Cache management system and method

1. A computer-implemented method, executed on a computing device, comprising:

determining a queue depth and a flush rate for each of a plurality of pending data queues associated with a cache system of a backend system, thus defining a queue depth/flush rate pair for each of the plurality of pending data queues, wherein each of the plurality of pending data queues is associated with, and configured to temporarily store all data to be written to, a respective volume group chosen from the plurality of volume groups, thus defining a plurality of pending data queue volume group pairs, and wherein the queue depth is based upon, at least in part, an amount of data temporarily stored within the respective pending data queue of each pending data queue volume group pair and the flush rate is based upon, at least in part, the rate at which the data is written to the respective volume group of each pending data queue volume group pair;

determining a predicted drain time for each of the plurality of pending data queues based, at least in part, upon the queue depth/flush rate pair, thus defining a plurality of predicted drain times that are respectively associated with the plurality of pending data queue volume group pairs;

comparing the plurality of predicted drain times to one another to identify a high-utilization volume group, chosen from the plurality of volume groups, wherein the high-utilization volume group utilizes a higher-than-average portion of the cache system based upon, at least in part, the predicted drain times for each of the plurality of pending data queues; and

regulating the rate at which data is allowed to enter a high-utilization pending data queue based, at least in part, upon the queue depth/flush rate pair defined for the high-utilization pending data queue including delaying the issuance of write acknowledgements for write requests associated with data entering the high-utilization pending data queue and not delaying the issuance of write acknowledgements for write requests associated with data entering a non-high-utilization pending data queue.

2. The computer implemented method of claim 1 wherein regulating the rate at which data is allowed to enter the high-utilization pending data queue includes:

throttling the issuance of write acknowledgements to a host that is providing the data to the high-utilization pending data queue so that the rate at which data is entering the high-utilization pending data queue is equal to the rate at which data is leaving the high-utilization pending data queue.

3. A computer program product residing on a non-transitory computer readable medium having a plurality of instructions stored thereon which, when executed by a processor, cause the processor to perform operations comprising:

determining a queue depth and a flush rate for each of a plurality of pending data queues associated with a cache system of a backend system, thus defining a queue depth/flush rate pair for each of the plurality of pending data queues, wherein each of the plurality of pending data queues is associated with, and configured to temporarily store all data to be written to, a respective volume group chosen from the plurality of volume groups, thus defining a plurality of pending data queue volume group pairs, and wherein the queue depth is based upon, at least in part, an amount of data temporarily stored within the respective pending data queue of each pending data queue volume group pair and the flush rate is based upon, at least in part, the rate at which the data is written to the respective volume group of each pending data queue volume group pair;

determining a predicted drain time for each of the plurality of pending data queues based, at least in part, upon the queue depth/flush rate pair, thus defining a plurality of predicted drain times that are respectively associated with the plurality of pending data queue volume group pairs;

comparing the plurality of predicted drain times to one another to identify a high-utilization volume group, chosen from the plurality of volume groups, wherein the high-utilization volume group utilizes a higher-than-average portion of the cache system based upon, at least in part, the predicted drain times for each of the plurality of pending data queues; and

regulating the rate at which data is allowed to enter a high-utilization pending data queue based, at least in part, upon the queue depth/flush rate pair defined for the high-utilization pending data queue including delaying the issuance of write acknowledgements for write requests associated with data entering the high-utilization pending data queue and not delaying the issuance of write acknowledgements for write requests associated with data entering a non-high-utilization pending data queue.

4. The computer program product of claim 3 wherein regulating the rate at which data is allowed to enter the high-utilization pending data queue includes:

throttling the issuance of write acknowledgements to a host that is providing the data to the high-utilization pending data queue so that the rate at which data is entering the high-utilization pending data queue is equal to the rate at which data is leaving the high-utilization pending data queue.

5. A computing system including a processor and memory configured to perform operations comprising:

determining a queue depth and a flush rate for each of a plurality of pending data queues associated with a cache system of a backend system, thus defining a queue depth/flush rate pair for each of the plurality of pending data queues, wherein each of the plurality of pending data queues is associated with, and configured to temporarily store all data to be written to, a respective volume group chosen from the plurality of volume groups, thus defining a plurality of pending data queue volume group pairs, and wherein the queue depth is based upon, at least in part, an amount of data temporarily stored within the respective pending data queue of each pending data queue volume group pair and the flush rate is based upon, at least in part, the rate at which the data is written to the respective volume group of each pending data queue volume group pair;

determining a predicted drain time for each of the plurality of pending data queues based, at least in part, upon the queue depth/flush rate pair, thus defining a plurality of predicted drain times that are respectively associated with the plurality of pending data queue volume group pairs;

comparing the plurality of predicted drain times to one another to identify a high-utilization volume group, chosen from the plurality of volume groups, wherein the high-utilization volume group utilizes a higher-than-average portion of the cache system based upon, at least in part, the predicted drain times for each of the plurality of pending data queues; and

regulating the rate at which data is allowed to enter a high-utilization pending data queue based, at least in part, upon the queue depth/flush rate pair defined for the high-utilization pending data queue including delaying the issuance of write acknowledgements for write requests associated with data entering the high-utilization pending data queue and not delaying the issuance of write acknowledgements for write requests associated with data entering a non-high-utilization pending data queue.

6. The computing system of claim 5 wherein regulating the rate at which data is allowed to enter the high-utilization pending data queue includes:

throttling the issuance of write acknowledgements to a host that is providing the data to the high-utilization pending data queue so that the rate at which data is entering the high-utilization pending data queue is equal to the rate at which data is leaving the high-utilization pending data queue.