Pipeline planning for low latency storage system
AssigneeNetApp, Inc.
InventorDavid Slik
Life of this patent
Estimating the term…
Abstract
A write request including payload data is received. The payload data of the write request is stored in a staging area of a storage manager. A transformation pipeline is determined based, at least in part, on an attribute of the write request. The transformation pipeline is queued for execution. Data fragments are generated based, at least in part, on the payload data and the transformation pipeline. The data fragments are transmitted to a plurality of enclosures.
Claims21
A method comprising:
- receiving a write request including payload data;
- storing the payload data of the write request in a staging area of a storage manager;
- determining a transformation pipeline based, at least in part, on an attribute of the write request;
- queuing the transformation pipeline for execution;
- generating data fragments based, at least in part, on the payload data and the transformation pipeline;
- and
- transmitting the data fragments to a plurality of enclosures.
A method of operating a storage manager, the method comprising:
- receiving a read request including a data object identifier associated with a requested data object;
- retrieving a transformation recipe associated with the data object identifier and a fragment layout associated with the data object identifier;
- generating a transformation pipeline based, at least in part, on the transformation recipe;
- retrieving data fragments from a plurality of data storage devices based, at least in part, on the fragment layout;
- and
- queuing the transformation pipeline for execution on the data fragments.
A non-transitory computer readable storage medium storing instructions for operating a storage manager, wherein the instructions comprise program code to:
- receive a write request including payload data;
- store the payload data of the write request in a staging area of the storage manager;
- determine a transformation pipeline based, at least in part, on an attribute of the write request;
- queue the transformation pipeline for execution;
- generating data fragments based, at least in part, on the payload data and the transformation pipeline;
- and
- transmit the data fragments to a plurality of enclosures.
Dependencies are inferred from the claim text.