Learn about the ONTAP ARP/AI evaluation period for block-device workloads
Beginning with ONTAP 9.17.1, ARP/AI requires an evaluation period to determine if entropy levels for block-device workloads are suitable for ransomware protection. These workloads include SAN LUNs and hypervisor virtual disks (for example, VMware virtual disks in NFS datastores and, beginning with ONTAP 9.17.1P5, Hyper-V, KVM, and OpenStack virtual disks) stored in ONTAP volumes. After ARP is enabled on an eligible volume, ARP/AI actively monitors and protects the volume during the evaluation period while simultaneously determining an optimal encryption threshold. Detection and alerts can occur during the evaluation period using a conservative threshold while baseline thresholds are being established over the course of several days. ARP distinguishes between suitable and unsuitable workloads in the evaluated volume and, if the workloads are determined to be suitable for protection, automatically sets an encryption threshold based on evaluation period statistics.
Supported workloads and evaluation applicability
The block-device evaluation period applies in the following scenarios:
-
SAN volumes
-
LUN-based workloads presented as block devices to hosts or hypervisors.
-
-
NAS volumes that contain hypervisor virtual disks automatically detected by ONTAP
-
Supported hypervisors include VMware, Hyper-V, KVM, and OpenStack virtual disks stored in NFS or SMB datastores.
-
Within these volumes:
-
The evaluation period is applicable to attacks detected based on entropy changes inside the guest filesystem of the virtual disk (for example, ransomware operating on files within the guest OS mapped to a LUN or virtual disk).
-
The evaluation period is not applicable to attacks detected based on entropy and file-extension changes made directly to the virtual disk files from the hypervisor host (for example, ransomware operating directly on
.vmdkfiles from an ESXi NFS datastore mount point). These direct-to-disk attacks use a different detection path that does not rely on the block-device evaluation period.
Version support for block-device and hypervisor detection
-
ONTAP 9.17.1
-
Introduces the block-device evaluation period for SAN volumes.
-
Enables ARP/AI attack detection inside SAN LUNs and inside VMware virtual disks stored in ONTAP NFS datastores.
-
-
ONTAP 9.17.1P5 and later
-
Expands ARP/AI block-device detection to hypervisor virtual disks such as Hyper-V, KVM, and OpenStack.
-
Applies the same block-device evaluation logic and thresholds to these additional hypervisor workloads when they are detected by ONTAP.
-
Understand entropy evaluation
During the evaluation period, the system collects continuous encryption statistics in 10-minute intervals from supported block-device and hypervisor workloads. ARP periodic snapshots are also continuously created every four hours. If the encryption percentage within an interval exceeds the optimal encryption threshold identified for this volume, an alert is triggered, an Anti_ransomware_attack_backup snapshot is created, and snapshot retention time is increased on any periodic ARP snapshots.
You can confirm that the evaluation is active by running the following command and confirming a status of evaluation_period. If a volume is not eligible for evaluation, the evaluation status will not be displayed.
security anti-ransomware volume show -vserver <svm_name> -volume <volume_name>
Example response:
Vserver Name : vs1 Volume Name : v1 State : enabled Attack Probability : none Attack Timeline : - Number of Attacks : - Attack Detected By : - Block device detection status : evaluation_period
You can monitor encryption detection in real time by running the following command. The command returns a histogram showing the amount of data in each encryption percentage range. The histogram is updated every 10 minutes.
security anti-ransomware volume entropy-stat show-encryption-percentage-histogram -vserver <svm_name> -name <lun_name> -duration real_time
Example response:
Vserver Name Entropy Range Seen N Time Data Written ---------- ---------------- --------------- -------------- ------------- vs0 lun1 0-5% 4 100MB vs0 lun1 6-10% 10 900MB vs0 lun1 11-15% 20 40MB vs0 lun1 16-20% 10 70MB vs0 lun1 21-25% 60 450MB vs0 lun1 26-30% 4 100MB vs0 lun1 31-35% 10 900MB vs0 lun1 36-40% 20 40MB vs0 lun1 41-45% 0 0 vs0 lun1 46-50% 0 0 vs0 lun1 51-55% 0 0 vs0 lun1 56-60% 0 0 vs0 lun1 61-65% 0 0 vs0 lun1 66-70% 0 0 vs0 lun1 71-75% 0 0 vs0 lun1 76-80% 0 0 vs0 lun1 81-85% 0 0 vs0 lun1 86-90% 0 0 vs0 lun1 91-95% 0 0 vs0 lun1 96-100% 0 0 20 entries were displayed.
Suitable workloads and adaptive thresholds
The evaluation ends with one of the following results for both SAN LUN workloads and hypervisor virtual disks evaluated through block-device detection:
-
The workload is suitable for ARP. ARP automatically sets the adaptive threshold to higher than 10% of the maximum encryption percentage seen during the evaluation period. ARP also continues statistics collection and creates periodic ARP snapshots.
-
The workload is unsuitable for ARP. ARP automatically sets the adaptive threshold to the maximum encryption percentage seen during the evaluation period. ARP also continues statistics collection and creates periodic ARP snapshots, but the system ultimately recommends disabling ARP on the volume.
After the evaluation period ends, ARP automatically sets the adaptive threshold based on the evaluation results.
You can determine the evaluation results by running the following command. Volume suitability is indicated in the Block device detection status field:
security anti-ransomware volume show -vserver <svm_name> -volume <volume_name>
Example response:
Vserver Name : vs1 Volume Name : v1 State : enabled Attack Probability : none Attack Timeline : - Number of Attacks : - Attack Detected By : - Block device detection status : Active_suitable_workload Block device evaluation start time : 5/16/2025 01:49:01
You can also show the value threshold adopted as a result of the evaluation:
security anti-ransomware volume attack-detection-parameters show -vserver <svm_name> -volume <volume_name>
Example response:
Vserver Name : vs_1
Volume Name : vm_2
Block Device Auto Learned Encryption Threshold : 10
...