Microsoft AZ-104 Exam Dumps, Practice Test Questions

The spr parameter restricts which transport protocols may be used with the SAS, such as HTTPS-only access. It protects how a download request reaches the service, not what filename the browser receives in Content-Disposition. The partner already downloads successfully; changing transport policy cannot supply the requested response header or override the filename.

The sp field limits which operations the SAS authorizes, including reading or writing the Blob. Granting write permission would widen the partner’s authority and would not instruct a browser to rename a downloaded file. Changing this signed field without producing a valid new signature would also invalidate the URL. Keep read permission and adjust the signed response header instead.

The rsct response override controls the Content-Type header, which tells clients the MIME type of returned data, such as application/pdf. MIME type can affect handling, but it does not set the attachment filename. The scenario asks for Content-Disposition on this signed link, so rscd must be included and signed rather than relying on rsct alone.

LRS keeps replicas within a single datacenter, protecting against some hardware faults but not independently surviving the loss of that site or its availability zone.

GRS adds asynchronous copies in a paired region, which the business specifically does not need. Its regional protection and cost exceed the stated zonal design target.

RA-GRS also replicates to another region and supports reads from the secondary. The additional secondary-read capability is unnecessary when only local zonal resilience is requested.

Geo-redundant storage copies data asynchronously to a second Azure region, so GRS supplies geographic durability. However, ordinary GRS does not offer the application a readable secondary endpoint before a failover. Because the reporting service must fetch blobs during primary-region failure without initiating failover first, replication alone is insufficient; it needs a read-access geo-redundant configuration.

Locally redundant storage keeps copies within a single region, which supports resilience to local hardware failures but does not give the application a secondary geographical endpoint. LRS has no independently readable remote regional copy for the described outage. Choosing LRS would leave the report reader dependent on the unavailable primary region rather than providing the required disaster-read continuity.

Zone-redundant storage distributes data across availability zones within the primary region, improving tolerance of a zone failure. The specified outage affects the entire region, and the service explicitly requires reads from a separate secondary region before failover. ZRS by itself does not maintain that geographic copy or expose a secondary regional read endpoint, despite being more resilient than LRS to zonal events.

GRS maintains an asynchronously replicated geographic secondary copy. LRS doesn't have geographic replication, so changing to LRS is not just a DNS access toggle. Assuming secondary data remains would overstate disaster-recovery coverage after the cost-reduction change.

LRS replicates data locally within a single Azure datacenter, not synchronously across Azure regions. GRS, by contrast, uses asynchronous replication to a secondary paired region. This option reverses both the topology and the resilience implications and would mislead a disaster-recovery review.

A redundancy conversion changes the account's replica configuration, not necessarily which region is primary. Azure Storage account failover is a separate controlled operation that promotes the geographic secondary for disaster recovery. Switching GRS to LRS isn't an automatic failover plus restoration of geo-redundancy.

ZRS protects against zone-level events within a region but does not by itself maintain a secondary regional copy for a full regional outage.

RA-GRS adds secondary read capability to cross-region replication, but does not inherently provide ZRS-style synchronous copies across primary availability zones.

LRS keeps replicas in a single datacenter facility within one region and does not supply the two separate resilience layers requested.

Moving from Archive to an online tier requires rehydration processing and potential charges. An instantaneous metadata change is not a safe assumption for converting account redundancy.

RA-GZRS also does not support Archive tier, and secondary readability is not the same as archive rehydration. The account must address the archive incompatibility before conversion.

Versioning can preserve historical objects but does not alter which redundancy configurations support Archive. Archived content remains a compatibility problem.

LRS does not spread copies across availability zones or provide a secondary regional endpoint, so it misses both additional requirements.

ZRS tolerates eligible zone failures in one region but lacks the requested separately readable cross-region secondary.

GRS replicates geographically, but does not promise zone redundancy in the primary nor customer read access to secondary before failover.

That behavior is not guaranteed for the geo-replication leg, so this claim could lead to an unsupported zero-data-loss design assumption.

Read access to a secondary does not change GRS's asynchronous cross-region write replication. RA-GRS allows reading the asynchronously replicated secondary and cannot guarantee every latest primary write reached the paired region before completion.

ZRS places data across zones in one region, not across paired regions. It cannot meet a cross-region replication objective by itself.

A primary acknowledgment does not guarantee the remote secondary has replicated the object. Assuming strong consistency can produce incorrect reports.

The secondary endpoint is read-only before failover. The client cannot force synchronization by writing the missing item there.

Authentication validity and replication timing are independent. The described symptom is an expected eventual-consistency window, not proof of token expiration.

Geo replication is asynchronous. A region loss can occur before the latest primary writes reach the secondary.

Deletion destroys the account rather than executing its supported failover workflow. It would undermine recovery.

Lifecycle manages blob aging and tier/deletion actions, not account-level disaster failover or service endpoints.

LRS maintains copies within the local data center/region; it is not a single unprotected physical disk. The missing layer is geographic redundancy.

Recreating the storage account risks losing the recovered data and does not make the failed region available. The supported operation is to plan re-replication after failover.

Unplanned failover does not immediately recreate the former cross-region redundancy. The original region may remain unavailable and asynchronous geo copy requires reconfiguration and time.