IO requests that is a write IO request, determining an optimal write IO request size by analyzing IO sequences on an IO bus having mixed read and write IO requests at a fixed point in time to determine an optimal write IO request size for that fixed point in time so that the optimal write IO request size can be applied to a subsequent write IO request;
segmenting the write IO request into a plurality of sub-IO write requests, each having an the optimal write IO request size; and
interleaving sub-IO write requests with read IO requests for servicing to avoid impact in performance to the read IO requests for a mixed IO workload.
14. The system of claim 13 , wherein determining an optimal write IO request size comprises dynamically computing an optimal sub-IO write request size during run time to achieve optimal read IO request latency for a mixed IO workload.
15. The system of claim 13 , wherein determining an optimal write IO request size comprises automatically selecting an optimal write IO request size from a predetermined list of write IO request sizes based on a previous read IO request size.
16. The system of claim 13 , wherein interleaving sub-IO write requests with read IO requests comprises interleaving the sub-IO write requests with the read IO requests while maintaining a transaction priority of the IO request transactions.
17. The system of claim 16 , wherein the transaction priority is maintained by servicing sub-IO write requests and pending read IO requests based on an arrival time and a completion rate of the respective IO requests.
18. The system of claim 13 , wherein sub-IO write requests are interleaved by grouping sub-IO write requests with child read IO requests of the associated IO request transaction or child read IO requests of other read IO request transactions to be serviced.