Emc Corporation

Emc IP Holding Company LLC

Infinidat Ltd.

Qun Fan

Stephen Richard Ives

Ofer Michael

Rong Yu

Orit Levin-Michael

Haim Helman

Haim Kopylovitz

Steven T McClure

Dan Aharoni

Hui Wang

Yechiel Yochai

Peng Yin

Eli Shagam

Jerome Cartmell

John F Madden, Jr

Robert DeCrescenzo

Roii Raz

John W Lefferts

Pei-Ching Hwang

Venu Nayar

Hagit Brit-Artzi

Michael Dudnik

Gabi Hershkovitz

Michael J Scharland

Salil Dangi

Scott Auchmoody

Scott H Ogata

Zvi Gabriel Benhanokh

Jiahui Wang

Hagay Dagan

Ran Margalit

A method for service level agreement allocation of permanent storage layer resources of a storage system, the method may include monitoring, by a control layer of the storage system, actual performances of the storage system that are related to multiple logical volumes; calculating actual-to-required relationships between the actual performances and service level agreement defined performances of the multiple logical volumes; allocating, for different logical volumes of the multiple logical volumes, allowable amounts of pending access requests; wherein the allocating is based on, at least, the actual-to-required relationships; receiving, by the control layer, received access requests aimed to one or more logical volumes of the logical volumes; and sending to the permanent storage layer requests to serve at least some of the received access requests, based on the allowable amounts of pending access requests.

Service level agreement based management of pending access requests

A method for service level agreement (SLA) allocation of resources of a cache memory of a storage system, the method may include monitoring, by a control layer of the storage system, actual performances of the storage system that are related to multiple logical volumes; calculating actual-to-required relationships between the actual performances and SLA defined performances of the multiple logical volumes; assigning caching priorities, to different logical volumes of the multiple logical volumes; wherein the assigning is based on, at least, the actual-to-required relationships; and managing, based on at least the caching priorities, a pre-cache memory module that is upstream to the cache module and is configured to store write requests that (i) are associated with one or more logical volumes of the different logical volumes and (ii) are received by the pre-cache memory module at points in time when the cache memory is full; wherein the managing comprises transferring one or more write requests from the pre-cache memory module to the cache memory when the cache memory is not full.

Service level agreement based management of a pre-cache module

Protected data is divided into a plurality of segments. For each segment in the plurality of segments, a fingerprint is generated. The plurality of fingerprints is grouped into group(s) of fingerprints and for each group of fingerprints, a group fingerprint is generated. It is determined if the group fingerprint being processed is stored in a cache and if so, the group fingerprint being processed is flagged as one for which a virtual copy will be performed.

Client side cache for deduplication backup systems

Described are techniques for processing a data request for a device of a data storage system. A data request directed to a device that is a virtualized device of the data storage system is received. The data storage system has a plurality of devices comprising one or more local devices and one or more virtualized devices. Each of the local devices is a storage device included in the data storage system and each of the virtualized devices is an external storage device with respect to the data storage system and is presented as a local storage device of the data storage system. First processing is performed by a disk adapter in connection with servicing the data request for the device. The first processing includes sending the request to an external storage device corresponding to the virtualized device over a connection in accordance with path information obtained from discovery processing.

Techniques for processing data requests directed to virtualized devices

Described are techniques for processing a data operation in a data storage system. A front-end component of the data storage system receives the data operation. In response to receiving the data operation, the front-end component performs first processing. The first processing includes determining whether the data operation is a read operation requesting to read a data portion which results in a cache miss; and if said determining determines that the data operation is a read operation resulting in a cache miss, performing read miss processing. Read miss processing includes sequential stream recognition processing performed by the front-end component to determine whether the data portion is included in a sequential stream.

Techniques for data prefetching

Described are techniques for performing data migration for a source logical volume and a target. The target is configured as storage for another mirror of the source logical volume prior to copying data from the source logical volume to the target, and if the target is configured storage of another logical volume of the data storage system, the configured storage is remapped as storage for another mirror of the source logical volume prior to copying data from the source logical volume to the target. One or more invalid bits are set indicating that the target does not contain a valid copy of data from the source logical volume. Data is copied from the first mirror of the source logical volume to the target. Invalid bits are cleared as data portions of the first mirror of the source logical volume are copied to the target.

Techniques for performing data migration

Described are techniques for representing and using RAID group consistency information. RAID group consistency information for a slice of data included on a device may be obtained by mapping the slice to an index. The index is one of a plurality of indices each associated with a structure indicating RAID group inconsistencies for a different portion of slices of data of the device. If the index is not associated with a structure, it may be determined that there is no RAID group inconsistency for the slice. If the index is associated with a structure, a hint bit associated with the slice may be determined wherein the hint bit encodes summary consistency information for a plurality of slices. A determination is made as to whether the hint bit is set, and, if the hint bit is not set, it is determined that there is no RAID group inconsistency for the slice.

Techniques for representing and storing RAID group consistency information

Described are techniques for determining where to store data and parity information in a RAID group for a data track having a track number. A stripe number is provided. A first member identifier is determined which identifies a RAID group member at which data information of said data track is stored. A second member identifier of a second member of said RAID group is determined to store a first set of parity information for said data track. A third member identifier of a third member of said RAID group is determined to store a last set of parity information for said data track. It is determined whether the third member identifier is less than the second member identifier. If the third member identifier is less than said second member identifier, the first member identifier is updated. Also described are techniques for determining a track identifier of a data track stored at a location in a RAID group.

Techniques for determining physical data layout of RAID devices

Described are techniques for performing data recovery processing for two failed devices included in a set of devices. It is determined whether the two failed devices are each one of the data devices, a horizontal parity device, or a diagonal parity device. Each of the data devices is partitioned into N&#x2212;1 data segments. The horizontal parity device includes N horizontal parity segments each including parity information for a horizontal grouping of N&#x2212;1 of the data segments. The diagonal parity device includes N diagonal parity segments each including parity information for a diagonal grouping of N&#x2212;1 of the data segments. Recovery processing is performed to recover two failed devices in accordance with the diagonal parity device, the horizontal parity device, and the data devices.

Techniques for fault tolerant data storage

Accessing data memory includes writing data to a first memory location and to a second memory location in response to a request to write data to a memory address that corresponds to both locations, where the first and second memory locations are mirrored, in response to a request to read data from the memory address, reading data from the first memory location or the second memory location based on load balancing, and accessing data from the second memory location in response to a request to access data at the memory address when memory hardware corresponding to the first memory location has failed. Accessing the data memory may include requesting access to a specific one of the first and second memory locations. The memory address may contain a portion that is common to both the first memory location and the second memory location. Hardware coupled to the memory may cause data written using the memory address to be automatically written to the first memory location and the second memory location.

Mirrored memory

Handling a faulting memory of a pair of mirrored memories includes initially causing a non-faulting memory of the pair of mirrored memories to service all read and write operations for the pair of mirrored memories, determining that hardware corresponding to the faulting memory of the pair of mirrored memories has been successfully replaced to provide a new memory, in response to the new memory being provided, causing data to be copied from the non-faulting memory to the new memory while data is being read to and written from the non-faulting memory, and, in response to successful copying to the new memory, causing writes to be performed to both memories of the pair of mirrored memories and selecting one of the pair of mirrored memories for read operations when one or more read operations are performed. Handling a faulting memory may also include, in response to a write being performed to the non-faulting memory while data is being copied from the non-faulting memory to the new memory, causing the write to be performed to the non-faulting memory and the new memory.

Handling memory faults for mirrored memory

Growing community of inventors

Northborough, MA, United States of America

Qun Fan