| 摘要 |
Systems capable of transformation of logical data objects for storage and methods of operating thereof are provided. One method includes identifying among a plurality of requests addressed to the storage device two or more “write” requests addressed to the same logical data object, deriving data chunks corresponding to identified “write” requests and transforming the derived data chunks, grouping the transformed data chunks in accordance with the order the requests have been received and in accordance with a predefined criteria, generating a grouped “write” request to the storage device, and providing mapping in a manner facilitating one-to-one relationship between the data in the obtained data chunks and the data to be read from the transformed logical object. The method further includes obtaining an acknowledging response from the storage device, multiplying the obtained acknowledging response, and sending respective acknowledgements to each source that initiated each respective “write” request. |
| 主权项 |
1. A method for transformation of logical data objects for storage in a distributed computing environment, comprising:
transmitting, by a plurality of servers via a distributed storage network, a plurality of data chunks and a plurality of write requests to write the data chunks to a plurality of storage devices, each data chunk associated with a respective write request and each storage device comprising one or more logical data objects (LOs); intercepting, by a processor in a transformation system coupled to the plurality of servers via the distributed storage network such that the distributed storage network is located between the transformation system and the plurality of servers and coupled to the plurality of storage devices via the distributed storage network such that the distributed storage network is located between the transformation system and the plurality of storage devices, each of the plurality of data chunks and each of the plurality of write requests to write the data chunks to the plurality of storage devices prior to each of the plurality of data chunks and each of the plurality of write requests reaching the plurality of storage devices, wherein:
the interception is transparent to the plurality of storage devices and to the server, andeach data chunk is associated with a respective write request; transforming, by the processor, each data chunk; determining which server transmitted each transformed data chunk and each respective write request; grouping each transformed data chunk and each associated write request that was transmitted by the same server to generate a plurality of groups of transformed data chunks and respective write requests; determining if a target LO for storing a particular group of transformed data chunks in a respective storage device includes sufficient free space to accommodate the particular group of transformed data chunks; if the target LO includes sufficient free space to accommodate the particular group of transformed data chunks:
transmitting the particular group of transformed data chunks and their respective write requests to the target LO, andwriting the particular group of transformed data chunks to the target LO; and if the target LO includes insufficient free space to accommodate the particular group of transformed data chunks:
determining if the particular group of transformed data chunks meets a predefined criteria related to a size of the particular group of transformed data chunks,if the size of the particular group of transformed data chunks meets the predefined criteria:
allocating a new LO in the respective storage device,splitting the group of transformed data chunks and their associated respective write requests into a plurality of parts,transmitting a first part of the group of transformed data chunks and their associated respective write requests to the target LO,transmitting a second part of the group of transformed data chunks and their associated respective associated write requests to the new LO,writing the first part of the plurality of parts to the target LO, andwriting the second part of the plurality of parts to the new LO, andif the size of the particular group of transformed data chunks fails to meet the predefined criteria:
allocating the new LO,transmitting the group of transformed data chunks and their respective write requests to the new LO, andwriting the particular group of transformed data chunks to the new LO. |
