Breadcrumbs

Hashed File Synchronisation

Query name

Hashed File Synchronisation

Server Asset Type

Hashed File stage

Analysis Type

Hashed Files

Description

Identifies jobs featuring Hashed Files that do not read and write from the same stage.

Issue

Using a Hashed file in the middle of a job flow serves as a point of synchronisation: the write operation on the inbound link completes before the read operation on the outbound link.

Where the read occurs from a different stage from the write, the conversion of the overall job design to parallel may result in a design where a synchronisation issue occurs.

This potentially make them unsuitable for conversion to a dataset, as Hashed Files support behaviours and usage patterns that datasets cannot replicate.

Actions

This helps customers…

  • Placeholder


See also