About backup anomaly detection
NetBackup can now detect anomalies in backup metadata. It can detect any unusual job data in the data backup flow. For example, it can detect a file count or a file size that is different than the usual count or size.
Note:
By default, the anomaly detection algorithm runs on the NetBackup primary server. If you see any impact on the primary server because of the anomaly detection process, you can configure a media server to detect anomalies.
The following backup job metadata, attributes, or features are verified during backup anomaly detection:
Backup image size
Number of backup files
Data that is transferred in KB
Deduplication rate
Backup job completion time
Any unusual deviation in these backup job attributes is considered to be an anomaly and is notified using the NetBackup web UI.
Starting with 10.4, NetBackup can detect anomalies for Oracle workload only for the image size attribute. If an Oracle workload job fails multiple times with status code 5407, NetBackup flags it as an anomaly.
The workflow of the backup anomaly detection and notification is as follows:
Table: Workflow
Step | Description |
|---|---|
Step 1 | Install or upgrade NetBackup software on the primary server and the media server. |
Step 2 | Enable the primary server to detect backup anomalies. See Detecting backup anomalies on the primary server. By default, the anomaly detection algorithm runs on the NetBackup primary server. If you see any impact on the primary server because of the anomaly detection process, you can configure a media server to detect anomalies. |
Step 3 | Configure anomaly detection settings using the NetBackup web UI. |
Step 4 | View the anomalies using the NetBackup web UI. |