Cloning Steps ============ Run preclone: adpreclone.pl appsTier adpreclone.pl dbTier Copy files Configure: adcfgclone.pl appsTier adcfgclone.pl dbTier Run AutoConfig Validate services In Oracle E-Business Suite (EBS), AutoConfig is a core utility that automatically configures and maintains all application and database tier configuration files based on a centralized set of values. Reads values from XML/context files Applies them to templates Generates all required config files across the system Reads context XML Validates parameters Applies templates Rewrites configuration files Updates database profile options (in some cases) Why ADOP is important? Before R12.2 (using adpatch), patching required full system downtime ❌ With ADOP: Users can continue working during patching βœ… Downtime is limited only to the cutover phase πŸ”Ή Key Concept Behind ADOP ADOP uses dual file systems + database editions: 1. File Systems Run File System (fs1) β†’ Currently active system Patch File System (fs2) β†’ Where patches are applied 2. Database Feature Uses Edition-Based Redefinition (EBR) β†’ Allows multiple versions of database objects simultaneously πŸ”Ή ADOP Cycle (Very Important for Interviews) ============================================== One-Line Answer (For Interview) β€œADOP is the online patching utility in Oracle EBS R12.2 that uses dual file systems and Edition-Based Redefinition to apply patches with minimal downtime.” There are 5 main phases: 1. Prepare Sets up patch edition Synchronizes file systems adop phase=prepare 2. Apply Applies patches to patch edition (fs2) adop phase=apply patches=123456 3. Finalize Prepares system for cutover Compiles objects adop phase=finalize 4. Cutover (Only Downtime Phase 🚨) Switches: fs2 β†’ becomes run fs1 β†’ becomes patch adop phase=cutover 5. Cleanup Removes old editions Frees space adop phase=cleanup πŸ”Ή ADOP Cycle Flow (Simple View) Prepare β†’ Apply β†’ Finalize β†’ Cutover β†’ Cleanup =========================================================================== 6. Stage the Installation Media Download EBS R12 software and build the staging area . Unzip into staging directory: /u01/install/stage πŸ”Έ 7. Start Rapid Install Main installation tool: πŸ‘‰ Oracle Rapid Install Run: cd startCD/Disk1/rapidwiz ./rapidwiz πŸ”Έ 8. Choose Installation Type Options: Vision Instance (Demo) Production Instance (Real setup) πŸ”Έ 9. Configure Installation Parameters You will provide: Database SID Ports Hostnames Domain name APPS password πŸ”Έ 10. Run Installation Rapid Install will: Install Oracle Database Configure Application Tier Setup services ⏱️ Takes: 4–8 hours πŸ”Έ 11. Verify Installation After completion: Check services: adstrtal.sh apps/apps Access URL: http://hostname:8000/OA_HTML/AppsLogin πŸ”Έ 12. Run AutoConfig πŸ‘‰ AutoConfig adautocfg.sh πŸ”Έ 13. Apply Required Patches Apply latest AD/TXK patches Apply CPU patches πŸ”Έ 14. Post Installation Steps Change default passwords Configure Workflow Mailer Setup printers Enable backups πŸ”Ή R12.2 Additional Steps (Important πŸ”₯) For R12.2 (latest version): Configure WebLogic Server Setup dual file system (fs1 & fs2) Validate ADOP πŸ”Ή Basic Architecture After Installation Database Tier (oracle user) Application Tier (applmgr user) WebLogic + OHS running The most substantial difference is the inclusion of a weblogic server in place of 10.1.3 Oracle Home. Developer home (10.1.2) remains in both the release πŸ”Ή Common Interview Questions from Installation ❓ What is Rapid Install? πŸ‘‰ GUI tool used to install EBS and database together ❓ What are key users? oracle β†’ DB owner applmgr β†’ Apps owner ❓ What is staging area? πŸ‘‰ Location where installation files are stored before install πŸ”Ή Real-Time Issues During Installation 🚨 Issue: Rapidwiz fails βœ” Check logs: /tmp $INST_TOP/logs 🚨 Issue: Port conflict βœ” Change port pool during install 🚨 Issue: Missing OS packages βœ” Install dependencies and restart installation πŸ”Ή One-Line Interview Answer β€œOracle EBS R12 installation is done using Rapid Install, which installs and configures both database and application tiers after preparing OS, staging software, and providing configuration details.” πŸš€ Pro Tip For L2/L3 interviews, always mention: Rapidwiz AutoConfig Dual file system (R12.2) Post-install validation Dataguard ========== πŸ”₯ 1. Explain Oracle Data Guard Architecture Answer: Oracle Data Guard maintains high availability and disaster recovery using: Primary Database β†’ Production DB Standby Database(s) β†’ Replica(s) Types of Standby: Physical Standby Logical Standby Snapshot Standby πŸ”₯ 2. Difference between Physical and Logical Standby Answer: Feature Physical Standby Logical Standby Redo Apply Block level SQL level Structure Exact copy Can differ Use case DR Reporting πŸ”₯ 3. What are Data Guard Modes? Answer: Maximum Protection β†’ No data loss, may impact availability Maximum Availability β†’ No data loss (preferred) Maximum Performance β†’ Best performance, possible data loss πŸ”₯ 4. What is Redo Transport and Apply? Answer: Redo Transport β†’ Sending redo logs from primary to standby Redo Apply β†’ Applying redo logs on standby Processes: LGWR, ARCH, RFS, MRP πŸ”₯ 5. Explain Data Guard Broker Answer: πŸ‘‰ Oracle Data Guard Broker Centralized management tool Uses DGMGRL Automates failover/switchover πŸ”₯ 6. What is Switchover vs Failover? Answer: Switchover β†’ Planned role reversal (no data loss) Failover β†’ Unplanned (during disaster) πŸ”₯ 7. How do you perform Switchover? Answer: ALTER DATABASE COMMIT TO SWITCHOVER TO STANDBY; Or using broker: DGMGRL> switchover to standby_db; πŸ”₯ 8. How do you monitor Data Guard? Answer: SELECT database_role, open_mode FROM v$database; SELECT process, status FROM v$managed_standby; πŸ”₯ 9. What is Real-Time Apply? Answer: Redo is applied directly from standby redo logs (SRL) without waiting for archive logs. πŸ”₯ 10. What is Gap Detection and Resolution? Answer: Gap = Missing archive logs Detected via: SELECT * FROM v$archive_gap; Resolution: Automatic (FAL server/client) Manual copy if needed πŸ”₯ 11. Standby is lagging behind. What will you do? Answer: Check lag: SELECT name, value FROM v$dataguard_stats; Verify network latency Check apply process (MRP) Increase parallel apply Tune I/O πŸ”₯ 12. MRP process is not running. How to fix? Answer: ALTER DATABASE RECOVER MANAGED STANDBY DATABASE DISCONNECT; Check: Logs availability Disk space Alert log πŸ”₯ 13. How do you handle archive log missing? Answer: Identify gap: SELECT * FROM v$archive_gap; Copy missing logs manually Register: ALTER DATABASE REGISTER LOGFILE 'file_name'; πŸ”₯ 14. Primary DB crashed. What will you do? Answer: Perform failover: DGMGRL> failover to standby_db; Open standby: ALTER DATABASE OPEN; Rebuild old primary as standby πŸ”₯ 15. What is FSFO (Fast-Start Failover)? Answer: Automatic failover using observer process. πŸ”₯ 16. How do you check Data Guard health? Answer: Check: SELECT * FROM v$dataguard_stats; Broker: DGMGRL> show configuration; πŸ”₯ 17. What is Standby Redo Log (SRL)? Answer: Required for Real-Time Apply. Should match number of primary redo groups + 1. πŸ”₯ 18. How do you add standby redo logs? Answer: ALTER DATABASE ADD STANDBY LOGFILE SIZE 500M; πŸ”₯ 19. Data Guard sync issue (transport lag) Answer: Check: SELECT * FROM v$dataguard_stats; Fix: Network tuning Compression Increase bandwidth πŸ”₯ 20. How do you rebuild standby? Answer: Take RMAN backup Restore on standby Configure redo apply πŸ”₯ 21. What is Snapshot Standby? Answer: Temporary read-write standby for testing; later converted back. πŸ”₯ 22. How do you enable Data Guard Broker? Answer: ALTER SYSTEM SET dg_broker_start=true; πŸ”₯ 23. What is Observer in FSFO? Answer: Monitors primary DB and triggers automatic failover. πŸ”₯ 24. How do you troubleshoot ORA- errors in Data Guard? Answer: Check alert logs Check: SELECT * FROM v$dataguard_status; Validate network / redo logs πŸ”₯ 25. What are common Data Guard issues? Answer: Archive gap Lag MRP stopped Network issues Disk full πŸ”₯ Real-Time L3 Scenario Question ❓ Standby DB is 2 hours behind. Root cause? Strong Answer: Check transport lag vs apply lag Verify network throughput Check MRP performance Review I/O bottlenecks Analyze archive log generation rate Tune parallel apply πŸš€ L3 Interview Expectations At L3 level, interviewer expects: Deep understanding of redo flow Strong troubleshooting approach Knowledge of failover/switchover commands Experience with real production incidents πŸ”₯ One-Line Answer (L3 Level) β€œOracle Data Guard ensures disaster recovery by maintaining synchronized standby databases using redo transport and apply, with advanced features like Broker and FSFO for automation.” If you want next: πŸ”₯ Data Guard architecture diagram (visual explanation) 🧠 Top 20 tricky interview traps πŸ“Š Step-by-step DR drill runbook (real production) 1. LOG_ARCHIVE_CONFIG Purpose: Defines Data Guard configuration (DB names involved) LOG_ARCHIVE_CONFIG='DG_CONFIG=(PRIMARY,STANDBY)' βœ” Must include all DB_UNIQUE_NAME values βœ” Required for redo transport πŸ”Ή 2. LOG_ARCHIVE_DEST_n (Most Important ⭐) Purpose: Defines where redo logs are sent Example: LOG_ARCHIVE_DEST_1='LOCATION=/arch VALID_FOR=(ALL_LOGFILES,ALL_ROLES) DB_UNIQUE_NAME=PRIMARY' LOG_ARCHIVE_DEST_2='SERVICE=standby LGWR SYNC AFFIRM VALID_FOR=(ONLINE_LOGFILES,PRIMARY_ROLE) DB_UNIQUE_NAME=STANDBY' Key attributes: SERVICE β†’ TNS entry for standby LGWR / ARCH β†’ Redo transport method SYNC / ASYNC β†’ Protection mode AFFIRM / NOAFFIRM β†’ Disk write confirmation πŸ”Ή 3. LOG_ARCHIVE_DEST_STATE_n Purpose: Enable/disable redo destination LOG_ARCHIVE_DEST_STATE_2=ENABLE Options: ENABLE DEFER (temporarily stop sending redo) πŸ”Ή 4. FAL_SERVER (Fetch Archive Log Server) Purpose: Specifies standby DB to fetch missing logs FAL_SERVER=PRIMARY πŸ”Ή 5. FAL_CLIENT Purpose: Defines standby DB name for communication FAL_CLIENT=STANDBY πŸ”Ή 6. DB_UNIQUE_NAME Purpose: Unique name for each DB in Data Guard setup DB_UNIQUE_NAME=PRIMARY βœ” Must be different for primary & standby πŸ”Ή 7. STANDBY_FILE_MANAGEMENT Purpose: Controls file creation on standby STANDBY_FILE_MANAGEMENT=AUTO Options: AUTO β†’ Automatically create datafiles MANUAL β†’ Manual intervention required πŸ”Ή 8. REMOTE_LOGIN_PASSWORDFILE Purpose: Required for redo transport authentication REMOTE_LOGIN_PASSWORDFILE=EXCLUSIVE πŸ”Ή 9. LOG_ARCHIVE_FORMAT Purpose: Defines archive log naming LOG_ARCHIVE_FORMAT=%t_%s_%r.arc πŸ”Ή 10. DB_FILE_NAME_CONVERT Purpose: Maps primary file paths to standby DB_FILE_NAME_CONVERT='/prod/data','/standby/data' πŸ”Ή 11. LOG_FILE_NAME_CONVERT Purpose: Maps redo log paths LOG_FILE_NAME_CONVERT='/prod/log','/standby/log' πŸ”Ή 12. ARCHIVE_LAG_TARGET Purpose: Forces log switch after specified seconds ARCHIVE_LAG_TARGET=1800 βœ” Helps reduce data loss window πŸ”₯ SYNC vs ASYNC (Interview Favorite ⭐) SYNC Zero data loss Waits for standby acknowledgment ASYNC Better performance Possible data loss πŸ”₯ LGWR vs ARCH LGWR Real-time redo shipping Used in SYNC mode ARCH Ships archived logs Slower πŸ”₯ Example Real Production Setup LOG_ARCHIVE_CONFIG='DG_CONFIG=(PROD,DR)' LOG_ARCHIVE_DEST_1='LOCATION=/arch VALID_FOR=(ALL_LOGFILES,ALL_ROLES) DB_UNIQUE_NAME=PROD' LOG_ARCHIVE_DEST_2='SERVICE=DR LGWR SYNC AFFIRM VALID_FOR=(ONLINE_LOGFILES,PRIMARY_ROLE) DB_UNIQUE_NAME=DR' LOG_ARCHIVE_DEST_STATE_2=ENABLE FAL_SERVER=PROD FAL_CLIENT=DR STANDBY_FILE_MANAGEMENT=AUTO πŸ”₯ L3 Scenario Question ❓ Redo logs are not shipping to standby. What will you check? Answer: Check parameters: SHOW PARAMETER log_archive_dest; Verify status: SELECT dest_id, status, error FROM v$archive_dest; Check network (TNS) Verify listener Check standby availability Check alert logs πŸ”₯ Common Mistakes (Interview Trap 🚨) ❌ Missing DB_UNIQUE_NAME ❌ Incorrect SERVICE name ❌ DEST_STATE set to DEFER ❌ Password file mismatch πŸ”Ή One-Line Interview Answer β€œRedo parameters in Data Guard control how redo logs are generated, transported, and applied between primary and standby databases using parameters like LOG_ARCHIVE_DEST_n, FAL_SERVER, and DB_UNIQUE_NAME.” If you want next: RAC Interview Questions ======================= 1. What is Oracle RAC? Answer: πŸ‘‰ Oracle Real Application Clusters Oracle RAC allows multiple instances to run on different nodes accessing the same database, providing: High availability Scalability Fault tolerance πŸ”₯ 2. Explain RAC Architecture Answer: Multiple nodes (servers) Each node runs an instance Shared storage (ASM or SAN) Interconnect network (private network) Key components: Clusterware ASM Voting Disk OCR πŸ”₯ 3. What is Cache Fusion? Answer: πŸ‘‰ Cache Fusion Transfers data blocks between nodes via interconnect Avoids disk I/O Uses global cache πŸ”₯ 4. What are Voting Disk and OCR? Answer: Voting Disk - Maintains cluster membership ,Determines node eviction OCR (Oracle Cluster Registry) - Stores cluster configuration πŸ”₯ 5. What is SCAN in RAC? Answer: πŸ‘‰ Single Client Access Name Single hostname for client connections Resolves to multiple IPs Load balances connections πŸ”₯ 6. What are RAC background processes? Answer: LMON β†’ Global enqueue service monitor LMD β†’ Lock manager daemon LMS β†’ Global cache service LCK β†’ Lock process πŸ”₯ 7. What is Interconnect? Answer: Private network used for: Cache Fusion Block transfer between nodes πŸ”₯ 8. What is Node Eviction? Answer: When a node is removed from cluster due to: Network failure Heartbeat loss Voting disk issues πŸ”₯ 9. What is Split Brain? Answer: Condition where cluster nodes lose communication and act independently. βœ” Prevented using Voting Disk πŸ”₯ 10. What is Services in RAC? Answer: Logical grouping of workloads: Used for load balancing Can be moved between nodes πŸ”₯ 11. RAC performance issue – high GC waits Answer: Check: SELECT event, total_waits FROM v$system_event WHERE event LIKE 'gc%'; Root causes: Hot blocks Poor SQL design Excessive block transfers Fix: Partition tables Reduce contention Tune SQL πŸ”₯ 12. One node is down – what happens? Answer: Remaining nodes continue Sessions may reconnect Services failover πŸ”₯ 13. How do you check cluster status? Answer: crsctl stat res -t olsnodes -n πŸ”₯ 14. How do you start/stop RAC? Answer: crsctl start cluster -all crsctl stop cluster -all πŸ”₯ 15. What is ASM? Answer: πŸ‘‰ Automatic Storage Management Manages database storage,Provides striping & mirroring πŸ”₯ 16. What is Instance Recovery in RAC? Answer: When one instance crashes: Other instance recovers its transactions πŸ”₯ 17. What is Load Balancing in RAC? Answer: Client-side (SCAN) Server-side (services) πŸ”₯ 18. What is FAN? Answer: πŸ‘‰ Fast Application Notification Notifies apps about node/service failure πŸ”₯ 19. What is TAF? Answer: πŸ‘‰ Transparent Application Failover Automatically reconnects sessions πŸ”₯ 20. How do you troubleshoot node eviction? Answer: Check logs: $GRID_HOME/log//cssd/ Check: Network latency Interconnect issues Voting disk πŸ”₯ 21. What is CRS? Answer: Cluster Ready Services manages cluster resources. πŸ”₯ 22. What is GES and GCS? Answer: GES β†’ Global Enqueue Service (locks) GCS β†’ Global Cache Service (blocks) πŸ”₯ 23. What happens if interconnect fails? Answer: Nodes lose communication Node eviction occurs πŸ”₯ 24. RAC DB is slow only during peak time Answer: Check: GC waits Load imbalance Fix: Reassign services Tune SQL Add nodes πŸ”₯ 25. How do you monitor RAC performance? Answer: AWR reports (per instance & global) GV$ views: SELECT * FROM gv$session; πŸ”₯ Real-Time L3 Scenario ❓ Users report slowness only in RAC (not single instance) Strong Answer: Check GC waits Identify hot blocks Analyze interconnect latency Review SQL causing contention Check load distribution across nodes πŸš€ L3 Interview Expectations At L3 level, interviewer expects: Deep understanding of Cache Fusion & GC waits Strong troubleshooting for: Node eviction Interconnect issues Performance problems Hands-on with CRS, ASM, services πŸ”₯ One-Line Answer (L3 Level) β€œOracle RAC enables multiple instances to access a shared database with high availability using Cache Fusion and Clusterware, ensuring scalability and fault tolerance.” If you want next: πŸ”₯ RAC troubleshooting scenarios (very tough) 🧠 ASM interview questions πŸ“Š RAC architecture diagram (visual explanation) Virtual Private Database ============================= VPD (Virtual Private Database) in Oracle Corporation databases is a security feature that enables row-level and column-level access control. It’s also known as Fine-Grained Access Control (FGAC). Instead of giving users access to entire tables, VPD dynamically filters data so users only see what they’re allowed to see. 🧠 How VPD works (simple idea) When a user runs a query like: SELECT * FROM employees; Oracle automatically rewrites it behind the scenes to something like: SELECT * FROM employees WHERE department_id = SYS_CONTEXT('USERENV','SESSION_USER_DEPT'); πŸ‘‰ This filtering is invisible to the user. βš™οΈ Key components of VPD 1. Policy Function A PL/SQL function that returns a WHERE clause condition. Example: CREATE OR REPLACE FUNCTION emp_policy ( schema_name VARCHAR2, table_name VARCHAR2 ) RETURN VARCHAR2 AS BEGIN RETURN 'department_id = SYS_CONTEXT(''USERENV'',''SESSION_USER'')'; END; 2. Policy Attached to a table using the package: DBMS_RLS.ADD_POLICY Example: BEGIN DBMS_RLS.ADD_POLICY( object_schema => 'HR', object_name => 'EMPLOYEES', policy_name => 'EMP_POLICY', function_schema => 'HR', policy_function => 'EMP_POLICY', statement_types => 'SELECT' ); END; 3. Application Context Stores session-specific values used in the policy. Example: SYS_CONTEXT('USERENV','SESSION_USER') You can also define custom contexts. ============== Core AWS Architecture How would you design a highly available web application on AWS? Use multiple Availability Zones, an Application Load Balancer, Auto Scaling groups, Amazon RDS Multi-AZ, S3 for static assets, CloudFront for caching, and Route 53 for DNS. What is the difference between scalability and elasticity? Scalability means the system can handle growth. Elasticity means resources automatically expand or shrink based on demand. What is the AWS Well-Architected Framework? It is a set of best practices across pillars such as operational excellence, security, reliability, performance efficiency, cost optimization, and sustainability. How do you design for fault tolerance in AWS? Deploy across multiple AZs, use health checks, load balancing, Auto Scaling, backups, replication, and managed services where possible. What is the difference between vertical and horizontal scaling? Vertical scaling means increasing instance size. Horizontal scaling means adding more instances. EC2 and Compute What are the different EC2 pricing models? On-Demand, Reserved Instances, Savings Plans, Spot Instances, and Dedicated Hosts. When would you use Spot Instances? For fault-tolerant, flexible workloads such as batch jobs, CI/CD workers, data processing, and stateless applications. What is an Auto Scaling group? A group of EC2 instances that automatically adjusts capacity based on policies, schedules, or metrics. What is the difference between an AMI and an EC2 instance? An AMI is a template used to launch instances. An EC2 instance is the running virtual server. How do you make an EC2-based application highly available? Place instances in multiple AZs behind a load balancer and use Auto Scaling. VPC and Networking What is a VPC? A logically isolated network in AWS where you can launch resources. What is the difference between a public and private subnet? A public subnet has a route to an Internet Gateway. A private subnet does not directly route to the internet. What is a NAT Gateway used for? It allows instances in private subnets to access the internet for outbound traffic while preventing inbound internet access. What is the difference between Security Groups and Network ACLs? Security Groups are stateful and attached to resources. NACLs are stateless and applied at the subnet level. How would you connect an on-premises data center to AWS? Use Site-to-Site VPN for encrypted connectivity or AWS Direct Connect for dedicated private connectivity. Storage What is the difference between S3, EBS, and EFS? S3 is object storage. EBS is block storage for EC2. EFS is managed shared file storage. What are S3 storage classes? Standard, Intelligent-Tiering, Standard-IA, One Zone-IA, Glacier Instant Retrieval, Glacier Flexible Retrieval, Glacier Deep Archive, and others depending on workload needs. How do you secure data in S3? Use bucket policies, IAM policies, Block Public Access, encryption, versioning, access logs, and least privilege. What is S3 versioning? It keeps multiple versions of objects, helping with accidental deletion or overwrite recovery. When would you use EFS instead of EBS? Use EFS when multiple EC2 instances need shared file storage. Databases What is the difference between RDS and DynamoDB? RDS is managed relational database service. DynamoDB is managed NoSQL key-value/document database. What is RDS Multi-AZ? A high-availability setup where AWS maintains a synchronous standby replica in another AZ. What is the difference between read replicas and Multi-AZ? Read replicas improve read scalability. Multi-AZ improves availability and failover. When would you choose Aurora? When you need high performance, managed relational database features, replication, and high availability. How do you back up databases in AWS? Use automated backups, manual snapshots, point-in-time recovery, cross-region backups, and AWS Backup. Security and IAM What is IAM? Identity and Access Management controls access to AWS services and resources. What is the principle of least privilege? Give users, roles, and services only the permissions they need. What is the difference between an IAM user, group, role, and policy? A user is an identity, a group contains users, a role is assumed temporarily, and a policy defines permissions. How do you secure AWS root account access? Enable MFA, avoid using it for daily tasks, remove access keys, and use IAM roles/users for operations. What is KMS used for? AWS Key Management Service manages encryption keys for securing data. Load Balancing and DNS What are the types of AWS load balancers? Application Load Balancer, Network Load Balancer, Gateway Load Balancer, and Classic Load Balancer. When would you use an ALB vs NLB? ALB is for HTTP/HTTPS Layer 7 routing. NLB is for high-performance Layer 4 TCP/UDP traffic. What is Route 53? AWS’s DNS service that supports domain registration, routing policies, and health checks. What are Route 53 routing policies? Simple, weighted, latency-based, failover, geolocation, geoproximity, and multivalue answer routing. Monitoring and Operations What is CloudWatch? A monitoring service for metrics, logs, alarms, and dashboards. What is CloudTrail? A service that records AWS API activity for auditing and compliance. What is AWS Config? A service that tracks resource configuration changes and evaluates compliance. How do you troubleshoot high latency in an AWS application? Check CloudWatch metrics, load balancer metrics, application logs, database performance, network latency, scaling policies, and external dependencies. Migration and Hybrid Cloud How would you migrate an application to AWS? Assess the workload, choose migration strategy, design target architecture, migrate data, test, cut over, and optimize. What are the 7 Rs of migration? Rehost, replatform, refactor, repurchase, retire, retain, and relocate. How do you migrate a database to AWS? Use AWS DMS, native backup/restore, replication, snapshots, or database-specific migration tools. Scenario-Based Questions A website is experiencing sudden traffic spikes. What would you do? Use Auto Scaling, ALB, CloudFront, caching, database read replicas, and performance monitoring. An application must survive an AZ failure. How would you design it? Deploy across multiple AZs, use load balancing, Multi-AZ databases, redundant NAT gateways, and replicated storage. How would you design a disaster recovery solution? Choose a DR strategy: backup and restore, pilot light, warm standby, or multi-site active-active. How would you reduce AWS cost? Right-size resources, use Savings Plans or Reserved Instances, delete unused resources, use S3 lifecycle policies, use Spot Instances, and monitor with Cost Explorer. How would you secure a three-tier application? Public subnet for load balancer, private subnets for app servers, isolated/private subnets for databases, security groups, IAM roles, encryption, WAF, and logging. How would you handle secrets in AWS? Use AWS Secrets Manager or SSM Parameter Store, encrypt with KMS, rotate secrets, and avoid hardcoding credentials. How would you design a serverless application? Use API Gateway, Lambda, DynamoDB, S3, EventBridge, Step Functions, CloudWatch, and IAM roles. How would you improve performance of a global application? Use CloudFront, Route 53 latency-based routing, Global Accelerator, regional deployments, and caching. How would you design for compliance and auditing? Use CloudTrail, Config, GuardDuty, Security Hub, IAM Access Analyzer, encryption, logging, and least privilege.