摘要:
Described herein is a flash remapping (FR) layer in a storage operating system for utilizing flash memory as a secondary permanent storage device in a storage system. The FR layer collects particular information (specified by collection parameters) of received access requests for data stored on primary storage devices of the storage system. Based on the collected information and a predetermined access pattern (specified by pattern parameters), the FR layer selects data sets on the primary storage devices to be transferred permanently to flash memory, whereby subsequent access requests to the selected data sets are redirected to flash memory. New parameters may be received by the FR layer (from a user or program) to dynamically reconfigure the functions of the FR layer. The FR layer may be implemented in the operating system without requiring other code of the storage operating system to be modified.
摘要:
The techniques introduced here provide for efficient management of storage resources in a modern, dynamic data center through the use of virtual storage appliances. Virtual storage appliances perform storage operations and execute in or as a virtual machine on a hypervisor. A storage management system monitors a storage system to determine whether the storage system is satisfying a service level objective for an application. The storage management system then manages (e.g., instantiates, shuts down, or reconfigures) a virtual storage appliance on a physical server. The virtual storage appliance uses resources of the physical server to meet the storage related needs of the application that the storage system cannot provide. This automatic and dynamic management of virtual storage appliances by the storage management system allows storage systems to quickly react to changing storage needs of applications without requiring expensive excess storage capacity.
摘要:
A network storage server system includes a distributed object store, a presentation layer, a metadata subsystem, and a content management subsystem. The object store has no namespace and provides location-independent addressing of data objects. The presentation layer provides multiple interfaces for accessing data stored in the object store, including a NAS interface and a Web Service interface, and provides at least one namespace for accessing data via the NAS interface or the Web Service interface. The Web Service interface allows access to stored data via the namespace or without using the namespace (“raw object” access). The metadata subsystem stores user-specified and/or system-generated metadata relating to data objects and allows data objects to be identified and retrieved by searching on the metadata. The content management subsystem autonomously manages lifecycles of data objects according to user-specified policies, based on metadata associated with the data objects and tracked by the metadata subsystem.
摘要:
A distributed object store in a network storage system uses location-independent global object identifiers (IDs) for stored data objects. The global object ID enables a data object to be seamlessly moved from one location to another without affecting clients of the storage system, i.e., “transparent migration”. The global object ID can be part of a multilevel object handle, which also can include a location ID indicating the specific location at which the data object is stored, and a policy ID identifying a set of data management policies associated with the data object. The policy ID may be associated with the data object by a client of the storage system, for example when the client creates the object, thus allowing “inline” policy management. An object location subsystem (OLS) can be used to locate an object when a client request does not contain a valid location ID for the object.
摘要:
Deduplication of data using a low-latency random read memory (LLRRM) is described herein. Upon receiving a block, if a matching block stored on a disk device is found, the received block is deduplicated by producing an index to the address location of the matching block. In some embodiments, a matching block having a predetermined threshold number of associated indexes that reference the matching block is transferred to LLRRM, the threshold number being one or greater. Associated indexes may be modified to reflect the new address location in LLRRM. Deduplication may be performed using a mapping mechanism containing mappings of deduplicated blocks to matching blocks, the mappings being used for performing read requests. Deduplication described herein may reduce read latency as LLRRM has relatively low latency in performing random read requests relative to disk devices.
摘要:
Deduplication of data using a low-latency random read memory (LLRRM) is described herein. Upon receiving a block, if a matching block stored on a disk device is found, the received block is deduplicated by producing an index to the address location of the matching block. In some embodiments, a matching block having a predetermined threshold number of associated indexes that reference the matching block is transferred to LLRRM, the threshold number being one or greater. Associated indexes may be modified to reflect the new address location in LLRRM. Deduplication may be performed using a mapping mechanism containing mappings of deduplicated blocks to matching blocks, the mappings being used for performing read requests. Deduplication described herein may reduce read latency as LLRRM has relatively low latency in performing random read requests relative to disk devices.
摘要:
Deduplication of data using a low-latency random read memory (LLRRM) is described herein. Upon receiving a block, if a matching block stored on a disk device is found, the received block is deduplicated by producing an index to the address location of the matching block. In some embodiments, a matching block having a predetermined threshold number of associated indexes that reference the matching block is transferred to LLRRM, the threshold number being one or greater. Associated indexes may be modified to reflect the new address location in LLRRM. Deduplication may be performed using a mapping mechanism containing mappings of deduplicated blocks to matching blocks, the mappings being used for performing read requests. Deduplication described herein may reduce read latency as LLRRM has relatively low latency in performing random read requests relative to disk devices.
摘要:
Deduplication of data on disk devices based on a threshold number (THN) of sequential blocks is described herein, the threshold number being two or greater. Deduplication may be performed when a series of THN or more received blocks (THN series) match a sequence of THN or more stored blocks (THN sequence), whereby a sequence comprises blocks stored on the same track of a disk device. Deduplication may be performed using a block-comparison mechanism comprising metadata entries of stored blocks and a mapping mechanism containing mappings of deduplicated blocks to their matching blocks. The mapping mechanism may be used to perform later read requests received for the deduplicated blocks. The deduplication described herein may reduce the read latency as the number of seeks between tracks may be reduced. Also, when a seek to a different track is performed, the seek time cost is spread over THN or more blocks.
摘要:
Deduplication of data on disk devices based on a threshold number (THN) of sequential blocks is described herein, the threshold number being two or greater. Deduplication may be performed when a series of THN or more received blocks (THN series) match a sequence of THN or more stored blocks (THN sequence), whereby a sequence comprises blocks stored on the same track of a disk device. Deduplication may be performed using a block-comparison mechanism comprising metadata entries of stored blocks and a mapping mechanism containing mappings of deduplicated blocks to their matching blocks. The mapping mechanism may be used to perform later read requests received for the deduplicated blocks. The deduplication described herein may reduce the read latency as the number of seeks between tracks may be reduced. Also, when a seek to a different track is performed, the seek time cost is spread over THN or more blocks.
摘要:
Methods and system for securely capturing workloads at a live network for replaying at a test network. The disclosed system captures file system states and workloads of a live server at the live network. In one embodiment the captured data is anonymized to protect confidentiality of the data. A file system of a test server at the test network is mirrored from a captured state of the live server. An anonymized version of the captured workloads is replayed as a request to the test server. A lost or incomplete command is recreated from the states of the live server. An order of the commands during replay can be based on an order in the captured workload, or based on a causal relationship. Performance characteristics of the live network are determined based on the response to the replayed command.