Across established ZDLRA environments, fine-grained configuration parameters and operational behaviors are frequently under-optimized, leading to measurable impacts on performance, recoverability SLAs, and security posture.
This article examines 10 commonly overlooked yet high-impact practices that can materially improve system efficiency and resilience.

  1. Not configuring autotuned_reserved_space during database onboarding when future capacity growth is challenging to forecast?
    The main confusion usually comes from the concept of “reserved space,” the minimum space assigned by the administrator for a specific protected database and can be manually increased or decreased as needed based on space needed to maintain recovery window settings. In release 21.1, a new feature was added that allows ZDLRA to automatically handle reserved space on your behalf.
    This Autotuned_reserved_space parameter is set at the protection policy level and can be either YES or NO (default is NO). When set to YES, ZDLRA automatically manages reserved space for all databases that belong to the policy and do not have disk_reserved_space explicitly set.This relieves administrator effort to adjust reserved space based on backup space needs – which can increase and decrease over time depending on database activities, to meet recovery window settings for each protected database.
  2. RMAN script includes a value other than 1 for filesperset ?
    When backing up to ZDLRA, the recommendation for filesperset for datafile backup is 1. This ensures 1 file per backupset which aligns with the virtual full backup design, leading to faster virtual Level 0 creation as soon as the datafile is received.
  3. Still leveraging complex firewall configuration to manage replication network traffic between production and cyber vault ZDLRA?
    Virtual Air Gap reduces operational complexity by putting air-gap control directly in ZDLRA, minimizing dependence on networking and other IT teams. With new RACLI commands, administrators can quickly open, close, or specify time-bound replication between the production (upstream) and vault (downstream) ZDLRA systems. This simplifies implementation and day-to-day management, enables predictable backup sync windows, improves efficiency, and strengthens security by keeping replication access tightly controlled.

    More details at: https://blogs.oracle.com/maa/introducing-zdlra-virtual-air-gap-v-gap
  4. Not copying ZDLRA catalog export to an external share? It’s easy.
    Every two hours alternating between compute nodes 1 and 2, a background process on the system exports the necessary EM agent, OSB catalog, and Recovery Appliance catalog information (e.g. replicated/tape/cloud backup metadata, protected database/VPC users, protection policies, etc.) to a ZDLRA export bundle file on the local compute node.

    In the event of a complete system failure, this export bundle file is used to rebuild the Recovery Appliance catalog, including users and policies as of the time the backup was taken.

    At the end of the Recovery Appliance DR/rebuild process:
    • Protected databases should be able to restart their backups with initial L0s as soon as the system is back online, with no changes to database-side configuration.
    • Protected databases should have access to Recovery Appliance backups (including tape, cloud, and replicated copies) as soon as the system is back online
    • Following command line can be used to setup the nfs path for export.
      racli configure export –nfs_path=/ranfs/smaug/ra_export
      Options
      The NFS must be added using “racli add nfs” prior to running this command.
      RA export bundle will be exported to this path in addition to the default location.
      –nfs_path  <Specifies the path to the NFS share>
  5. Not using LACP to leverage maximum network bandwidth for backup, restore, and replication? This could impact your RTO and replication performance.
    Using LACP for ZDLRA backups(supported with 10/25/100G network cards), aggregates multiple links to boost overall throughput, balance load across RMAN channels and provide seamless resilience with automatic interface failover—all while simplifying operations via a single logical interface and enabling non-disruptive maintenance.
  6. Wondering how ZDLRA plays a role in a cyber vault environment?
    A ZDLRA in a network-isolated vault location maintains operational database backups that can be quickly restored into a separate clean room to support minimum viable business services (database and application infrastructure). Synchronization of vault backups with production is accelerated due to incremental-forever replication model instead of replicating full backups, while still providing full point-in-time recovery. As a result, the vault network only needs to be opened for short time intervals to complete synchronization, unlike generic backup solutions that require longer windows to replicate scheduled full backups.
  7. Are you taking a new Level 0 backup after key rotation on the protected database?
    Space efficient backup backups work with RMAN to maintain the virtual backup lineage without requiring new full backups, when moving from non-TDE to TDE-enabled databases or performing TDE master key rotations. The incremental-forever approach continues to create virtual full backups on the Recovery Appliance, eliminating the need for post-rotation full backups
  8. Struggling with ZDLRA key metrics on state of health?
    The System Activity Report (SAR) captures protection health through appliance status, number of protected databases, backup success rate, redo ingest latency, and adherence to recovery windows/point-in-time recovery. It summarizes replication and air-gap posture via incremental-forever backlog, average sync window, last successful sync, and any notable exceptions. Capacity and performance (utilization, headroom, ingest/restore throughput, dedupe) are paired with key incidents, and prioritized recommendations.
    The System Activity Report (SAR) is available as :
    • Command line option using RACLI
      >racli run diagnostics  –tag=sar
      See more details In this MoS note: Zero Data Loss Recovery Appliance System Activity Script -KB514522
    • Enterprise Manager ZDLRA Reports
  9. Want to know more about managing ZDLRA through Oracle Enterprise Manager 24ai ?
    Oracle Enterprise Manager 24ai provides a unified platform to manage and monitor ZDLRA environments. With the ZDLRA plug-in, users gain visibility into appliance health, backup activity, storage utilization, and deployment topology, along with alerts, performance metrics, and redo transport monitoring.
    Building on this, the ZDLRA Management pack provides centralized backup management capabilities including onboarding of protected databases, configuration of backup policies, and scheduling of backup and archival operations. It supports fleet-level management across large database groups with both UI-driven and EM CLI-based workflows. Additional capabilities include visibility into recoverability status, SLA-related metrics, and reporting for backup operations and protection status
  10. Not leveraging Live updates for CVE patching? Keep up with CVE fixes without any downtime.
    Exadata Live Update enables the ZDLRA to address Operating Systems based CVSS by upgrading the security components in the software without a complete Exadata Image Update, which eliminates the reboot of the hosts in the cluster.
    More Details In this MoS note: Zero Data Loss Recovery Appliance Software Updates Guide – KB212582

And one more innovation : How can I leverage ZDLRA capabilities in the cloud?

If you are looking to extend ZDLRA capabilities into the cloud, consider the Zero Data Loss Autonomous Recovery Service for a fully managed, cloud-based equivalent of on-premises ZDLRA functionality. 

For organizations not yet ready to migrate databases to OCI, Zero Data Loss Cloud Protect provides a seamless option to safeguard on-premises databases by leveraging Autonomous Recovery Service  as a secure data protection/backup target, enabling cloud-based protection without requiring relocation of databases to the cloud.

Zero Data Loss Recovery Appliance continues to evolve with feature advancements that streamline operations and improve overall efficiency across backup and recovery. Systematically addressing these overlooked practices with leveraging latest innovations improves recovery objectives and strengthens resilience against ransomware threats. As a whole, these targeted optimizations can drive higher throughput, predictable recovery outcomes, and improved operational robustness.