In this blog post How Australian Businesses Should Prepare for Microsoft Fabric Recovery we will explain what Microsoft’s new default recovery setting means, where its protection stops, and what Australian technology leaders should do before relying on it.
At a high level, Microsoft Fabric brings data storage, reporting, analytics and artificial intelligence tools into one platform. Businesses use it to manage items such as lakehouses, data warehouses, notebooks, pipelines and real-time dashboards. If one of these items is accidentally deleted, the business impact can range from a broken report to a complete interruption of an operational process.
Item Recovery adds a safety net. Instead of immediately removing a supported item, Fabric uses “soft delete”, which temporarily places it in a workspace recycle bin. Authorised users can restore the item before the recovery period expires.
What is changing in Microsoft Fabric
Microsoft made Fabric Item Recovery generally available on 30 June 2026. From 23 August 2026, Microsoft plans to enable it automatically for tenants where an administrator has not already made an explicit choice.
These previously unconfigured tenants will receive a three-day recovery window for supported item types. If your organisation has already enabled, configured or explicitly disabled Item Recovery, Microsoft’s automatic change should not override that decision.
The three-day default is better than having no protection, but it should not be mistaken for a complete recovery policy. A long weekend, staff absence or delayed incident report could consume most of that window before the IT team even knows an item is missing.
Why this matters to mid-market organisations
Fabric environments often grow faster than their governance processes. A reporting project begins with two people, then expands to finance, operations, sales and external consultants. Before long, dozens of users can create, change or remove business-critical items.
Consider a 180-person logistics company using Fabric for daily delivery reporting. During a workspace clean-up on Friday afternoon, someone deletes a pipeline that loads tracking data overnight.
The problem may not become visible until Monday morning, when managers find their dashboards are showing Friday’s figures. With recovery enabled, the pipeline may be restored in minutes. Without it, the data team could spend hours rebuilding the pipeline, testing its connections and investigating whether any data was lost.
The value is not simply technical convenience. It is fewer business interruptions, less expensive rework and less time spent explaining missing reports to senior management.
Five steps to prepare for default Item Recovery
1. Check your tenant before Microsoft makes the decision for you
Do not assume your existing setting is known or appropriate. Ask your Fabric administrator to open the Fabric admin portal, review Tenant settings and locate Item Recovery.
Confirm whether the setting is enabled, disabled or still unconfigured. Record the result, the selected retention period and who approved it.
This simple review prevents an automatic platform default from quietly becoming your organisation’s recovery policy. It also gives the CIO or technology leader a clear answer when asked how deleted Fabric assets are protected.
2. Choose a recovery window based on business impact
The automatic three-day window is a baseline, not necessarily the right answer. Microsoft’s administration controls allow organisations that deliberately configure Item Recovery to select a longer retention period, currently between seven and 90 days.
Choose the period by considering how quickly deletion is likely to be detected. A failed daily finance process may be noticed within hours, while a monthly compliance report might not be opened for several weeks.
Longer retention also has a cost. Soft-deleted items continue to consume OneLake storage, Microsoft Fabric’s central data storage layer, at the normal storage rate. Small amounts of computing capacity may also be used for background maintenance.
For many mid-market organisations, 14 to 30 days will provide more practical breathing room than three days. The right number depends on the value and volume of your data, internal reporting cycles and privacy obligations.
3. Identify what is protected and what is not
Item Recovery only applies to supported Fabric item types. Coverage includes many commonly used assets, such as lakehouses, warehouses, notebooks, data pipelines and machine learning items, but Microsoft is continuing to expand support.
Create an inventory of your business-critical Fabric assets and confirm whether each item type can be recovered. Pay particular attention to custom integrations, recently introduced Fabric features and processes that rely on several connected items.
It is also important to distinguish between item recovery, workspace recovery and OneLake file recovery. They protect different layers and can have different retention periods.
A recycle bin for an individual pipeline does not automatically solve every failure involving its underlying files, credentials, external connections or dependent systems. Your recovery plan needs to cover the complete business process, not just the visible Fabric item.
4. Tighten permissions around recovery and permanent deletion
Workspace Contributors, Members and Administrators can restore recoverable items. Workspace Administrators can also permanently delete items from the recycle bin before the retention period expires.
That makes access control important. If too many people have administrator rights, a malicious or careless user may be able to delete an item and then remove the recoverable copy.
Review who has access to production workspaces and remove unnecessary administrative permissions. This supports the Essential Eight, the Australian government’s cybersecurity framework, which includes restricting powerful administrative access.
Where possible, separate day-to-day development from production administration. Developers may need to build and update items, but permanent deletion should be limited to a smaller group and handled through an approved change process.
The same identity and access principles discussed in our Microsoft 365 tenant hardening roadmap should be applied to Fabric administration.
5. Test recovery instead of assuming it works
A recovery feature only provides value if your team knows how to use it under pressure. Create a test item in a non-production workspace, delete it, locate it in the workspace recycle bin and restore it.
Check that its configuration, permissions, data relationships and dependent processes return as expected. Record how long recovery takes and who needs to approve the action.
Add these instructions to your incident response runbook. Larger Fabric environments should also consider using the Fabric REST API, which allows approved recovery tasks to be performed through automation rather than relying entirely on manual portal steps.
Testing also supports the Essential Eight’s regular backups strategy. Australian Signals Directorate guidance expects organisations to test restoration, not simply confirm that a backup or recovery feature exists.
Item Recovery is not a replacement for backup or disaster recovery
Soft delete is designed mainly for accidental removal. It does not provide complete protection against data corruption, unsupported item deletion, an expired retention window, compromised administrator accounts or a broader service and regional outage.
Critical Fabric solutions may also need version control for item definitions, separate copies of essential data, documented rebuild procedures and recovery arrangements outside the affected workspace. This is the same layered approach we recommend in our guide to high availability and disaster recovery for Microsoft Foundry agents.
Australian organisations should also consider privacy requirements. Australian Privacy Principle 11 requires covered organisations to take reasonable steps to protect personal information from loss and unauthorised access. However, retaining deleted data for longer than necessary can create its own privacy and governance concerns.
Your chosen recovery period should therefore involve IT, business owners, security and privacy stakeholders. It should not be selected solely because 90 days sounds safer.
A practical checklist for technology leaders
- Confirm whether Item Recovery is enabled, disabled or unconfigured.
- Decide whether three days is enough for your reporting and operational cycles.
- Inventory critical Fabric items and confirm recovery support.
- Review workspace roles and limit permanent deletion rights.
- Document who can approve and perform a recovery.
- Run a test deletion and restoration in a safe workspace.
- Monitor the storage cost of longer retention periods.
- Maintain separate backup, version control and disaster recovery measures.
Turn Microsoft’s safety net into a business control
Default Item Recovery is a welcome improvement. It can turn an accidental deletion from a serious operational incident into a short support task.
However, the default three-day window is only a starting point. The real business outcome comes from selecting an appropriate retention period, controlling administrator access, testing restoration and protecting the parts of your Fabric environment that Item Recovery does not cover.
CloudProInc is a Melbourne-based Microsoft Partner with more than 20 years of enterprise IT experience. If you are unsure whether your Fabric recovery settings match your business, Essential Eight or privacy requirements, we are happy to take a practical look at your current setup—no strings attached.
Discover more from CPI Consulting
Subscribe to get the latest posts sent to your email.