Netapp, Inc.

Hewlett-packard Development Company, L.p.

Vijay Srinath

Subramaniyam V Pooni

Vikram Krishnamurthy

Rajkumar Mangalore

James Namboorikandathil Joseph

Yuedong Mu

Michael Robin Eisler

Ajith Balachandran

Rahul Pradeep

Somesh Kumar

Mohit Devarakonda

Ramakrishna Rao Yadala

Krishna Murthy Chandraiah Setty Narasingarayanapeta

Rakesh Bhargava M R

John Keith Fullbright

Shilpa Kumar

Namboorikandathil Joseph James

A method performed by one or more processing resources of one or more computer systems is disclosed. The method comprises receiving an object at a first of a plurality of nodes from a second of the plurality of storage nodes within a cluster switch fabric, examining a value associated included within the received object, wherein the value is associated with a clock value of the second node and updating a clock operating at the first node with the received value.

Mechanism to maintain data compliance within a distributed file system

One or more techniques and/or computing devices are provided for restarting a dump backup. For example, a phase offset table is used to track a directory transfer phase offset and a file transfer phase offset of phases of a dump backup. An offset map is used to map inodes, of data being backed up, to offsets within a backup image within which the data is being backed up. The phase offset table and/or the offset map are evaluated using a bytes written value to identify a phase during which the dump backup aborted and to identify a restart point. Accordingly, the dump backup is restarted at the restart point. In this way, the dump backup may be restarted during any phase and/or at any point using the bytes written value, the phase offset table, and the offset map without the need for additional information such as a file history.

Granular dump backup restart

Techniques and components for backing up data are disclosed. A first subset of data units is selected from a plurality of data units of a storage system to create a first partial baseline for backup of the first storage system. The number of data units in the first subset of data units is determined based on a window size parameter n. Data within the storage system that has changed since a previous backup operation are identified. A first backup including the first partial baseline and the first changed data is written to a second storage system. For a data loss event, the techniques include selecting, based on a time of the determined data loss event in combination with n, the first backup and n&#x2212;1 additional backups, wherein each of the n&#x2212;1 additional backups include a partial baseline and changed data. The first backup and the n&#x2212;1 additional backups are written to a third storage system.

Data backup with rolling baselines

A backup tool can manage multi-level backup into a cloud and restoration from the cloud. The backup tool can request a data source to stream backup data to the backup tool, and the backup tool can then generate data objects from the data stream for storing into the cloud. The backup tool generates the data objects in accordance with serialization of the data stream. The order of the data objects resulting from the data stream serialization is encoded into the names of the data objects. In addition, the backup tool encodes the backup level into the object names. With sequencing and backup level encoded into the data object names, the data objects can be stored in the cloud for later restoration.

Managing multi-level backups into the cloud

First partial baseline data of a first storage system is identified. First changed data of the first storage system is identified. The first changed data comprises data that has changed since a previous point in time. First backup data is written to a second storage system. The first backup data comprises the first partial baseline data and the first changed data. After writing the first backup data to the second storage system, second partial baseline data of the first storage system is identified. The second partial baseline data does not include the first partial baseline data. Second changed data of the first storage system is identified. The second changed data comprises data that has changed since writing the first backup data. Second backup data is written to the second storage system. The second backup data comprises the second partial baseline data and the second changed data.

The method and apparatus is utilized in order to generate a persistent path to a SCSI device for a host. In an embodiment, a SCSI device is queried for path information related thereto, and if path information is returned, a SCSI command requesting identifier data is issued to the SCSI device. The identifier data is used to determine a unique identifier (UID), from which is generated a UID-based device file for the SCSI device that is independent from the path information.

Method and apparatus for generating persistent path identifiers

A method, by which a user mode application obtains all physical paths that point to a logical unit on a newly-discovered small computer system interface (SCSI) device, may include: sending an input output control command to a kernel component regarding future discovery of any SCSI device; awaiting discovery of a new SCSI device; awaiting generation, by the kernel component, of a unique identifier (UID) and at least one set of physical path information mapping thereto for a logical unit on the newly-discovered SCSI device; and receiving, from the kernel component, the UID and the at least one set of physical path information mapping thereto for the logical unit of the newly-discovered SCSI device.

Method and apparatus for identifying multiple paths to discovered SCSI devices and specific to set of physical path information

The method includes querying for one or more logical unit numbers (LUNs) pertaining to a small computer system interface device, each LUN representing a potential path from a host to the SCSI device. Response data indicative of multiple LUNs to the single SCSI device is treated as separate instances of independent SCSI devices, with each separate instance representing a different SCSI separate instances of independent SCSI devices, with each separate instance representing a different SCSI device structure. A unique identifier (UID) is calculated for each SCSI device structure, from which a device file is generated based on the UID and contains UID and path information that differentiates between multiple paths from the host to the SCSI device.

Method and apparatus for identifying multiple paths to a SCSI device using a calculated unique identifier

A method and arrangement, for use in a system having a host and one or more small computer system interface (SCSI) devices, are described for enabling a user application accessing the system to communicate with one or more of the SCSI devices. The method and arrangement serve to overcome a device node limitation observed in standard Linux, which limits the number of devices a user application may communicate with to 128 SCSI disk (sd) devices or 256 SCSI generic (sg) devices. The method and arrangement provide a pass through capability by allowing the user application to directly talk to any SCSI device by using virtual handles, thereby overcoming the device node limitation observed in standard Linux.

Growing community of inventors

Bengaluru, India

Vijay Srinath