Free AZ-120 practice test: 30 questions with answers

The Database Migration Option (DMO) of Software Update Manager is SAP's supported way to combine a change of database platform to HANA with a migration or upgrade. Replicating VMs only moves the Oracle database as it is, restoring an Oracle backup does not change the platform, and copying tables with Data Factory does not give a consistent SAP system.

SAPS is the SAP-defined, hardware-independent measure of application performance that SAP sizing uses. IOPS and network throughput describe storage and network capacity rather than SAP workload, and clock speed does not reflect the overall capacity of a server.

SAP HANA writes its redo log on every commit, so log writes need low latency, while data volumes see large bursts of writes at savepoints. Separating them keeps log latency stable. The other options describe effects that do not drive the design, and the last one would disable a log that HANA needs for recovery.

Azure Hybrid Benefit lets Software Assurance Windows Server and SQL Server licences be used on Azure VMs. Reserved Instances lower compute price without reusing licences, Spot pricing offers interruptible capacity, and Dev/Test pricing is limited to non-production use.

A rehost, or lift-and-shift, moves the existing servers and database with minimal change, which suits a short cutover and a fixed database platform. Rebuilding on HANA, rewriting custom code or redesigning business processes each add scope and risk that these constraints rule out.

SAPS and memory size the application and database servers, and peak IOPS and throughput size the data and log storage, so both belong in the estimate. The SAP GUI version and the number of transport requests do not determine compute or storage capacity.

Azure Migrate discovers and assesses on-premises servers and uses Azure Site Recovery replication to move them with a short cutover. Storage Explorer copies files without discovery or replication, Backup restore points are not kept in sync, and Cost Management forecasts spend without moving workloads.

The SAP Product Availability Matrix lists which SAP releases, operating systems and databases are supported together, which is the compatibility check needed before migrating. Quick Sizer estimates hardware demand, the Pricing Calculator estimates cost, and Azure Advisor reviews existing Azure resources.

SAP publishes the list of certified IaaS platforms for SAP HANA, which is where a VM size is confirmed as certified. A Marketplace listing, a portal filter or an Advisor score does not show SAP certification for a specific VM size.

The VM extension for SAP solutions passes Azure infrastructure metrics to the SAP host so that SAP monitoring can show the VM's resource use. Installing the SAP GUI, encrypting database files and replacing load balancer health probes are not what the extension does.

SAP Deployment Automation Framework is an open-source project that uses Terraform for infrastructure and Ansible for SAP installation, run from the customer's own pipelines. Azure Center for SAP solutions is a managed service, exported ARM templates do not install SAP, and Site Recovery is for replication.

A private endpoint gives a PaaS service a private IP inside the virtual network, and public network access can then be disabled. Private endpoints do not expose the SAP application to the internet and do not remove the need for network security groups.

A proximity placement group keeps the grouped VMs in the same datacenter to reduce latency between them. Traffic Manager routes client requests, availability sets spread VMs across fault domains, and a paired region is for distant disaster recovery, so none of these places VMs together.

Azure Policy evaluates resources at deployment and, with deny effects, blocks creation of resources in non-approved regions or without required tags. Budgets only notify on spend, Defender for Cloud scores security posture, and action groups send notifications when alerts fire.

Management groups organize subscriptions into a hierarchy, and policies assigned to a group are inherited by the subscriptions below it. Tags and locks do not pass policy down a hierarchy, and Lighthouse delegates management to a service provider rather than organizing subscriptions.

A SAML 2.0 trust lets the SAP system accept sign-ins issued by Microsoft Entra ID, which provides single sign-on. Bastion gives secure administrative access, opening a public port does not authenticate users, and storing passwords in Key Vault does not enable federated sign-in.

Pacemaker needs fencing (STONITH) so that a node whose state it cannot confirm is powered off, which prevents split-brain. Load balancer probe timing, a public IP or an extra NIC for GUI traffic do not let the cluster confirm node state.

Availability zones are physically separate datacenters within a region, so VMs spread across zones survive a datacenter failure. Availability sets only isolate VMs from rack and host failures inside one datacenter, a proximity placement group co-locates VMs, and a paired region is a cross-region disaster recovery design.

Synchronous HANA System Replication makes each commit wait until the secondary has the log, so committed data is not lost when the primary fails within the supported distance between zones. Asynchronous replication, nightly log shipping and periodic snapshots all leave a window of unreplicated data.

Log backups record the transaction log between full backups, which allows restore to a point in time with Azure Backup for SAP HANA. Daily VM restore points and disk snapshots cannot recover to an arbitrary moment, and soft delete only protects files that were deleted.

Asynchronous HANA System Replication keeps a copy in the DR region without making each commit wait across the long distance between regions. Synchronous replication across regions adds latency to every commit, a nightly export loses a day of data, and load balancer rules cannot move database replication.

Paired regions are updated in a staggered order, and geo-redundant storage replicates data to the paired region, both useful in disaster recovery planning. Paired regions do not share datacenters, and nothing synchronizes HANA transactions between them automatically.

A floating IP lets the load balancer keep the cluster's virtual address with whichever node is currently active. Load balancing does not encrypt traffic, nodes keep their own addresses separately, and skipping the health probe would break failover detection.

Azure Monitor for SAP solutions collects SAP-specific metrics from HANA and NetWeaver and presents them with dashboards and alerts. Service Health reports Azure platform incidents, Advisor gives recommendations, and Cost Management tracks spend rather than SAP application health.

Scheduled Events exposes upcoming platform maintenance to the VM through the instance metadata service, which lets the application prepare gracefully. Maintenance configurations schedule customer-controlled windows, while Service Health and Advisor report information that the VM cannot read directly.

Stopping idle non-production VMs on a schedule with an automation runbook cuts compute charges while keeping the systems available in working hours. Spot VMs can be evicted at any time, enlarging VMs raises cost, and archive storage is unsuitable for active database files.

Azure Advisor analyzes resource configuration and usage and gives recommendations across the five pillars, including right-sizing. Policy dashboards report compliance, Network Watcher diagnoses network traffic, and Migrate assesses on-premises servers before migration.

Trend reviews for right-sizing and watching storage throughput against its limits both help keep SAP performance steady over time. Turning off monitoring removes the evidence needed to tune, and permanent oversizing wastes cost without managing performance.

Rolling patches move the SAP resources to the other node before each node is updated, so the service stays available. Patching both nodes together causes an outage, skipping patches leaves known vulnerabilities, and availability sets do not replace patching.

A Log Analytics workspace collects and stores logs from VMs, alerts and diagnostics so they can be queried together with Kusto Query Language. A Recovery Services vault holds backup data, Key Vault holds secrets and keys, and tags are labels rather than log stores.