OO

Free OO practice questions

10 questions from our OO 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 OO practice exams

Question 1 of 10, Data Protection Operations (SnapMirror, SnapVault, SVM DR)

A backup team wants the DR cluster to keep 52 weekly Snapshot copies while the source keeps only 2 weekly copies. The source Snapshot policy already labels its weekly copies with the SnapMirror label weekly. What must be configured on the destination?

  1. AA MirrorAllSnapshots policy with -keep 52 on the relationship so every replicated weekly copy is retained longer
  2. BA modified source Snapshot policy keeping 52 weekly copies, because the destination always mirrors the source retention
  3. CA vault-type policy rule, snapmirror policy add-rule -snapmirror-label weekly -keep 52, on the relationship policy Correct answer
  4. DA local Snapshot policy on the destination volume with a weekly schedule that keeps 52 copies of the DP volume

Answer and explanation: Retention that differs from the source is a vault behavior: a rule in a vault-type policy (such as XDPDefault or MirrorAndVault) matches the SnapMirror label and sets the keep count on the destination. A mirror policy such as MirrorAllSnapshots simply reflects the source Snapshot copies and has no separate keep per label. Local Snapshot schedules cannot run on a read-only DP volume, and raising source retention would consume capacity on production, which the requirement avoids.

Question 2 of 10, Data Protection Operations (SnapMirror, SnapVault, SVM DR)

An operator must protect a production volume on cluster A by creating a new asynchronous relationship to an existing destination volume on cluster B. The cluster and SVM peers are already in place. Which sequence of commands, run on cluster B, establishes the relationship and performs the baseline transfer?

  1. Asnapmirror resync -source-path svmA:vol_db -destination-path svmB:vol_db_dst, which creates the relationship and copies the baseline in one step
  2. Bsnapmirror initialize -source-path svmA:vol_db -destination-path svmB:vol_db_dst, then snapmirror create to register the schedule afterwards
  3. Csnapmirror create -source-path svmA:vol_db -destination-path svmB:vol_db_dst -policy MirrorAllSnapshots, then snapmirror update -source-path svmA:vol_db on cluster A
  4. Dsnapmirror create -source-path svmA:vol_db -destination-path svmB:vol_db_dst -policy MirrorAllSnapshots, then snapmirror initialize on the destination path Correct answer

Answer and explanation: SnapMirror relationships are created and initialized from the destination cluster: snapmirror create defines the relationship and snapmirror initialize performs the baseline transfer. Running update on the source cluster is wrong because updates are destination-driven and require an initialized relationship. Initialize cannot run before the relationship exists, and resync is used to re-establish a broken relationship from a common Snapshot copy, not to perform a first baseline.

Question 3 of 10, Performance Monitoring and Tuning

qos statistics volume latency show reports a large Cluster latency component for a busy CIFS share. The share's volume lives on node 2 and the clients connect to a data LIF currently on node 4. Which action MOST directly reduces this latency component?

  1. AMove the volume to an SSD aggregate on node 4 and leave the LIF in place
  2. BMigrate or rehome the data LIF to a port on node 2, the node that owns the volume's aggregate Correct answer
  3. CIncrease the SMB credits granted by the SVM so that more requests can be in flight across the cluster interconnect at once
  4. DCreate a FlexCache volume of the share on node 4 so that reads are cached locally for clients on that LIF

Answer and explanation: The Cluster component measures time spent on the cluster interconnect when requests arrive on a LIF that is not on the node owning the volume, so putting the LIF on node 2 removes the indirect path. Moving the volume to node 4 would also work but adds a data move when a quick LIF change suffices, making it less direct. SMB credits do not affect interconnect latency, and FlexCache is meant for read-heavy remote or distributed access, not for fixing indirect access inside one cluster.

Question 4 of 10, Performance Monitoring and Tuning

An operator plans to add a new workload to node 3 and wants to know how much more work the node can absorb before latency rises sharply. Which metric is designed for this question?

  1. AThe physical-used-percent of node 3's aggregates, because performance capacity follows the amount of data stored on them
  2. BThe average CPU busy percentage of node 3 from sysstat, since any value below 100% leaves the same amount of safe headroom
  3. CThe number of volumes on node 3 compared with the platform maximum, because each volume consumes a fixed share of performance
  4. DThe node's headroom (performance capacity used), from the resource_headroom_cpu object or Unified Manager Correct answer

Answer and explanation: Headroom, or performance capacity used, compares current utilization with the point at which latency starts climbing, which is exactly what is needed to decide whether a node can absorb more load. Capacity used on aggregates does not indicate performance limits. Raw CPU busy values do not capture the latency curve, and volume count is a configuration limit unrelated to workload intensity.

Question 5 of 10, Upgrades, Patching and Lifecycle Maintenance

Before starting planned maintenance on an HA pair, an operator wants to prevent AutoSupport from automatically opening support cases for the expected takeover and giveback events. Which command should be run?

  1. Asystem node autosupport invoke -node * -type all -message MAINT=4h Correct answer
  2. Bevent notification destination delete -name support to stop forwarding events to NetApp
  3. Csystem node autosupport modify -node * -state disable for the duration of the window
  4. Dsystem node autosupport modify -node * -support disable -transport smtp until the work ends

Answer and explanation: Sending an AutoSupport message with MAINT=<duration> tells NetApp that a maintenance window is in progress so case creation is suppressed for that period. Disabling AutoSupport entirely loses diagnostic history and must be remembered and re-enabled. Deleting an EMS notification destination does not affect AutoSupport case creation, and changing the support or transport settings disrupts reporting rather than flagging maintenance.

Question 6 of 10, Upgrades, Patching and Lifecycle Maintenance

A cluster has SnapMirror relationships replicating from site A (source) to site B (destination). Both sites will be upgraded to a newer ONTAP release. Which upgrade order does NetApp recommend?

  1. AUpgrade the source cluster at site A first so new Snapshot copy formats are created before replication
  2. BUpgrade both clusters at the same time so that their ONTAP versions never differ during the window
  3. CBreak all SnapMirror relationships, upgrade the clusters in any order, and then perform new baselines
  4. DUpgrade the destination cluster at site B first, and then upgrade the source cluster at site A afterwards Correct answer

Answer and explanation: Upgrading the destination first ensures the destination always runs the same or a later ONTAP version than the source, which keeps replication compatible during the window. Upgrading the source first can leave the destination older than the source, which is the unsupported direction for some relationship types. A simultaneous upgrade is not required, and breaking relationships and re-baselining is unnecessary and costly.

Question 7 of 10, Troubleshooting and Recovery

A volume reports 40% free space, yet applications fail with "No space left on device" when creating small files. What should the operator check and change?

  1. ACheck volume show -fields space-guarantee and change it to none because thick volumes refuse new files near full
  2. BCheck storage aggregate show-space and add disks, because the aggregate is full even though the volume itself still reports free space
  3. CCheck volume show -fields files,files-used and raise the limit with volume modify -files; the inodes are exhausted Correct answer
  4. DCheck volume show -fields percent-snapshot-space and reduce the Snapshot reserve because reserve blocks are not available to new files

Answer and explanation: Many small files can exhaust the volume's inode count while blocks remain free, producing ENOSPC; volume modify -files raises the maximum. The Snapshot reserve does not block file creation while active file system space is free. A full aggregate under a thin volume would cause write failures but would not explain free volume space with file-creation-only errors, and the space guarantee setting does not limit file creation in this way.

Question 8 of 10, Troubleshooting and Recovery

An HA pair shows "Partial giveback" in storage failover show after maintenance. storage failover show-giveback indicates that some aggregates were not returned because a subsystem vetoed the giveback. What is the appropriate next step?

  1. ARun storage aggregate relocation start for the remaining aggregates, which is the normal way to finish an HA giveback
  2. BDisable and re-enable storage failover to clear the partial giveback state
  3. CResolve the veto reason that show-giveback reports, then rerun storage failover giveback -ofnode for the node Correct answer
  4. DReboot the partner node with system node reboot to force the aggregates home

Answer and explanation: Giveback vetoes come from subsystems such as CIFS sessions or ongoing operations; the operator should address the reason and retry the giveback; -override-vetoes exists but should be used only when the impact is understood. Rebooting the node risks a further takeover and outage. Aggregate relocation is for planned moves between partners, not completing giveback, and disabling failover removes HA protection without resolving the veto.

Question 9 of 10, Ransomware Protection and Security Operations. Choose 2.

Which TWO measures make Snapshot copies on a NAS volume tamperproof so that a compromised administrator account cannot delete them before retention ends? (Choose two.)

  1. AIncrease the Snapshot reserve to 50% so that autodelete never removes copies when the volume fills
  2. BInitialize the SnapLock compliance clock on the nodes and set a retention period in the Snapshot policy schedule Correct answer
  3. CAdd the Snapshot copies to a FlexClone volume so that they are owned by the clone and cannot be deleted
  4. DEnable Snapshot copy locking on the volume with volume modify -snapshot-locking-enabled true Correct answer
  5. ESet the Snapshot policy schedule prefix to Anti_ransomware_backup so that ARP protects the copies

Answer and explanation: Tamperproof Snapshot copies require Snapshot locking on the volume plus an initialized SnapLock compliance clock and a retention period, after which locked copies cannot be deleted by any administrator until they expire. A FlexClone only blocks deletion of its parent Snapshot copy while the clone exists, which an administrator can remove. A larger reserve does not stop manual deletion, and naming copies like ARP copies gives no protection.

Question 10 of 10, Monitoring, Alerting and Capacity Management. Choose 2.

Which TWO configurations let a third-party monitoring system collect ONTAP data securely and receive alerts? (Choose two.)

  1. AAdd the monitoring host as a trap destination with system snmp traphost add Correct answer
  2. BConfigure an SNMPv1 community with write access for the monitoring host
  3. CEnable telnet on the cluster management LIF for counter polling
  4. DSet AutoSupport to send a copy of every message to the monitoring host
  5. ECreate an SNMPv3 user with security login create -application snmp -authentication-method usm Correct answer

Answer and explanation: SNMPv3 users authenticated with USM provide secure polling, and a traphost receives SNMP traps for alerts. SNMPv1 communities are not secure, and write access is unnecessary for monitoring. Telnet is insecure and not used for counter collection, and AutoSupport messages are diagnostic bundles rather than a monitoring feed.