Abstract: Described is a process and device for accessing data stored in multiple logical volumes. The data are replicated on first and second storage elements, such as the redundant hard disk drives of a disk mirror. The multiple logical volumes are divisible into a first logical volume and a second logical volume. All read requests targeting the first logical volume are directed to one of the first and second storage elements. Read requests targeting the second logical volume are asymmetrically interleaved between the first and second storage elements. An asymmetric interleave ratio is determined and implemented that substantially balances the read requests to the multiple logical volumes between the first and second storage elements.

Abstract: Destaging activities in a data storage system are controlled by providing a write pending list of elements, where each element is defined to store information related to a cache memory data element for which a write to storage is pending, and maintaining the write pending list so that destaging of a data element can be based on the maturity of the pending write.

Abstract: A method for implementing the invention is carried out in a data-storage system having a data storage unit that includes at least two constituent data storage elements. Each of the constituent data storage elements is either in a first state or a second state. The method includes providing a data structure having an entry corresponding to the data storage unit. The entry includes status information indicating whether at least one constituent data storage element of the data storage unit is in the first state. These entries are updated as necessary following any changes in state of the constituent data storage element. Scanning the data storage units instead of the data storage elements provides a more efficient way to locate data storage elements in the first state, particularly where such data storage elements are rare.

Abstract: This invention is a system and method for preparing workload data that may be accessed in a data storage environment and then replayed for testing or other reasons. The invention has the capability to prepare a trace of I/Os to a data storage system for an extended period of time for replaying these I/Os back to that or another system. Variables may also be manipulated through a preparation process and is particularly useful for customizing benchmarking tests, or consolidation, or trouble-shooting, or capacity planning.

Abstract: Described are techniques and criteria used in connection with cache management. The cache may be organized as a plurality of memory banks in which each memory bank includes a plurality of slots. Each memory bank has an associate control slot that includes groups of extents of tags. Each cache slot has a corresponding tag that includes a bit value indicating the availability of the associated cache slot, and a time stamp indicating the last time the data in the slot was used. The cache may be shared by multiple processors. Exclusive access of the cache slots is implemented using an atomic compare and swap instruction. The time stamp of slots in the cache may be adjusted to indicate ages of slots affecting the amount of time a particular portion of data remains in the cache.

Abstract: A seek minimizing process for use in storage controller of a data storage system is presented. The process collects statistics for physical devices storing mirrored logical volumes. The statistics include utilization and activity level. The process determines from the statistics the top n busiest physical devices. The process evaluates each of the n physical devices for mirror policy adjustment that will result in reduced seek time when the utilization of the physical device is greater than a threshold, for example, 50%.

Abstract: Destaging activities in a data storage system are controlled by providing a write pending list of elements, where each element is defined to store information related to a cache memory data element for which a write to storage is pending, and maintaining the write pending list so that destaging of a data element can be based on the maturity of the pending write.

Abstract: This invention is a system and method for preparing workload data that may be accessed in a data storage environment and then replayed for testing or other reasons. The invention has the capability to prepare a trace of I/Os to a data storage system for an extended period of time for replaying these I/Os back to that or another system. Variables may also be manipulated through a preparation process and is particularly useful for customizing benchmarking tests, or consolidation, or trouble-shooting, or capacity planning.

Abstract: This invention is a system and method for preparing workload data that may be accessed in a data storage environment and then replayed for testing or other reasons. The invention has the capability to prepare a trace of I/Os to a data storage system for an extended period of time for replaying these I/Os back to that or another system. Variables may also be manipulated through a preparation process and is particularly useful for customizing benchmarking tests, or consolidation, or trouble-shooting, or capacity planning.

Abstract: A prefetch process that generates prefetch tasks for short sequences that are no longer than n tracks in length. The value of n is selected as 8. The prefetch process maintains a history of short sequences, uses that history to predict an expected length of a current sequence and generates a short prefetch task based on that prediction. The historical short sequence data is stored in histograms, each histogram being associated with a different logical volume. The histograms store a cumulative count of sequence occurrences of a given sequence length for each sequence length in a range of 1 track to n tracks. The process applies a probability-based threshold to its prediction to control the aggressiveness of the prefetch task to be generated. The threshold is adjusted based on system activity level metrics, such as processor utilization and average memory access time.

Abstract: Techniques and criteria are used in connection with promoting a slot within a cache in the form of a replacement queue. A cache slot may be promoted based on an inequality that considers the following criteria: probability of losing a cache hit, gaining a cache hit, and the price or cost associated with promoting a slot. The foregoing criteria may be used in accordance with a predetermined promotion policy when the replacement queue is in a locked state and an unlocked state, or only when the replacement queue is in a locked state. Different costs may be associated with the state of the replacement queue as locked or unlocked as the replacement queue may be locked in connection with operations that are performed on the replacement queue. The cost associated with a locked replacement queue may be different than the cost associated with an unlocked replacement queue. Different thresholds and values associated with the foregoing criteria may be specified as dynamic system parameters.