Greener software defined storage stack
A method for managing client resources by receiving a desired load factor representing the number of instructions being executed per second (IOPS) to implement an application on a set of cores of a client device, based on the desired load factor and a latency factor, determining a maximum number of IOPS that can be executed by the cores of the client device before reaching system saturation, determining a pattern of the IOPS being executed on the set of cores based on historical IOPS information for the latency factor, and based on the historical IOPS information, determining to execute the IOPS on a subset of the set of cores.
1. A method of managing client device resources comprising:
receiving a desired load factor representing a number of instructions executed in input/output operations per second (IOPS) to implement an application on a set of cores of the client device;
based on the desired load factor and a received latency factor of the application, determining a maximum number of IOPS that can be executed by the set of cores of the client device before reaching system saturation on the set of cores;
determining a pattern of the IOPS being executed on the set of cores by using a received historical IOPS information for the received latency factor as an input to a machine learning model;
based on the pattern, the desired load factor, the received latency factor, and the maximum number of IOPS, determining to execute the application on at least one core of the set of cores, wherein the at least one core is a minimum number of cores to execute the application; and
executing the application on the at least one core.
2. The method of claim 1 , wherein the received historical IOPS information represents a first number of IOPS executed at a time prior to a current time.
3. The method of claim 2 , wherein determining the pattern based on the received historical IOPS information further comprises determining that a second number of IOPS to be executed by the set of cores is lower than a third number of IOPS currently being executed by the set of cores.
4. The method of claim 1 , further comprising:
making a second determination based on the received historical IOPS information that a first number of IOPS to be executed by the set of cores is higher than a second number of IOPS currently being executed by the set of cores;
determining, based on the second determination, to execute the application on at least one additional core in addition to the at least one core; and
executing the application one the at least on additional core in addition to the at least one core.
5. A non-transitory computer readable medium comprising computer readable program code, which when executed by a computer processor enables the computer processor to perform a method for client device resource management, the method comprising:
receiving a desired load factor representing a number of instructions executed in input/output operations per second (IOPS) to implement an application on a set of cores of a client device;
based on the desired load factor and a received latency factor of the application, determining a maximum number of IOPS that can be executed by the set of cores of the client device before reaching system saturation on the set of cores;
determining a pattern of the IOPS being executed on the set of cores by using a received historical IOPS information for the received latency factor as an input to a machine learning model;
based on the pattern, the desired load factor, the received latency factor, and the maximum number of IOPS, determining to execute the application on at least one core of the set of cores, wherein the at least one core is a minimum number of cores to execute the application; and
executing the application on the at least one core.
6. The non-transitory computer readable medium of claim 5 , wherein the received historical IOPS information represents a first number of IOPS executed at a time prior to a current time.
7. The non-transitory computer readable medium of claim 6 , wherein the determining based on the received historical IOPS information comprises determining that a second number of IOPS to be executed by the set of cores is lower than a third number of IOPS currently being executed by the set of cores.
8. The non-transitory computer readable medium of claim 5 , further comprising:
making a second determination based on the received historical IOPS information that a first number of IOPS to be executed by the set of cores is higher than a second number of IOPS currently being executed by the set of cores;
determining, based on the second determination, to execute the application on at least one additional core in addition to the at least one core;
executing the application on the at least one additional core in addition to the at least one core.
9. A method of managing system resources comprising:
receiving a desired load factor representing a number of instructions executed in input/output operations per second (IOPS) to implement an application on a set of client devices;
based on the desired load factor and a received latency factor of the application, determining a maximum number of IOPS that can be executed by the system before reaching system saturation on the set of client devices;
determining a pattern of the IOPS being executed on the set of client devices by using a received historical IOPS information for the received latency factor as an input to a machine learning model;
based on the pattern, the desired load factor, the received latency factor, and the maximum number of IOPS, determining to execute the application on at least one client device of the set of client devices, wherein the at least one client device is a minimum number of client devices to execute the application; and
executing the application on the at least one client device.
10. The method of claim 9 , wherein the received historical IOPS information represents a first number of IOPS executed at a time prior to a current time.
11. The method of claim 10 , wherein the determining based on the received historical IOPS information comprises determining that a second number of IOPS to be executed by the set of client devices is lower than a third number of IOPS currently being executed by the set of client devices.
12. The method of claim 9 , further comprising:
making a second determination based on the received historical IOPS information that a first number of IOPS to be executed by the set of client devices is higher than a second number of IOPS currently being executed by the set of client devices;
determining, based on the second determination, to execute the application on at least one additional client device in addition to the at least one client device; and
executing the application on the at least one additional client device in addition to the at least one client device.
13. The method of claim 9 , wherein determining to execute the application on the at least one client device of the set of client devices is further based on memory availability of the set of client devices.