Exam Bundle

The incident report records lag at failover and validates the recovered transaction state against the approved RPO. That result is evidence for another corrective change: Monitor replication state and evaluate potential data loss before an unplanned asynchronous geo failover, following business RPO policy. The alternative problem is: A database team treats an Azure SQL active geo-replica as synchronously consistent and starts a data-critical failover without checking replication lag. The current Cedarfield Hospitality issue is instead: A database DR test promotes the secondary SQL Managed Instance and then reconnects applications without validating data consistency or application write permission. Its decisive acceptance evidence would be: The promoted endpoint permits intended application writes and the sample transaction reconciliation matches the agreed recovery checkpoint.

A controlled failback test shows only the current authorized primary accepts writes while the previous primary is safely fenced. That result is evidence for another corrective change: Implement supported cluster or failover-group role enforcement and application endpoint fencing before allowing writes after recovery. The alternative problem is: A post-failover application continues writing to a former primary after network recovery because the architecture lacks split-brain protection and role validation. The current Cedarfield Hospitality issue is instead: A database DR test promotes the secondary SQL Managed Instance and then reconnects applications without validating data consistency or application write permission. Its decisive acceptance evidence would be: The promoted endpoint permits intended application writes and the sample transaction reconciliation matches the agreed recovery checkpoint.

The secondary transitions to the intended online writable state and accepts an application connection after log application. That result is evidence for another corrective change: Apply required remaining log backups and perform the supported final recovery step before opening the secondary for production access. The alternative problem is: A SQL Server log shipping database is left in an unrecovered state during a planned disaster switch and users cannot connect to it. The current Cedarfield Hospitality issue is instead: A database DR test promotes the secondary SQL Managed Instance and then reconnects applications without validating data consistency or application write permission. Its decisive acceptance evidence would be: The promoted endpoint permits intended application writes and the sample transaction reconciliation matches the agreed recovery checkpoint.

Implement supported cluster or failover-group role enforcement and application endpoint fencing before allowing writes after recovery. That relates to a separate SQL administration problem: A post-failover application continues writing to a former primary after network recovery because the architecture lacks split-brain protection and role validation. The Meadowbrook Finance requirement is instead: A database DR test promotes the secondary SQL Managed Instance and then reconnects applications without validating data consistency or application write permission. That approach does not directly establish the required fix; the requested acceptance evidence is: The promoted endpoint permits intended application writes and the sample transaction reconciliation matches the agreed recovery checkpoint.

Monitor secondary compute, replica lag and connection capacity under representative read and disaster recovery workload. That relates to a separate SQL administration problem: A region-paired Azure SQL deployment has a secondary database that is too small for reporting load and the team mistakes successful replication for recovery readiness. The Meadowbrook Finance requirement is instead: A database DR test promotes the secondary SQL Managed Instance and then reconnects applications without validating data consistency or application write permission. That approach does not directly establish the required fix; the requested acceptance evidence is: The promoted endpoint permits intended application writes and the sample transaction reconciliation matches the agreed recovery checkpoint.

Use an appropriate failover group with stable read-write listener routing and supported failover policy when automatic app-level redirection is required. That relates to a separate SQL administration problem: An Azure SQL Database has active geo-replication configured but the application still connects only to the primary and cannot recover automatically when that region fails. The Meadowbrook Finance requirement is instead: A database DR test promotes the secondary SQL Managed Instance and then reconnects applications without validating data consistency or application write permission. That approach does not directly establish the required fix; the requested acceptance evidence is: The promoted endpoint permits intended application writes and the sample transaction reconciliation matches the agreed recovery checkpoint.

The destination database remains synchronized and post-cutover logins and scheduled operations pass independent validation tests. That result is evidence for another corrective change: Follow the supported Managed Instance link workflow and separately migrate required instance-level dependencies and application configuration. The alternative problem is: A multi-instance hybrid migration uses an availability-group-based link to SQL Managed Instance but assumes all on-premises logins and server jobs are migrated automatically. The current Meadowbrook Finance issue is instead: A database DR test promotes the secondary SQL Managed Instance and then reconnects applications without validating data consistency or application write permission. Its decisive acceptance evidence would be: The promoted endpoint permits intended application writes and the sample transaction reconciliation matches the agreed recovery checkpoint.

A post-failover client opens a new session through the listener to the current writable primary without code changes. That result is evidence for another corrective change: Use the supported availability group listener and required DNS or multi-subnet client settings for application connections. The alternative problem is: A team configures availability group replicas but hardcodes one VM hostname into the connection string and continues to fail after primary moves. The current Meadowbrook Finance issue is instead: A database DR test promotes the secondary SQL Managed Instance and then reconnects applications without validating data consistency or application write permission. Its decisive acceptance evidence would be: The promoted endpoint permits intended application writes and the sample transaction reconciliation matches the agreed recovery checkpoint.

A controlled failback test shows only the current authorized primary accepts writes while the previous primary is safely fenced. That result is evidence for another corrective change: Implement supported cluster or failover-group role enforcement and application endpoint fencing before allowing writes after recovery. The alternative problem is: A post-failover application continues writing to a former primary after network recovery because the architecture lacks split-brain protection and role validation. The current Meadowbrook Finance issue is instead: A database DR test promotes the secondary SQL Managed Instance and then reconnects applications without validating data consistency or application write permission. Its decisive acceptance evidence would be: The promoted endpoint permits intended application writes and the sample transaction reconciliation matches the agreed recovery checkpoint.

Monitor replication state and evaluate potential data loss before an unplanned asynchronous geo failover, following business RPO policy. That relates to a separate SQL administration problem: A database team treats an Azure SQL active geo-replica as synchronously consistent and starts a data-critical failover without checking replication lag. The Meadowbrook Finance requirement is instead: A SQL Server log shipping database is left in an unrecovered state during a planned disaster switch and users cannot connect to it. That approach does not directly establish the required fix; the requested acceptance evidence is: The secondary transitions to the intended online writable state and accepts an application connection after log application.

Follow the supported Managed Instance link workflow and separately migrate required instance-level dependencies and application configuration. That relates to a separate SQL administration problem: A multi-instance hybrid migration uses an availability-group-based link to SQL Managed Instance but assumes all on-premises logins and server jobs are migrated automatically. The Meadowbrook Finance requirement is instead: A SQL Server log shipping database is left in an unrecovered state during a planned disaster switch and users cannot connect to it. That approach does not directly establish the required fix; the requested acceptance evidence is: The secondary transitions to the intended online writable state and accepts an application connection after log application.

Inspect backup, copy and restore job history and restore the log shipping chain with suitable schedules and alerts. That relates to a separate SQL administration problem: An on-premises SQL Server log shipping secondary is consistently hours behind because copy and restore jobs fail without monitoring. The Meadowbrook Finance requirement is instead: A SQL Server log shipping database is left in an unrecovered state during a planned disaster switch and users cannot connect to it. That approach does not directly establish the required fix; the requested acceptance evidence is: The secondary transitions to the intended online writable state and accepts an application connection after log application.

The promoted endpoint permits intended application writes and the sample transaction reconciliation matches the agreed recovery checkpoint. That result is evidence for another corrective change: Use the documented Managed Instance failover and post-failover validation procedure including replica state and application identity. The alternative problem is: A database DR test promotes the secondary SQL Managed Instance and then reconnects applications without validating data consistency or application write permission. The current Meadowbrook Finance issue is instead: A SQL Server log shipping database is left in an unrecovered state during a planned disaster switch and users cannot connect to it. Its decisive acceptance evidence would be: The secondary transitions to the intended online writable state and accepts an application connection after log application.

A controlled outage follows the documented manual or automatic policy and the test record identifies the correct trigger. That result is evidence for another corrective change: Check the configured failover policy, grace period and supported automatic-failover scope and align the runbook with actual behavior. The alternative problem is: An Azure SQL failover group is set to manual but the operations runbook claims automatic recovery will occur after any transient outage. The current Meadowbrook Finance issue is instead: A SQL Server log shipping database is left in an unrecovered state during a planned disaster switch and users cannot connect to it. Its decisive acceptance evidence would be: The secondary transitions to the intended online writable state and accepts an application connection after log application.

A post-failover client opens a new session through the listener to the current writable primary without code changes. That result is evidence for another corrective change: Use the supported availability group listener and required DNS or multi-subnet client settings for application connections. The alternative problem is: A team configures availability group replicas but hardcodes one VM hostname into the connection string and continues to fail after primary moves. The current Meadowbrook Finance issue is instead: A SQL Server log shipping database is left in an unrecovered state during a planned disaster switch and users cannot connect to it. Its decisive acceptance evidence would be: The secondary transitions to the intended online writable state and accepts an application connection after log application.

Use an appropriate failover group with stable read-write listener routing and supported failover policy when automatic app-level redirection is required. That relates to a separate SQL administration problem: An Azure SQL Database has active geo-replication configured but the application still connects only to the primary and cannot recover automatically when that region fails. The Alpine Healthcare requirement is instead: A SQL Server log shipping database is left in an unrecovered state during a planned disaster switch and users cannot connect to it. That approach does not directly establish the required fix; the requested acceptance evidence is: The secondary transitions to the intended online writable state and accepts an application connection after log application.

Use the supported availability group listener and required DNS or multi-subnet client settings for application connections. That relates to a separate SQL administration problem: A team configures availability group replicas but hardcodes one VM hostname into the connection string and continues to fail after primary moves. The Alpine Healthcare requirement is instead: A SQL Server log shipping database is left in an unrecovered state during a planned disaster switch and users cannot connect to it. That approach does not directly establish the required fix; the requested acceptance evidence is: The secondary transitions to the intended online writable state and accepts an application connection after log application.

Configure and monitor Always On availability group health, synchronization and appropriate planned-failover readiness. That relates to a separate SQL administration problem: A SQL Server on Azure VM application requires database-level high availability but its Availability Group replica has not synchronized before a planned failover. The Alpine Healthcare requirement is instead: A SQL Server log shipping database is left in an unrecovered state during a planned disaster switch and users cannot connect to it. That approach does not directly establish the required fix; the requested acceptance evidence is: The secondary transitions to the intended online writable state and accepts an application connection after log application.

The secondary restore lag declines into the accepted range and all three log shipping stages complete in sequence. That result is evidence for another corrective change: Inspect backup, copy and restore job history and restore the log shipping chain with suitable schedules and alerts. The alternative problem is: An on-premises SQL Server log shipping secondary is consistently hours behind because copy and restore jobs fail without monitoring. The current Alpine Healthcare issue is instead: A SQL Server log shipping database is left in an unrecovered state during a planned disaster switch and users cannot connect to it. Its decisive acceptance evidence would be: The secondary transitions to the intended online writable state and accepts an application connection after log application.

After controlled regional failover the application reconnects using the listener and commits approved transactions on the new primary. That result is evidence for another corrective change: Use an appropriate failover group with stable read-write listener routing and supported failover policy when automatic app-level redirection is required. The alternative problem is: An Azure SQL Database has active geo-replication configured but the application still connects only to the primary and cannot recover automatically when that region fails. The current Alpine Healthcare issue is instead: A SQL Server log shipping database is left in an unrecovered state during a planned disaster switch and users cannot connect to it. Its decisive acceptance evidence would be: The secondary transitions to the intended online writable state and accepts an application connection after log application.

The promoted endpoint permits intended application writes and the sample transaction reconciliation matches the agreed recovery checkpoint. That result is evidence for another corrective change: Use the documented Managed Instance failover and post-failover validation procedure including replica state and application identity. The alternative problem is: A database DR test promotes the secondary SQL Managed Instance and then reconnects applications without validating data consistency or application write permission. The current Alpine Healthcare issue is instead: A SQL Server log shipping database is left in an unrecovered state during a planned disaster switch and users cannot connect to it. Its decisive acceptance evidence would be: The secondary transitions to the intended online writable state and accepts an application connection after log application.

Apply required remaining log backups and perform the supported final recovery step before opening the secondary for production access. That relates to a separate SQL administration problem: A SQL Server log shipping database is left in an unrecovered state during a planned disaster switch and users cannot connect to it. The Alpine Healthcare requirement is instead: A post-failover application continues writing to a former primary after network recovery because the architecture lacks split-brain protection and role validation. That approach does not directly establish the required fix; the requested acceptance evidence is: A controlled failback test shows only the current authorized primary accepts writes while the previous primary is safely fenced.

Use the supported availability group listener and required DNS or multi-subnet client settings for application connections. That relates to a separate SQL administration problem: A team configures availability group replicas but hardcodes one VM hostname into the connection string and continues to fail after primary moves. The Alpine Healthcare requirement is instead: A post-failover application continues writing to a former primary after network recovery because the architecture lacks split-brain protection and role validation. That approach does not directly establish the required fix; the requested acceptance evidence is: A controlled failback test shows only the current authorized primary accepts writes while the previous primary is safely fenced.

Check the configured failover policy, grace period and supported automatic-failover scope and align the runbook with actual behavior. That relates to a separate SQL administration problem: An Azure SQL failover group is set to manual but the operations runbook claims automatic recovery will occur after any transient outage. The Alpine Healthcare requirement is instead: A post-failover application continues writing to a former primary after network recovery because the architecture lacks split-brain protection and role validation. That approach does not directly establish the required fix; the requested acceptance evidence is: A controlled failback test shows only the current authorized primary accepts writes while the previous primary is safely fenced.

A controlled primary-node failure starts the SQL instance on the alternate node with the same database files accessible. That result is evidence for another corrective change: Implement a supported FCI storage and cluster topology with witness, network and service dependencies correctly configured. The alternative problem is: A SQL Server Failover Cluster Instance on Azure VMs lacks a properly configured shared storage or supported cluster storage solution, preventing failover. The current Alpine Healthcare issue is instead: A post-failover application continues writing to a former primary after network recovery because the architecture lacks split-brain protection and role validation. Its decisive acceptance evidence would be: A controlled failback test shows only the current authorized primary accepts writes while the previous primary is safely fenced.

The promoted endpoint permits intended application writes and the sample transaction reconciliation matches the agreed recovery checkpoint. That result is evidence for another corrective change: Use the documented Managed Instance failover and post-failover validation procedure including replica state and application identity. The alternative problem is: A database DR test promotes the secondary SQL Managed Instance and then reconnects applications without validating data consistency or application write permission. The current Alpine Healthcare issue is instead: A post-failover application continues writing to a former primary after network recovery because the architecture lacks split-brain protection and role validation. Its decisive acceptance evidence would be: A controlled failback test shows only the current authorized primary accepts writes while the previous primary is safely fenced.

The incident report records lag at failover and validates the recovered transaction state against the approved RPO. That result is evidence for another corrective change: Monitor replication state and evaluate potential data loss before an unplanned asynchronous geo failover, following business RPO policy. The alternative problem is: A database team treats an Azure SQL active geo-replica as synchronously consistent and starts a data-critical failover without checking replication lag. The current Alpine Healthcare issue is instead: A post-failover application continues writing to a former primary after network recovery because the architecture lacks split-brain protection and role validation. Its decisive acceptance evidence would be: A controlled failback test shows only the current authorized primary accepts writes while the previous primary is safely fenced.

Use an appropriate failover group with stable read-write listener routing and supported failover policy when automatic app-level redirection is required. That relates to a separate SQL administration problem: An Azure SQL Database has active geo-replication configured but the application still connects only to the primary and cannot recover automatically when that region fails. The Northgate Telecom requirement is instead: A post-failover application continues writing to a former primary after network recovery because the architecture lacks split-brain protection and role validation. That approach does not directly establish the required fix; the requested acceptance evidence is: A controlled failback test shows only the current authorized primary accepts writes while the previous primary is safely fenced.

Configure and monitor Always On availability group health, synchronization and appropriate planned-failover readiness. That relates to a separate SQL administration problem: A SQL Server on Azure VM application requires database-level high availability but its Availability Group replica has not synchronized before a planned failover. The Northgate Telecom requirement is instead: A post-failover application continues writing to a former primary after network recovery because the architecture lacks split-brain protection and role validation. That approach does not directly establish the required fix; the requested acceptance evidence is: A controlled failback test shows only the current authorized primary accepts writes while the previous primary is safely fenced.

Configure SQL Server availability group health and replica synchronization telemetry with actionable alerts. That relates to a separate SQL administration problem: An availability group becomes unhealthy but operators watch only VM CPU metrics and miss replica connection and synchronization failures. The Northgate Telecom requirement is instead: A post-failover application continues writing to a former primary after network recovery because the architecture lacks split-brain protection and role validation. That approach does not directly establish the required fix; the requested acceptance evidence is: A controlled failback test shows only the current authorized primary accepts writes while the previous primary is safely fenced.

A controlled replica disconnect is detected by the AG monitoring signal and the recovery runbook identifies the affected database. That result is evidence for another corrective change: Configure SQL Server availability group health and replica synchronization telemetry with actionable alerts. The alternative problem is: An availability group becomes unhealthy but operators watch only VM CPU metrics and miss replica connection and synchronization failures. The current Northgate Telecom issue is instead: A post-failover application continues writing to a former primary after network recovery because the architecture lacks split-brain protection and role validation. Its decisive acceptance evidence would be: A controlled failback test shows only the current authorized primary accepts writes while the previous primary is safely fenced.

The designated secondary handles the expected reads and a promotion rehearsal meets documented capacity thresholds. That result is evidence for another corrective change: Monitor secondary compute, replica lag and connection capacity under representative read and disaster recovery workload. The alternative problem is: A region-paired Azure SQL deployment has a secondary database that is too small for reporting load and the team mistakes successful replication for recovery readiness. The current Northgate Telecom issue is instead: A post-failover application continues writing to a former primary after network recovery because the architecture lacks split-brain protection and role validation. Its decisive acceptance evidence would be: A controlled failback test shows only the current authorized primary accepts writes while the previous primary is safely fenced.

After controlled regional failover the application reconnects using the listener and commits approved transactions on the new primary. That result is evidence for another corrective change: Use an appropriate failover group with stable read-write listener routing and supported failover policy when automatic app-level redirection is required. The alternative problem is: An Azure SQL Database has active geo-replication configured but the application still connects only to the primary and cannot recover automatically when that region fails. The current Northgate Telecom issue is instead: A post-failover application continues writing to a former primary after network recovery because the architecture lacks split-brain protection and role validation. Its decisive acceptance evidence would be: A controlled failback test shows only the current authorized primary accepts writes while the previous primary is safely fenced.