Netapp, Inc.

Network Appliance, Inc.

Panzura, Inc.

Brian Christopher Parkison

Susan M Coatney

David Brittain Bolen

Jeffrey S Kimmel

John Richard Taylor

John T Lloyd

Andrew P Davis

Loellyn Jane Cassell

Alan Stuart Driscoll

Stephen Wu

Rithin Kumar Shetty

Randy Yen-pang Chou

Vikas Yadav

David W Minnich

The disclosed embodiments disclose techniques for executing a cloud command for a distributed filesystem. Two or more cloud controllers collectively manage distributed filesystem data that is stored in one or more cloud storage systems; the cloud controllers ensure data consistency for the stored data, and each cloud controller caches portions of the distributed filesystem. During operation, a cloud controller presents a distributed-filesystem-specific capability to a client system as a file in the distributed filesystem (e.g., using a file abstraction). Upon receiving a request from the client system to access and/or operate upon this file, the client controller executes an associated cloud command. More specifically, the cloud controller initiates a specially-defined operation that accesses additional functionality for the distributed filesystem that exceeds the scope of individual reads and writes to a typical data file.

Executing a cloud command for a distributed filesystem

The disclosed embodiments disclose techniques for restoring an archived file in a distributed filesystem. Two or more cloud controllers collectively manage distributed filesystem data that is stored in one or more cloud storage systems; the cloud controllers ensure data consistency for the stored data, and each cloud controller caches portions of the distributed filesystem. Furthermore, cloud controllers may archive infrequently-accessed files in an archival cloud storage system. During operation, a cloud controller receives a request from a client system to access an archived file, and restores this archived file from the archival cloud storage system.

Restoring an archived file in a distributed filesystem

The disclosed embodiments disclose techniques for synchronizing file updates between two cloud controllers of a distributed filesystem. Two or more cloud controllers collectively manage distributed filesystem data that is stored in the cloud storage systems; the cloud controllers ensure data consistency for the stored data, and each cloud controller caches portions of the distributed filesystem. During operation, a cloud controller receives a request from a client to access a file in the distributed filesystem. The cloud controller sends a synchronization update request for the file to a second cloud controller and in response receives a synchronization update for the file from the second cloud controller.

Synchronizing file updates between two cloud controllers of a distributed filesystem

Lun clones are transferred from a primary server to a secondary server. The lun clones on the secondary server maintain the same data storage space saving relationship with backing luns of the lun clones as exists on the primary server. Incremental backups and restores of the lun clones between the primary and secondary servers involves transferring less than the total number of data blocks associated with the lun clone.

System and method for transferring and backing up luns and lun clones on primary and secondary servers

A system and method for allowing more rapid takeover of a failed filer by a clustered takeover partner filer in the presence of a coredump procedure (e.g. a transfer of the failed filer's working memory) is provided. To save time, the coredump is allowed to occur contemporaneously with the takeover of the failed filer's regular, active file service disks by the partner so that the takeover need not await completion of the coredump to begin. This is accomplished, briefly stated, by the following techniques. The coredump is written to a single disk that is not involved in regular file service, so that takeover of regular file services can proceed without interference from coredump. A reliable means for both filers in a cluster to identify the coredump disk is provided, which removes takeover dependence upon unreliable communications mechanisms. A means is provided for identification during takeover of the coredump disk and SCSI-3 reservations are employed to coordinate write access to (ownership of) shared disks, so as to prevent takeover from interfering with coredump while simultaneously preventing the failed filer is from continuing to write regular file system disks being taken over by its partner.

System and method for takeover of partner resources in conjunction with coredump

This invention provides a system and method for selecting and communicating a single disk (a 'coredump disk') for use in a coredump procedure by a failed file server (or filer). A selection method on the failed filer determines the 'best candidate' coredump disk according to a predetermined set of criteria. For example, the available disks that can receive coredump data are located and ordered so as to prefer disks that best match the coredump data size requirement, are least likely to be needed for normal service by the server; and require the least preparation to receive coredump data. Appropriate attributes on the selected coredump disk are written to indicate that a coredump is in progress, and the location of the coredump data. Upon reboot of the failed filer (or takeover by a cluster partner), the coredump disk is identified and the coredump data recovered by reading back appropriate attributes.

System and method of selection and communication of a disk for storage of a coredump

System and method for takeover of partner re-sources in conjunction with coredump

A system and method for allowing more rapid takeover of a failed filer by a clustered takeover partner filer in the presence of a coredump procedure (e.g. a transfer of the failed filer's working memory) is provided. To save time, the coredump is allowed to occur contemporaneously with the takeover of the failed filer's regular, active file service disks by the partner so that the takeover need not await completion of the coredump to begin. This is accomplished, briefly stated, by the following techniques. The coredump is written to a single disk that is not involved in regular file service, so that takeover of regular file services can proceed without interference from coredump. A reliable means for both filers in a cluster to identify the coredump disk is provided, which removes takeover dependence upon unreliable communications mechanisms. A means is provided for identification during takeover of the coredump disk and SCSI-3 reservations are employed to coordinate write access to (ownership of) shared disks, so as to prevent takeover from interfering with coredump while simultaneously preventing the failed filer from continuing to write regular file system disks being taken over by its partner.

Growing community of inventors

Santa Cruz, CA, United States of America

Brian Christopher Parkison