This browser is no longer supported.
Upgrade to Microsoft Edge to take advantage of the latest features, security updates, and technical support.
A proposal allocates all 40 TiB of ideal two-way-mirror logical capacity from an 80-TiB pool to current applications. Which review finding is most important?
Two-way mirroring always provides 80 TiB of logical capacity.
The plan hasn't accounted for overhead, repair reserve, infrastructure, growth, and reserve placement.
Deduplication guarantees enough capacity for any missing reserve.
A partition retains cluster quorum, but its storage can't supply sufficient usable data. What does a healthy witness do?
It provides the missing workload copy.
It overrides the volume's resiliency requirements.
It helps establish cluster ownership but can't make unavailable data readable.
A proposal combines a 4+4 hyperconverged rack-aware cluster, 4-millisecond inter-rack round-trip time, and external ReFS CSVs. The designer cites the SAN preview's 5-millisecond allowance. What should you conclude?
The design combines incompatible requirements from different architectures.
The design is valid because both architectures use two racks.
Adding a disk witness makes the local-drive design valid.
An ordinary three-node cluster uses three-way mirroring. Two nodes fail, leaving one data copy. Which statement is correct?
One remaining copy guarantees that the cluster and pool stay operational.
Data-copy survival is only one condition. Cluster and pool quorum can still prevent operation.
Dynamic quorum guarantees survival of any simultaneous two-node failure.
A LUN has four reported paths, but all four traverse the same switch. What can you conclude?
The data exists as four independent copies.
The switch can fail without affecting storage access.
The path count doesn't establish independence from that switch failure.
Using the capacity example in this module, what's the approximate budget for current workloads after 2 TiB physical overhead, 10 TiB physical reserve, three-way mirroring, 1 TiB logical infrastructure, and 3 TiB logical growth?
18.7 TiB.
22.7 TiB.
64 TiB.
A ReFS deduplication job has reached its 30-minute limit. What's the correct completion decision?
Assume that the volume is fully optimized because the duration elapsed.
Inspect ReFS job status, actual savings, workload impact, storage health, and the approved schedule state.
Run Disable-ReFSDedup to restore every block to its original physical layout.
Disable-ReFSDedup
A rack-aware volume retains usable data after a zone failure, but a VM has strict placement in the failed zone. What should you expect?
The witness stores and starts a replacement VM.
Storage protection automatically removes the placement restriction.
The VM doesn't fail over to the other zone. Recovery depends on the application and placement design.
A node returns after an outage and reports Up, but storage resynchronization is still active. Another node is scheduled for maintenance. What should you do?
Up
Continue because node reachability proves that protection is restored.
Wait for resynchronization and the other required health and placement conditions before reducing protection again.
Reformat the returning node's data drives to remove stale copies.
A CSV uses ReFS. What can you conclude about checksums?
Every file has data checksums enabled because the volume uses ReFS.
Checksums can reverse an accidental deletion replicated to every copy.
ReFS checksums metadata, but file-data checksum coverage depends on integrity streams and must be assessed for the workload.
An Azure dashboard still shows the previous healthy state after telemetry stops arriving. What should the operator conclude?
The current condition is unknown from that dashboard, so local evidence and the monitoring interruption need investigation.
No new samples mean that no new storage faults occurred.
The missing telemetry proves that the storage pool has failed.
An accidental deletion reaches every synchronous replica, but the storage system remains healthy. Which action addresses the lost data?
Rebalance data across the remaining drives.
Move the CSV to another owner node.
Recover from an appropriate known-good recovery point or application-supported recovery mechanism.
You must answer all questions before checking your work.
Was this page helpful?
Need help with this topic?
Want to try using Ask Learn to clarify or guide you through this topic?