Distribution of quantities of an increased workload portion into buckets representing operations
In some examples, a computing system receives an indication of an increased workload portion to be added to a workload of a storage system, the workload comprising buckets of operations of different characteristics. The computing system computes, based on quantities of operations of the different characteristics in the workload, factor values that indicate distribution of operations of the increased workload portion to the buckets of operations of the different characteristics, and distributes, according to the factor values, the operations of the increased workload portion into the buckets of operations of the different characteristics.
1. A non-transitory machine-readable storage medium comprising instructions that upon execution cause a computing system to:
receive an indication of an increased workload portion to be added to a workload of a storage system;
access a representation of the workload, the representation containing quantities of operations of different characteristics;
compute, based on the quantities of operations of the different characteristics included in the representation, factor values that indicate a distribution of quantities of operations of the increased workload portion to the different characteristics;
update, using the factor values, the representation to represent a modified workload that is a combination of the workload and the increased workload portion; and
predict a performance of the storage system using a predictive model based on the updated representation.
2. The non-transitory machine-readable storage medium of claim 1 , wherein the different characteristics comprise different input/output (I/O) sizes, and the representation contains quantities of operations of the different I/O sizes.
3. The non-transitory machine-readable storage medium of claim 1 , wherein the updated representation representing the modified workload contains updated quantities of operations of the different characteristics.
4. The non-transitory machine-readable storage medium of claim 3 , wherein the updated quantities of operations of the different characteristics comprise a first updated quantity of operations of a first characteristic and a second updated quantity of operations of a second characteristic different from the first characteristic, wherein the first updated quantity is based on a quantity of operations of the first characteristic in the workload and a quantity of operations of the first characteristic in the increased workload portion according to the distribution, and wherein the second updated quantity is based on a quantity of operations of the second characteristic in the workload and a quantity of operations of the second characteristic in the increased workload portion according to the distribution.
5. The non-transitory machine-readable storage medium of claim 1 , wherein the instructions upon execution cause the computing system to:
compute a first factor value of the factor values based on correlating a quantity of operations of a first characteristic of the different characteristics in the workload to a total quantity of operations in the workload.
6. The non-transitory machine-readable storage medium of claim 5 , wherein the quantity of operations of the first characteristic in the workload is in a window of time intervals, and the total quantity of operations in the workload is in the window of time intervals.
7. The non-transitory machine-readable storage medium of claim 6 , wherein the window is a rolling window, and the computing of the factor values is based on quantities of operations of the different characteristics in rolling windows of time intervals.
8. The non-transitory machine-readable storage medium of claim 5 , wherein the computing of the first factor value further comprises normalizing a correlation value produced by the correlating of the quantity of operations of the first characteristic to the total quantity of operations.
9. The non-transitory machine-readable storage medium of claim 8 , wherein the normalizing of the correlation value is based on a ratio of the quantity of operations of the first characteristic to the total quantity of operations.
10. The non-transitory machine-readable storage medium of claim 1 , wherein the predicting of the performance of the storage system comprises predicting a performance of one or more storage volumes in the storage system.
11. The non-transitory machine-readable storage medium of claim 1 , wherein the predicted performance comprises a predicted latency of the storage system or a resource of the storage system, or a predicted saturation of the storage system or a resource of the storage system.
12. A computing system comprising:
a processor; and
a non-transitory storage medium storing instructions executable on the processor to:
receive an indication of an increased workload portion to be added to a workload of a storage system, the workload represented by a representation containing quantities of input/output (I/O) operations of different characteristics;
compute, based on the quantities of I/O operations of the different characteristics included in the representation, factor values that indicate a distribution of quantities of I/O operations of the increased workload portion to the different characteristics;
generate, using the factor values, a modified representation to represent a modified workload that includes the workload and the increased workload portion; and
predict a performance of the storage system using a predictive model based on the modified representation.
13. The computing system of claim 12 , wherein the indication of the increased workload portion is based on an indication that the workload is to be increased by a specified percentage.
14. The computing system of claim 12 , wherein the representation comprises a histogram of the quantities of I/O operations of the different characteristics in a plurality of time intervals.
15. The computing system of claim 12 , wherein the modified representation representing the modified workload contains updated quantities of I/O operations of the different characteristics.
16. The computing system of claim 12 , wherein the instructions are executable on the processor to compute a respective factor value of the factor values based on correlating a quantity of I/O operations of a respective characteristic of the different characteristics in the workload to a total quantity of I/O operations in the workload.
17. The computing system of claim 16 , wherein the quantity of I/O operations of the respective characteristic in the workload is in a window of time intervals, and the total quantity of operations in the workload is in the window of time intervals.
18. The computing system of claim 16 , wherein the computing of the respective factor value further comprises normalizing a correlation value produced by the correlating of the quantity of operations of the respective characteristic to the total quantity of operations, wherein the normalizing of the correlation value is based on a ratio of the quantity of I/O operations of the respective characteristic to the total quantity of I/O operations.
19. A method performed by a system comprising a hardware processor, comprising:
receiving an indication of an increased workload portion to be added to a workload of a storage system;
access a representation of the workload, the representation containing quantities of operations of different input/output (I/O) sizes;
compute, based on the quantities of operations of the different I/O sizes included in the representation, factor values that indicate a distribution of quantities of operations of the increased workload portion to the different I/O sizes;
update, using the factor values, the representation to represent a modified workload that is a combination of the workload and the increased workload portion; and
predict a performance of the storage system using a predictive model based on the updated representation.
20. The method of claim 19 , wherein the computing of the factor values comprises computing each respective factor value of the factor values based on correlating a quantity of operations of a respective I/O size of the different I/O sizes in the workload to a total quantity of I/O operations in the workload.