Home > Storage > PowerScale (Isilon) > Industry Solutions and Verticals > Media and Entertainment > PowerScale OneFS: A Metadata Driven Approach to On Demand Tiering > SmartPools job engine
OneFS has a job engine that runs on a schedule to enforce File Policies that migrate data between storage tiers. The SmartPools job must walk the entire OneFS file system and make determinations about what data needs to be moved. This job can be resource intensive. On a large busy cluster, the SmartPools job might take a long time to complete.
To meet this challenge, OneFS has developed several features to allow speedy completion of File Pool Policies. One of these features, SmartPoolsTree jobs, runs the File Pool Policies only against a specific path in the file system. SmartPoolsTree is one of the OneFS features that this white paper leverages for doing on-demand data tiering. Crucially, SmartPoolsTree jobs can be triggered using the OneFS REST API, as will be described later.
The Unstructured Data Storage Tips blog has an excellent entry on another OneFS feature, the FilePolicy job. This job can also be leveraged for speeding up the evaluation of File Pool Policy criteria in a OneFS file system. The File Pool Policy job relies on change lists generated by the IndexUpdate job. The OneFS job engine includes many operations to determine which files should be evaluated against File Pool Policy criteria. However, the FilePolicy job is not used in this white paper, it is only mentioned here for completeness.
Figure 3. SmartPoolsTree job