Browse Prior Art Database

Applying an Operation to a Set of Slices Matching a Name Pattern

IP.com Disclosure Number: IPCOM000249434D
Publication Date: 2017-Feb-27
Document File: 2 page(s) / 26K

Publishing Venue

The IP.com Prior Art Database

Abstract

Disclosed is a method to perform efficient data transformation via using a name pattern match for all data within a container or a set of containers. The approach uses idempotent functors to asynchronously transform the slice data so each functor applied can guarantee that the data will be consistent, without a specified order of functor application.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 52% of the total text.

1

Applying an Operation to a Set of Slices Matching a Name Pattern

A system that provides the ability for a user to make a single call to transform all or a subset of objects within a dispersed storage (ds) container can be extremely convenient and useful for end users of the dispersed storage system.

One example is a single call to the storage system to delete all objects within a container. Instead of having to manually call delete on every object within the container, the end user makes a single call, which is immediately consistent from the perspective of outside the ds unit, but internal to the ds unit it may be asynchronously processed by the ds units later. In the delete case, a ds processing unit receives a call via the public Application Programming Interface (API) to delete a specified container and all objects within it. The ds processing unit then sends a message to its associated ds memory units to apply a given transformation, such as delete, to slices matching the provided pattern when possible. In the event that an input/output (I/O) request was sent to the ds memory unit for a slice with a pending transformation, the system automatically applies it before returning the result to the requesting ds processing unit. Therefore, this transformation provides immediate consistency to the end user as soon as the initial request has returned, without waiting on the transformations to be applied on every matching slice.

This type of request has many applications such as re-encryption, delete, or applying a change in access control lists (ACLs).

The novel contribution is a method to perform efficient data transformation via using a name pattern match for all data within a container or a set of containers. The approach uses idempotent functors to asynchronously transform the slice data so each functor applied can guarantee that the data will be...