Controlling Dataverse Storage Costs for Regulated Records with Long Term Retention and Elastic Tables

Model-driven applications built on Microsoft Dataverse tend to accumulate data quietly and continuously. Every case note, assessment, approval and audit entry adds rows to tables that were never designed to be pruned, and in regulated environments those rows cannot simply be deleted once they lose operational value. For aged care providers and government agencies, the result is a growing tension between record keeping duties and the cost of holding data in a live transactional store.

This article examines how Dataverse long term retention and elastic tables can help organisations manage that tension. The aim is to preserve records for as long as compliance obligations require while moving inactive data to lower cost storage that remains queryable for audit and investigation. Understanding how Dataverse meters storage, and where each capability fits, allows technical and decision making readers to plan capacity deliberately rather than react to escalating bills.

How Dataverse meters and charges for storage

Dataverse separates capacity into distinct types rather than charging for a single undifferentiated pool. Database storage covers the structured, relational rows that most tables use, file storage holds attachments and larger binary content, and log storage retains auditing records. Each type is priced differently, and the transactional database tier is the most expensive because it is optimised for fast reads and writes rather than economical long term holding.

In practice, the fastest growing contributors in regulated deployments are auditing logs and high frequency operational tables. A government agency capturing every change to a client record, or an aged care provider logging care interactions across thousands of residents, can find that historical data it must keep for years sits in the same premium tier as data staff query daily. Recognising this pattern early is important, because storage consumption compounds over time and retrofitting a strategy onto a mature environment is harder than designing one from the outset.

Long term retention for records you must keep but rarely touch

Dataverse long term retention addresses the common situation where records carry a legal obligation to be retained but no longer support day to day work. Administrators define a retention policy against a table, using conditions that identify which rows qualify, and Dataverse moves the matching data from the transactional database into managed, lower cost retention storage. The data remains within the Dataverse governance boundary, so it continues to inherit the platform's security and compliance controls rather than being exported to a loosely managed archive.

Retained data is held in a read only state and can be queried without restoring it to the live tables, which suits audit, discovery and investigation scenarios well. Analysts can reach the retained records through tools such as Microsoft Fabric and Azure Synapse Link, allowing reporting and legal hold requirements to be met without inflating the operational store. Because the retention decision is policy driven, an organisation can encode its records disposal schedule directly into the platform, retaining data for the mandated period and applying deletion only when retention obligations genuinely lapse.

Elastic tables for high volume operational data

Where the challenge is sheer volume and unpredictable throughput rather than long term holding, elastic tables offer a different approach. Elastic tables are backed by Azure Cosmos DB rather than the standard relational store, which allows them to scale horizontally and handle very large row counts and bursty write patterns. This design suits telemetry, sensor readings, high frequency device data and other workloads that would strain a conventional Dataverse table.

For aged care and government scenarios, elastic tables become relevant as connected devices and continuous monitoring generate data at a scale that traditional tables handle inefficiently. Adopting them involves trade offs, because they support a subset of relational features and behave differently from standard tables in areas such as transactions and certain query patterns. The practical guidance is to reserve elastic tables for genuinely high volume, semi structured workloads, and to pair them with a retention plan so that even fast growing operational data does not remain in an active tier longer than necessary.

Aligning retention with Australian record keeping duties

Storage tiering is only defensible when it maps to real obligations, and in the sectors this article addresses those obligations are specific. Australian Government agencies operate under record keeping frameworks overseen by the National Archives of Australia, including authorised retention and disposal schedules that dictate how long particular classes of record must be kept. Aged care providers face their own duties to retain care and clinical records, and neither sector can justify deleting data simply to reduce a storage bill.

A sound approach treats the retention policy inside Dataverse as an implementation of the organisation's approved disposal authority rather than an independent decision. That means documenting which tables hold records of enduring value, mapping each to its lawful retention period, and configuring policies so that data moves to retention storage during its dormant years and is disposed of only when authorised. It also means preserving auditability throughout, so that the movement of records between tiers is itself traceable and the organisation can demonstrate to regulators that access controls and integrity have been maintained across the record's full lifecycle.

Planning a deliberate storage strategy

Bringing these capabilities together starts with visibility into where storage is actually being consumed, because assumptions about which tables dominate are often wrong. Reviewing capacity reports, identifying the tables and logs that grow fastest, and classifying that data against retention obligations gives a factual basis for decisions. From there, retention policies and elastic tables can be applied where they deliver the most benefit, rather than uniformly across an environment.

The outcome that regulated organisations should seek is predictable cost alongside demonstrable compliance. By keeping active data in the transactional tier, moving dormant but legally required records to retention storage, and directing high volume workloads to elastic tables, an organisation aligns its architecture with both its budget and its statutory duties. This is an ongoing discipline rather than a one off exercise, and revisiting the strategy as data volumes and obligations evolve keeps the environment sustainable over the long term.

References

• Microsoft Learn: Overview of long term data retention in Dataverse
• Microsoft Learn: Elastic tables in Dataverse
• Microsoft Learn: Manage storage capacity in Power Platform
• National Archives of Australia: Records authorities