Time Machine is the correct feature for this requirement. While snapshots capture the state at a moment, a Time Machine provides continuous protection and maintains a log of changes, enabling precise Point-in-Time Recovery (PITR) for the entire database service (including its files) to any second within the retention window. This is essential for critical applications needing minimal RTO and precise recovery points.

When a clone’s performance differs from its source, storage configuration is a primary suspect. Era clones inherit the database configuration but are provisioned onto a target storage container defined in the Era blueprint. If this container has a lower Resilience Factor (e.g., RF2 instead of RF3) or is mapped to a lower-performance storage tier (e.g., All-Flash vs. Hybrid), it will directly impact I/O performance. This check precedes investigating compute or software versions.

The requirement calls for full automation of both protection (source snapshot) and provisioning (new clone). A Protection Policy automates the snapshot schedule on the source database. By adding a Clone Automation schedule to that policy, you can define a schedule (e.g., daily) and a target blueprint to automatically create a new clone from the latest snapshot, meeting the ‘ready by 6 AM’ requirement without manual steps.

Long-Term Retention (LTR) is the Era feature designed for this scenario. It allows you to send database backups (snapshots and logs) from the Time Machine to an immutable, write-once-read-many (WORM) compliant object store (like Nutanix Objects or AWS S3). This satisfies the requirements for separate storage, immutability, and extended retention periods far beyond what typical local snapshots provide.

‘Tracked’ mode is designed for databases on existing VMs where you want Era to provide protection (snapshots, clones) and insights, but you (or another tool) retain control over the server VM’s lifecycle (e.g., power state, sizing, patching). This is a key architectural choice, contrasting with ‘Provisioned’ mode where Era fully manages the underlying VM.

The most critical step before an Era update is to ensure no active Era operations are running. Initiating an update while a snapshot, clone, or refresh is in progress can lead to failed operations or database corruption. The Era UI typically warns about this, but explicitly checking and waiting for all operations to complete is a mandatory best practice for a smooth update.

Database Parameters (or ‘Database Properties’ in some contexts) within an Era blueprint are used to set specific database-level settings. For SQL Server, parameters like ‘Number of Log Files’ and ‘Initial Log File Size (MB)’ are configured here. The Software Profile defines the DBMS version, the Compute Profile defines CPU/Memory, and the Network Profile defines connectivity.

Creating a new clone with the ‘Thin Clone’ option is the best answer. Nutanix’s native snapshot and clone technology allows for the creation of a space-efficient, writable clone almost instantly, regardless of the source database size. It consumes minimal storage initially (only for metadata and changed blocks), meeting both the speed and efficiency requirements. Refreshing a clone is fast but requires one to already exist.

Era leverages the underlying Nutanix AOS storage services. Snapshot data is stored on the AHV storage container. Therefore, enabling encryption at the container level (via Prism Central storage policies) is the fundamental and correct way to ensure all Era snapshots (and clones) are encrypted at rest. This is transparent to the database engine and managed by the infrastructure.

Time Machine is the correct feature for this requirement. While snapshots capture the state at a moment, a Time Machine provides continuous protection and maintains a log of changes, enabling precise Point-in-Time Recovery (PITR) for the entire database service (including its files) to any second within the retention window. This is essential for critical applications needing minimal RTO and precise recovery points.0