System and method for an efficient cache warm-up

A method comprising:

• using computer hardware for performing:
• identifying, by a processor of a storage system, a data block stored on a primary storage of the storage system device based on one or more data block identification policies;
• determining latencies associated with the data block, the latencies including a cache memory latency specifying a first estimated amount of time to access the identified data block if the identified data block is stored in the cache memory, a primary storage device latency specifying a second estimated amount of time to access the identified data block from the primary storage device, and a future primary storage device latency specifying a third estimated amount of time to access the identified data block from the primary storage device at a future point in time;
• calculating an effective latency for the identified data block based on the cache memory latency, the primary storage device latency, and the future primary storage device latency;
• and
• storing, on the cache memory, the identified data block when the effective latency for the identified data block is above a threshold effective latency value.

A non-transitory computer readable storage medium containing one or more program instructions configured for execution by a processor and further configured to be stored on the non-transitory computer readable storage medium, the program instructions when executed operable to:

• receive a data block segment size;
• identify a data block segment stored on the primary storage device of a storage system based on an identification policy, each data block segment including a number of data blocks corresponding to the data block segment size;
• determine latencies associated with the identified the data block segment, the latencies including a cache memory latency that is a first estimated amount of time to access the identified data block segment if the identified data block segment is stored on a cache memory associated with a storage system, a primary storage device latency that is a second estimated amount of time to access the identified data block segment from the primary storage device, and a future primary storage device latency that is a third estimated amount of time to access the identified data block segment from the primary storage device at a future point in time;
• calculate an effective latency associated with the identified data block segment by subtracting the cache memory latency from the future primary storage device latency and dividing a result of the subtraction by the primary storage device latency;
• and
• copy, to the cache memory, the data block segment based on the effective latency.