Abstract
This disclosure describes techniques for data storage management. A storage management service is utilized to manage requests for datasets from multiple devices and users across an enterprise as well as perform storage management across multiple tiers of storage devices and/or services. Features of the storage management service include interception of user requests for datasets, and making suitable namespace adjustments and mounting the datasets (files) such that a user can access the datasets with minimal changes to their workflows. Additionally, dataset usage is monitored to determine a data temperature for the dataset based on a frequency of use and likelihood of a future request for the dataset. Based on the determined data temperature, the dataset is stored at a suitable storage system taking into account respective costs associated with different types of storage devices within the storage system. Further, the datasets can be stored and migrated between various tiers of storage to provide dynamic optimization of storage.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
N.A., "Machine Learning Data Storage Management", Technical Disclosure Commons, (July 18, 2024)
https://www.tdcommons.org/dpubs_series/7211