NS0-165

Free NS0-165 practice questions

10 questions from our NS0-165 practice tests, each with the correct answer and a written explanation. Read them with the answers shown, or switch to quiz mode and answer first.

Back to the NS0-165 practice exams

Question 1 of 10, Storage Platforms

A database team uses only iSCSI and FC, and wants every host path to every LUN to be active and optimized so that a controller failure does not require any path state changes. Which platform should be recommended?

  1. AAn ASA system, which uses symmetric active-active multipathing for SAN Correct answer
  2. BA FAS system using RAID-TEC aggregates with ALUA reporting disabled on igroups
  3. CAn ONTAP Select cluster with the LUNs placed on a FlexGroup volume across nodes
  4. DAn AFF A-Series system configured with Selective LUN Map on both HA nodes

Answer and explanation: ASA systems are block-optimized and present all paths as active-optimized through symmetric active-active multipathing. On unified AFF and FAS systems, ALUA marks partner paths as non-optimized, and Selective LUN Map only limits which nodes report paths. Disabling ALUA is not a supported way to make paths symmetric, and FlexGroup volumes do not host LUNs in this way.

Question 2 of 10, Core ONTAP

In a four-node ONTAP cluster, what is the purpose of epsilon?

  1. AIt reserves a portion of NVRAM on each node so that writes survive a power outage during a takeover
  2. BIt gives one node an extra fractional vote so that quorum can be decided when the cluster splits into two equal halves Correct answer
  3. CIt identifies the node that owns the SVM root volumes and the replicated cluster configuration database
  4. DIt marks the node that runs the cluster management LIF and the System Manager web service

Answer and explanation: Epsilon is a tie-breaking weighted vote held by one node so that, if an even-numbered cluster partitions evenly, one side can still hold quorum. NVRAM mirroring protects in-flight writes but has nothing to do with epsilon. The cluster management LIF can run on any node, and the replicated database is held by all healthy nodes rather than a single epsilon owner.

Question 3 of 10, Core ONTAP

A four-node cluster loses both nodes of one HA pair at the same moment. Epsilon is held by a node in the surviving HA pair. What happens to the cluster?

  1. AThe surviving nodes take over the failed pair's aggregates through storage failover across HA pairs
  2. BThe two surviving nodes keep quorum because epsilon breaks the tie, and they keep serving their own data Correct answer
  3. CThe whole cluster loses quorum and stops serving data until at least three of the four nodes are back online
  4. DThe surviving nodes form a new two-node cluster automatically and enable cluster HA on themselves

Answer and explanation: With two of four nodes healthy, epsilon gives the surviving side the extra weight needed to keep quorum, so its nodes continue serving their own data. Without epsilon on that side, quorum would be lost. Storage failover works only within an HA pair, so the failed pair's aggregates stay unavailable, and ONTAP does not automatically reform a new cluster.

Question 4 of 10, ONTAP Storage

A FabricPool aggregate is attached to an object store. The administrator wants both cold user data blocks and cold Snapshot blocks in volume archive1 to move to the cloud tier after the cooling period. Which tiering policy should be set?

  1. Aall
  2. Bsnapshot-only
  3. Cnone
  4. Dauto Correct answer

Answer and explanation: The auto policy tiers both cold Snapshot blocks and cold active file system blocks once they reach the cooling period. Snapshot-only, the default, tiers only blocks that are no longer referenced by the active file system. The all policy moves data to the cloud tier immediately without waiting for it to cool, and none keeps everything on the performance tier.

Question 5 of 10, Networking

A service provider hosts two tenants on one ONTAP cluster. Both tenants use the same private 10.10.0.0/16 address range and must be kept fully separate at the routing level. What should the administrator configure?

  1. ATwo subnets in the Default IPspace that both use the 10.10.0.0/16 range with different gateways
  2. BOne VLAN per tenant on the same ports, with both tenant SVMs left in the Default IPspace
  3. CA separate IPspace for each tenant, with that tenant's SVM, broadcast domains, and routes inside it Correct answer
  4. DTwo broadcast domains in the Default IPspace, one per tenant, each with its own failover group

Answer and explanation: IPspaces provide separate routing tables, so overlapping address ranges can exist on the same cluster for different tenants. Broadcast domains and VLANs separate layer 2 traffic, but SVMs in the same IPspace share one routing domain, so duplicate addresses would conflict. Two subnet objects with the same range in one IPspace are not allowed for the same reason.

Question 6 of 10, Storage Protocols and Connectivity

A LUN on node1 of a four-node cluster will be moved to an aggregate on node3, which is in a different HA pair. Selective LUN Map is at its default settings. What should the administrator do BEFORE starting the move so that the host does not lose optimized paths?

  1. ARemove node1 and its partner from the reporting nodes so that the host is forced to rediscover all paths
  2. BDisable ALUA on the igroup so the host treats every path in the cluster as active and optimized
  3. CAdd node3 and its partner as reporting nodes with lun mapping add-reporting-nodes, then rescan paths on the host Correct answer
  4. DCreate a new portset containing only the LIFs on node3 and bind it to the igroup after the move completes

Answer and explanation: With Selective LUN Map, only the owning node and its HA partner report paths, so the destination HA pair must be added as reporting nodes and the host must discover those paths before the move. Removing the current reporting nodes first would cut all existing paths. Binding a portset after the move is too late and only restricts paths, and ALUA should stay enabled so the host knows which paths are optimized.

Question 7 of 10, Storage Protocols and Connectivity

Volume app is junctioned at /app in SVM svm1, and its export policy allows client 10.1.1.20 read-write access. The client still gets access denied when mounting svm1:/app. The SVM root volume uses the default export policy. What should the administrator fix?

  1. ADisable NFSv3 on the SVM and enable NFSv4.1 so that the client mounts through a pseudo file system
  2. BChange the junction path of volume app to / so that the client mounts it without traversing the root volume
  3. CAdd a rule to the default policy used by the SVM root volume that gives the client at least read access Correct answer
  4. DSet the volume app security style to mixed so that both UNIX and NTFS clients can traverse the junction

Answer and explanation: An NFS client must be able to traverse every volume in the path, so the SVM root volume's export policy needs a rule that grants the client at least read access; the default policy has no rules out of the box. The junction path of the SVM root is reserved and cannot be reused for another volume. Security style changes and protocol version changes do not bypass export policy checks on the parent volume.

Question 8 of 10, Data Protection

A company needs to keep 90 daily Snapshot copies of volume finance on a backup cluster, while the source volume keeps only 7 daily copies. Which SnapMirror configuration meets this requirement?

  1. AA FlexClone of the source volume on the backup cluster that is refreshed daily by a Snapshot policy
  2. BA load-sharing mirror schedule that runs daily and keeps 90 copies of the volume on the backup cluster
  3. CA mirror relationship using the MirrorAllSnapshots policy with the destination Snapshot policy set to 90 copies
  4. DA vault relationship using the XDPDefault policy with a rule for the daily SnapMirror label and a retention count of 90 Correct answer

Answer and explanation: Vault relationships transfer Snapshot copies by SnapMirror label and keep them according to the destination policy's retention, so the backup can keep far more copies than the source. A MirrorAllSnapshots relationship keeps the same Snapshot copies as the source, and Snapshot policies do not run on DP destinations. Load-sharing mirrors and FlexClones are not designed for long-term backup retention across clusters.

Question 9 of 10, Data Protection

Two ONTAP clusters will replicate volumes with SnapMirror for the first time. Which network objects must exist on each node before a cluster peer relationship can be created?

  1. AIntercluster LIFs that can reach the intercluster LIFs on the other cluster Correct answer
  2. BData LIFs with the data-nfs service policy on the source and destination SVMs
  3. CCluster LIFs on the cluster interconnect ports with an MTU of 9000 on both clusters
  4. DNode management LIFs placed in a dedicated SnapMirror broadcast domain on both clusters

Answer and explanation: Cluster peering and SnapMirror traffic use intercluster LIFs, which must be able to communicate between the two clusters. Cluster LIFs carry only intra-cluster traffic on the cluster interconnect. Node management and data LIFs are not used for peer replication traffic unless an intercluster service policy is applied, which the default data and management roles do not include.

Question 10 of 10, Security

An administrator wants Snapshot copies of volume hr to be impossible to delete until their retention period ends, even by a cluster administrator, as protection against ransomware. Which feature meets this requirement?

  1. AA Snapshot policy with a larger count, so that copies are kept longer before rotation
  2. BTamperproof Snapshot copy locking, with a retention period on the Snapshot copies Correct answer
  3. CMulti-admin verification on the volume snapshot delete command for all administrators
  4. DA SnapMirror vault relationship to a second cluster with a 30-day retention rule

Answer and explanation: Snapshot copy locking uses the SnapLock compliance clock to make Snapshot copies undeletable until their retention time expires. A larger policy count still allows manual deletion. Multi-admin verification adds approvals but a group of administrators can still approve a deletion, and a vault keeps copies on another cluster without locking the copies on the source volume.