Netapp, Inc.

Network Appliance, Inc.

Silicon Graphics, Incorporated

James Leong

Rajesh Sundaram

Srinivasan Viswanathan

Scott A Schoenthal

Atul Goel

Douglas P Doucette

Stephen H Strange

Ratnesh Gupta

Anil Paul Thoppil

Qinghua Zheng

Srishylam Simharajan

Vitaly Revsin

Loellyn Jane Cassell

Vani Vully

James A Taylor

Laurie Costello

Eric Mowat

A copy-free transition tool converts storage objects from a source format associated with a source storage system to a destination format associated with a destination storage system. The transition tool exports configuration information associated with the source storage system to the destination storage system and brings the storage objects offline. Once the transition tool determines that the storage devices containing the storage objects are physically connected to the destination storage system, the tool can convert file systems from the source format to the destination format while leaving data and file layout unchanged. The tool can also modify metadata associated with each of the storage objects to conform to the destination format and store the modified metadata with the destination storage system.

Converting storage objects between formats in a copy-free transition

A data access request to a file system is decomposed into a plurality of lower-level I/O tasks. A logical combination of physical storage components is represented as a hierarchical set of objects. A parent I/O task is generated from a first object in response to the data access request. A child I/O task is generated from a second object to implement a portion of the parent I/O task. The parent I/O task is suspended until the child I/O task completes. The child I/O task is executed in response to an occurrence of an event that a resource required by the child I/O task is available. The parent I/O task is resumed upon an event indicating completion of the child I/O task. Scheduling of any child I/O task is not conditional on execution of the parent I/O task, and a state diagram regulates the child I/O tasks.

Method and apparatus for decomposing I/O tasks in a RAID system

Storage servers use a fast, non-volatile or persistent memory to store data until it can be written to slower mass storage devices such as disk drives. If the server crashes before a write can complete, the data remains safely stored in non-volatile memory. If the data cannot be committed to disk when the server reboots (e.g. because the destination mass storage device is unavailable), it is stored in a file. When the disk reappears, the data in the file may be used to restore a file or filesystem on the disk to a consistent state.

Preventing data loss in a storage system

Embodiments of the present invention disclose a technique for providing an indication whether data stored on a disk drive are invalid. As used herein, invalid data are data written prior to the disk drive being added to an array of the disk drives or data in a block that has become free and which has been removed from the corresponding parity block of the stripe. Knowing that the disk drive was written prior to the drive being added to the existing array or having data which has become invalid allows a storage server to ignore the invalid data and not to use it when computing parity (i.e., a data protection value computed as a result of a logical operation on data blocks in a stripe in the array of disk drives). This, in turn, eliminates the need to zero disk drives or to perform parity re-computation prior to using the disk drives.

System and method for eliminating zeroing of disk drives in RAID arrays

Storage servers use a fast, non-volatile or persistent memory to store data until it can be written to slower mass storage devices such as disk drives. If the server crashes before a write can complete, the data remains safely stored in non-volatile memory. If the data cannot be committed to disk when the server reboots (e.g. because the destination mass storage device is unavailable), it is stored in a file. When the disk reappears, the data in the file may be used to restore a file or file system on the disk to a consistent state.

Redundant Array of Inexpensive Disks (RAID) groups in a storage system are dynamically reconfigured by merging and splitting the RAID groups. When an indication of a change to data or system characteristics is received, disks in the RAID groups of the storage system can be reorganized to adapt to the change.

Dynamically adapting the fault tolerance and performance characteristics of a raid-based storage system by merging and splitting raid groups

A method for operating a data storage system is described. The method first constructs an I/O tree representing a logical configuration of storage devices coupled to the storage system, the I/O tree representing a flow of I/O operations to the storage devices. Elements of the I/O tree are represented by objects. A freeze condition is imposed on a selected object of the I/O tree in order to disable a portion of the storage devices serviced by the selected object. Configuration management operations are performed on the portion of the storage devices serviced by the selected object. The freeze condition is removed from the selected object in response to completion of the configuration management, in order to resume I/O operations to the portion of the storage devices serviced by the selected object.

Technique for coherent suspension of I/O operations in a RAID subsystem

A method and system for rapidly recovering data from a failed disk in a RAID disk group are disclosed. According to one aspect of the present invention, a RAID-based storage system identifies a particular disk in a RAID disk group as a 'dead' disk (e.g., incapable of servicing client-initiated requests in a timely manner). Accordingly, a spare disk is allocated to replace the 'dead' disk and client-initiated read/write requests are directed to the spare disk for servicing. In addition, a disk-to-disk copy operation is initiated. Without overwriting valid data on the target disk with stale data from the &#x201c;dead&#x201d; disk, the disk-to-disk copy operation copies data from the &#x201c;dead&#x201d; disk to the target by directly reading data from the &#x201c;dead&#x201d; disk while reconstructing only the data that cannot be read directly from the &#x201c;dead&#x201d; disk.

Method and system for rapidly recovering data from a 'dead' disk in a RAID disk group

The present invention implements an I/O task architecture in which an I/O task requested by the storage manager, for example a stripe write, is decomposed into a number of lower-level asynchronous I/O tasks that can be scheduled independently. Resources needed by these lower-level I/O tasks are dynamically assigned, on an as-needed basis, to balance the load and use resources efficiently, achieving higher scalability. A hierarchical order is assigned to the I/O tasks to ensure that there is a forward progression of the higher-level I/O task and to ensure that resources do not become deadlocked.

Method and apparatus for decomposing I/O tasks in a raid system

Method and apparatus for runtime resource deadlock avoidance in a raid system

A storage system with redundant mass storage device arrays determines that a mass storage device is non-responsive, and individually offlines the specific mass storage device for a temporary period. Reads to the offline mass storage device may be temporarily prevented from reaching the mass storage device, and may be serviced by a redundant array. Write to the offline mass storage device may also be temporarily prevented from reaching the mass storage device, and may be serviced with a temporary storage management that tracks and/or stores write transactions. The mass storage device can be synchronized with the write transactions on the basis of the tracking.

Selective disk offlining

A technique coherently suspends input/output (I/O) operations in a RAID subsystem of a storage system. A configuration tree of the RAID subsystem has a plurality of objects representing a logical configuration of storage devices coupled to the system. According to the technique, a 'freeze' condition may be imposed on an object of the configuration tree to suspend I/O operations directed to that object. In order to freeze, I/O operations underway ('in flight') in the RAID subsystem and directed to the object need to complete sufficiently so as to reach a recoverable state in the event the subsystem subsequently fails prior to an I/O restart procedure. Once a freeze condition has been imposed, new I/O requests directed to the object are inserted onto a freeze list of pending requests at the RAID subsystem and are blocked from processing until the object is &#x201c;unfrozen&#x201d; (i.e., the freeze condition is lifted).

Growing community of inventors

Hillsborough, CA, United States of America

James Leong