Abstract: Techniques for configuring a new data storage system generate IOPS distribution models from activity data of an existing data storage system and use the IOPS distribution models to generate a storage configuration for the new data storage system. Such techniques assign LUNs of the existing storage system to groups based on their input-out (I/O) characteristics and provide a storage configuration tailored to the expected IOPS distributions of each group of LUNs based on the IOPS distribution models and the activity data.

Abstract: Techniques for processing large data sets in parallel in a cloud-networked environment. The cloud-networked environment includes a plurality of remote systems and a cloud-based management system. The cloud-based management system can receive a large set of events from the respective remote systems, place the events, in the order they are received, in a task queue of an elastic scalable thread pool including a plurality of processing threads, process the respective events in the queue asynchronously and in parallel by the plurality of processing threads to produce a plurality of transaction outcomes, respectively, and generate a stream of the transaction outcomes in the same serial order as the events corresponding to the respective outcomes were received. By generating an ordered serial stream of transaction outcomes that corresponds to the respective events placed in the task queue, an analysis of conditions underlying the respective events can be performed more efficiently and accurately.

Abstract: There are disclosed techniques for use in configuring a data storage system. The techniques disclose defining a first workload including a first storage capacity requirement, a first IOPS requirement and a first skew value describing an expected workload profile. The techniques also disclose determining a percentage of the storage capacity requirement to be allocated from a storage tier based on a policy as well as allocating an amount of storage capacity from the storage tier in accordance with the percentage of the first storage capacity requirement. The techniques also disclose utilizing the first skew value to map the percentage of the first storage capacity requirement to a corresponding IOPS percentage to be handled by the storage tier. Further, the techniques disclose determining a first distribution of IOPS to the allocated storage capacity based on the IOPS percentage and the first IOPS requirement.

Abstract: Techniques for processing large data sets in parallel in a cloud-networked environment. The cloud-networked environment includes a plurality of remote systems and a cloud-based management system. The cloud-based management system can receive a large set of events from the respective remote systems, place the events, in the order they are received, in a task queue of an elastic scalable thread pool including a plurality of processing threads, process the respective events in the queue asynchronously and in parallel by the plurality of processing threads to produce a plurality of transaction outcomes, respectively, and generate a stream of the transaction outcomes in the same serial order as the events corresponding to the respective outcomes were received. By generating an ordered serial stream of transaction outcomes that corresponds to the respective events placed in the task queue, an analysis of conditions underlying the respective events can be performed more efficiently and accurately.

Abstract: There is disclosed a technique for use in memory allocation. In one embodiment, the technique comprises a method. The method comprises the step of determining the size of an item. The method also comprises the step of determining whether to allocate the item to stack or heap memory based on the size of the item.

Abstract: A technique evaluates a set of storage devices (e.g., magnetic disk drives, solid state drives, etc.). The technique involves receiving, by processing circuitry (e.g., a storage processor, a standalone computer, etc.), storage device evaluation factors which (i) map possible storage device error events to individual weights and (ii) map cumulative weights to recommended activities. The technique further involves receiving, by the processing circuitry, a storage device error log containing storage device error entries identifying actual storage device error events which were encountered by the set of storage devices while performing data storage operations over a period of time. The technique further involves analyzing, by the processing circuitry, the storage device error entries based on the storage device evaluation factors to produce a set of evaluation results identifying a set of recommended activities to be performed on the set of storage devices.

Abstract: A method is used in managing cache systems of storage systems. Information is gathered from a storage system to determine an estimated amount of time required to flush data of a nonvolatile cache of the storage system to storage devices. The data cached in the nonvolatile cache is associated with a set of logical objects configured to cache the data in the nonvolatile cache. Based on the information gathered from the storage system, the estimated amount of time required to flush the data of the nonvolatile cache is determined.

Abstract: The system and method filters out benign errors and triaging errors that are not filtered. The errors that are not filtered are triaged by categorizing the error and in response to the categorized error either resolving the error by executing code to proactively test the error and repair it; or collect the data necessary to perform deeper analysis by the customer service center and forwarding the collected data to the customer service center.