System and method for preemptively adjusting forecasted memory access workload to minimize data center carbon footprint
An information handling system executing a predictive task adjusting and carbon dioxide (CO2) minimizing system may comprise a hardware processor executing code instructions of a time series forecasting engine to predict a future CO2 emissions value for hardware component within a data center within a future time window that exceeds a CO2 threshold value determined for the hardware component where changes in CO2 emissions are due to a forecasted future load of memory accessing functions causing a change in energy consumption, the hardware processor executing code instructions to identify an adjusted forecasted future load of memory accessing functions whose execution is predicted to decrease CO2 emitted during the future time window below the CO2 threshold value, and the network interface device to transmit a workload redistribution instruction for implementing the adjusted forecasted future load of memory accessing functions at the data center.
1 . An information handling system executing a predictive task adjusting and carbon dioxide (CO2) minimizing system comprising:
a network interface device to receive operational telemetry measurements for a first hardware component that executes memory accessing functions within a data center, including power analytics, a load of the memory accessing functions identified for a memory maintenance function with a memory access type, a geographic location, and a determined CO2 emissions value, wherein the memory access type includes the memory accessing functions selected from a read memory access type, a write memory access type, and a memory maintenance type;
a local hardware processor executing code instructions of a time series forecasting engine modeling a relationship between changes in CO2 emissions values and changes in the operational telemetry measurements for the first hardware component that executes the memory accessing functions within the data center, to predict a future CO2 emissions value for the first hardware component within a future time window that exceeds a CO2 load-balancing threshold value determined for thate first hardware component where changes in CO2 emissions are due to a forecasted future load of the memory accessing functions causing a change in energy consumption at that first hardware component;
the local hardware processor executing code instructions of the predictive task adjusting and CO2 minimizing system to identify an adjusted forecasted future load of the memory accessing functions due to decreasing read compression rate or read decompression rate for the read memory access type, due to decreasing write compression rate or write deduplication rate for the write memory access type, or due to decreasing frequency of execution of the high-power consumption memory maintenance functions for the memory maintenance access type to cause a predicted decrease in CO2 emitted due to execution of the adjusted forecasted future load of the memory accessing functions during the future time window relative to the CO2 load-balancing threshold; and
the network interface device to transmit a workload redistribution instruction for implementing the adjusted forecasted future load of the memory accessing functions at a remote hardware processor of a second hardware component of the data center instead of the first hardware component when the adjusted forecasted future load of the memory accessing functions during the future time window are still above the CO2 load-balancing threshold.
2 . The information handling system of claim 1 , wherein the workload redistribution instructions limits the read compression rate or the write compression rate in the adjusted forecasted future load of the memory accessing functions executing at the second hardware component of the data center.
3 . The information handling system of claim 1 , wherein the workload redistribution instructions limits the read decompression rate in the adjusted forecasted future load of memory accessing functions executing at the second hardware component of the data center.
4 . The information handling system of claim 1 , wherein the workload redistribution instructions limits the write deduplication rate in the adjusted forecasted future load of the memory accessing functions executing at the hardware component of the data center.
5 . The information handling system of claim 1 , wherein the workload redistribution instructions limits an execution frequency for the high-power consumption memory maintenance function in the adjusted forecasted future load of the memory accessing functions executing at the second hardware component of the data center.
6 . The information handling system of claim 1 further comprising:
the local hardware processor executing code instructions to identify a first high-power consumption memory maintenance function in the forecasted future load of the memory accessing functions that is predicted to cause a largest percentage of the future CO2 emissions value relative to other memory maintenance functions in the forecasted future load of memory accessing functions, and identify the adjusted forecasted future load of the memory accessing functions based on adjustment to limit an execution frequency of the first high-powered consumption memory maintenance function.
7 . The information handling system of claim 1 further comprising:
the local hardware processor executing code instructions to identify a first memory access type selected from the read memory access type, the write memory access type, and the memory maintenance type in the forecasted future load of the memory accessing functions that is predicted to cause a largest percentage of the future CO2 emissions value relative to other memory accessing functions of the other memory access types in the forecasted future load of the memory accessing functions, and identify the adjusted forecasted future load of memory accessing functions based on adjustment of the first memory access type.
8 . A method for predictive memory access function task adjustment to minimize carbon dioxide (CO2) emissions executing on an information handling system comprising:
receiving, via a network interface device, operational telemetry measurements for a hardware component that executes memory accessing functions within a data storage system-, including power analytics, a load of the memory accessing functions each identified with a memory access type, a geographic location, and a determined CO2 emissions value, wherein the memory access type includes the memory accessing functions selected from a read memory access type, a write memory access type, and a memory maintenance type;
predicting, via a local hardware processor executing a time series forecasting engine modeling a relationship between changes in CO2 emissions values and changes in the operational telemetry measurements for a first hardware component that executes memory accessing functions within the data storage system, a future CO2 emissions value for the first hardware component within a future time window that exceeds a CO2 load-balancing threshold value determined for the hardware component, where changes in CO2 emissions are due to a forecasted future load of the memory accessing functions;
identifying a first memory maintenance function in the forecasted future load of the memory accessing functions, via the local hardware processor executing a predictive task adjusting and CO2 minimizing system, that is predicted to cause a largest percentage of the future CO2 emissions value relative to other memory accessing functions in the forecasted future load of the memory accessing functions;
determining an adjusted forecasted future load of the memory accessing functions, based on an adjustment of the first memory maintenance function with decreasing frequency of execution of the first memory maintenance function that is of the memory maintenance access type and, that is predicted to decrease CO2 emitted due to execution of the adjusted forecasted future load of the memory accessing functions during the future time window below the CO2 load-balancing threshold;
transmitting, via the network interface device, a workload redistribution instruction for implementing the adjusted forecasted future load of memory accessing functions to adjust the first memory maintenance function at the first hardware resource of the data storage system by decreasing the frequency of execution of the first memory maintenance function; and
transmitting, via the network interface device, a workload redistribution instruction for implementing the adjusted forecasted future load of memory accessing functions to adjust the first memory maintenance function at a second hardware resource having a remote hardware processor of the data storage system when the predicted decrease in CO2 emitted due to execution of the adjusted forecasted future load of the memory accessing functions during the future time window does not fall below the CO2 load-balancing threshold.
9 . The method of claim 8 further comprising:
identifying, via the local hardware processor, a first memory access type selected from the read memory access type of the write memory access type in the forecasted future load of memory accessing functions that is predicted to cause a largest percentage of the future CO2 emissions value relative to other memory access types in the forecasted future load of memory accessing functions, and identify the adjusted forecasted future load of memory accessing functions based on adjustment of the first memory access type by decreasing read compression rate or read decompression rate for the read memory access type, or by decreasing write compression rate or write deduplication rate for the write memory access type.
10 . The method of claim 8 , wherein the workload redistribution instructions specify an encryption scheme for the first memory maintenance function in the adjusted forecasted future load of the memory accessing functions executing at the first hardware component of the data storage system.
11 . The method of claim 8 , wherein the adjustment to the first memory maintenance function includes decreasing the frequency of execution of memory backup functions executing at the hardware component of the data storage system.
12 . The method of claim 8 , wherein the adjustment to the first memory maintenance function includes decreasing the frequency of execution of memory encryption functions executing at the hardware component of the data storage system.
13 . The method of claim 8 , wherein the adjustment to the first memory maintenance function includes decreasing the frequency of execution of memory snapshot functions executing at the hardware component of the data storage system.
14 . The method of claim 8 , wherein the adjustment to the first memory maintenance function includes decreasing the frequency of execution of memory replication functions executing at the hardware component of the data storage system.
15 . An information handling system executing a predictive task adjusting and carbon dioxide (CO2) minimizing system comprising:
a network interface device to receive operational telemetry measurements for a hardware component that executes memory accessing functions within a data storage system, including power analytics, a load of the memory accessing functions identified by a memory maintenance function and a memory access type, a data storage system geographic location, and a determined CO2 emissions value for a first hardware component of the data storage system, wherein the memory access type includes the memory accessing functions selected from a read memory access type, a write memory access type, and a memory maintenance type;
a local hardware processor executing code instructions of a time series forecasting engine modeling a relationship between changes in CO2 emissions values and changes in the operational telemetry measurements for the first hardware component that executes the memory accessing functions within the data storage system to predict a future CO2 emissions value for the first hardware component within a future time window;
the local hardware processor determining that the predicted future CO2 emissions value exceeds a CO2 load-balancing threshold value determined for the first hardware component due to a forecasted future load of the memory accessing functions;
the local hardware processor executing code instructions of the predictive task adjusting and CO2 minimizing system to identify a first memory access type in the forecasted future load of the memory accessing functions that is predicted to cause a largest percentage of the future CO2 emissions value relative to other memory access types in the forecasted future load of the memory accessing functions;
the local hardware processor executing code instructions of the predictive task adjusting and CO2 minimizing system to determine an adjusted forecasted future load of the memory accessing functions based on adjustment of the first memory access type that is predicted to decrease CO2 emitted due to execution of the adjusted forecasted future load of the memory accessing functions during the future time window below the CO2 load-balancing threshold, wherein the adjustment of the first memory type includes decreasing read compression rate or read decompression rate for the read memory access type, decreasing write compression rate or write deduplication rate for the write memory access type, or decreasing frequency of execution high-power consumption memory maintenance functions for the memory maintenance access type; and
the network interface device to transmit a workload redistribution instruction for implementing the adjusted forecasted future load of memory accessing functions to adjust the first memory access type at the first hardware component of the data storage system according to the adjustment of the first memory type having the predicted decrease CO2 emitted below the CO2 load-balancing threshold.
16 . The information handling system of claim 15 further comprising:
the local hardware processor executing code instructions to identify a first high-power consumption memory maintenance function as the first memory access type in the forecasted future load of the memory accessing functions that is predicted to cause a largest percentage of the future CO2 emissions value relative to other memory maintenance functions in the forecasted future load of the memory accessing functions, and determine the adjusted forecasted future load of the memory accessing functions based on adjustment of decreasing the frequency of execution of the first high-power consumption memory maintenance function.
17 . The information handling system of claim 15 , wherein the adjustment of the first memory access type includes decreasing of the read compression rate for memory reads executing at the first hardware component of the data storage system.
18 . The information handling system of claim 15 , wherein the adjustment of the first memory access type includes decreasing the write compression rate for memory writes executing at the first hardware component of the data storage system.
19 . The information handling system of claim 15 , wherein the adjustment of the first memory access type includes decreasing the write deduplication rate for memory writes executing at the first hardware component of the data storage system.
20 . The information handling system of claim 15 , wherein the adjustment of the first memory access type includes decreasing the read decompression rate for memory reads executing at the first hardware component of the data storage system.