Distributed Data Storage Optimization in Healthcare Cloud Systems

The rapid growth of healthcare-related data and research prompts investments in data storage infrastructure.Cloud systems offer on-demand storage and cost savings through resource pooling; however, service outages may prevent access to stored data, especially during clinical emergencies.Data must comply with privacy regulations, requiring patients' consent for cross-region data transfer routing.Moreover, data distribution across multiple geographic regions delays access.A proposed system architecture employs location-and tier-agnostic data generators, along with ingestion pipelines that validate and normalize data for storage.Healthcare data can be contracted to clouds run by third-party service providers.Although such systems reduce costs through resource pooling, they incur additional overhead for wide-area data accesses.The main cloud storing and managing the data must guarantee permanent access for the data owner, incorporating adequate penalties in service-level agreements (SLAs).Provided that the main cloud site is operational, sensitive data are stored there.Other clinical data can be accessed from replicas kept in other regions under a different privacy regime.Availability and consistency requirements therefore do not always match in low-cost cross-region accesses.These trade-offs must be managed appropriately.