OCI Migration Architect flashcards
OCI Migration Architect · Build recall with a focused review session.
32 cards. Read the term and the definition here. The deck at the top of this page is optional.
- Cloud Adoption Framework phases
- OCI's structured migration method: Assess (inventory and readiness), Plan (waves and business case), Design (architecture), Deploy (execute and operate). Each phase produces artifacts that gate the next phase.
- Six Rs migration strategies
- Rehost (lift and shift), Replatform (minor tuning), Repurchase (move to SaaS), Refactor or re-architect, Retire, and Retain. Assign exactly one strategy per workload with a documented rationale.
- Workload discovery inventory
- Per-workload record of platform, OS, CPU/RAM/storage, network dependencies, RTO/RPO, licensing, data volume, and owner. Missing fields block wave planning and architecture design.
- TCO comparison components
- Compute, storage, egress, licensing (BYOL versus license-included), support, and staff or training costs. OCI Cost Estimator produces the baseline for on-premises versus cloud comparison.
- OCVS live vMotion
- OCI VMware Solution supports live vMotion of VMs from on-premises vSphere over FastConnect or VPN into the OCVS SDDC, enabling rehost with near-zero downtime. Verify region and shape availability first.
- Assessment scoring model
- Rate each workload on business criticality, technical complexity, and migration effort from one to five. High criticality with low complexity enters early waves; high complexity becomes refactor or retain.
- AWS to OCI mapping
- EC2 maps to Compute, S3 to Object Storage, RDS to Base Database or Autonomous Database, ELB to Load Balancer, Route 53 to DNS, and Direct Connect to FastConnect.
- Exadata migration options
- On-premises Exadata moves to Exadata Database Service on OCI or Exadata Cloud@Customer. Oracle RAC moves to RAC on bare metal or Autonomous Database depending on compatibility.
- VCN CIDR planning rule
- Use a non-overlapping /16 per VCN and /24 per subnet. Overlapping CIDRs with on-premises or peered VCNs break VPN and FastConnect routing and force re-IP or NAT.
- FastConnect capacity
- Each port provides 1 Gbps or 10 Gbps, with up to eight ports per cross-connect group. FastConnect is private dedicated connectivity; VPN over internet is the fallback path.
- Internet versus NAT Gateway
- Internet Gateway serves public subnets with bidirectional internet traffic. NAT Gateway gives private subnets outbound-only internet access, blocking unsolicited inbound connections while allowing updates and patches.
- Service Gateway purpose
- Routes traffic to Oracle Services Network such as Object Storage and Autonomous Database without traversing the internet or NAT. Traffic stays on the OCI backbone and is private.
- Security lists versus NSGs
- Security lists apply at subnet level; network security groups apply at VNIC level and are preferred for tiered applications. Both are allow-only stateful rules with no explicit deny.
- Availability domains and fault domains
- Availability domains are physically isolated datacenters within a region; fault domains are groupings inside an AD. Design multi-AD for high availability and cross-region for disaster recovery.
- Load Balancer versus NLB
- Load Balancer is Layer 7 for HTTP/HTTPS with flexible bandwidth. Network Load Balancer is Layer 4, preserves source IP, supports UDP, and targets extreme performance or non-HTTP protocols.
- Block Volume performance tiers
- Lower Cost, Balanced, Higher Performance, and Ultra High Performance scale IOPS and throughput with VPUs per GB. Boot volumes default to Balanced; size for IOPS before migration.
- Object Storage tiers
- Standard, Infrequent Access, Archive, and auto-tiering. Cross-region replication is asynchronous; bucket versioning and lifecycle policies support migration rollback and retention.
- Route table behavior
- Route tables attach per subnet; each rule names a destination CIDR or service and a target such as Internet Gateway, NAT Gateway, Service Gateway, DRG, or Local Peering Gateway. Most specific prefix wins.
- Data Pump migration
- Oracle expdp and impdp perform logical export and import. Use for cross-platform, cross-version moves and schema-level migrations; requires staging storage sized for the dump set.
- GoldenGate replication
- Extracts changes from source and applies to target for minimal-downtime cutover or active-active. Requires supplemental logging on source and a stable, low-latency network path.
- Zero Downtime Migration
- ZDM automates on-premises-to-OCI database moves in physical mode (RMAN, Data Guard) or logical mode (Data Pump, GoldenGate). Physical requires same endianness and compatible versions.
- Autonomous Database compatibility
- No OS-level access, no SYSDBA, and no direct upgrade from very old releases. Validate features, character set, and dependencies before choosing Autonomous Database over Base Database Service.
- Non-Oracle database targets
- MySQL maps to MySQL HeatWave; PostgreSQL to PostgreSQL on OCI or Autonomous Database where compatible; SQL Server to SQL Server on OCI. Assess feature gaps before replatforming.
- Application migration order
- Migrate the database first or in parallel with replication, then the application tier, then DNS cutover. Stateless tiers move first when possible to reduce risk.
- Container services on OCI
- OKE runs Kubernetes; Container Instances runs simple workloads; Container Registry stores images; OCI DevOps provides CI/CD. Match image architecture (x86 or Arm) to the chosen shape.
- Bulk data transfer
- Use OCI Data Transfer Service for large offline datasets, or rsync and rclone over FastConnect. Verify checksums after transfer and reconcile file counts before cutover.
- Wave planning criteria
- Group workloads by dependency, business unit, and risk. Each wave has entry and exit criteria, a rollback plan, and a named owner. First wave uses low-risk, low-dependency workloads.
- Cutover runbook steps
- Freeze changes, final replication sync, validate data, switch DNS or load balancer, smoke test, monitor, then decommission source. Each step has a rollback trigger and time budget.
- Replication lag threshold
- GoldenGate lag, Data Guard apply lag, and RMAN incremental apply must be near zero before cutover. Define an acceptable threshold such as five minutes and abort if exceeded.
- DNS cutover technique
- Lower TTL to 60 to 300 seconds at least 24 hours before cutover. Use weighted or failover routing to shift traffic gradually and keep old records during the rollback window.
- Rollback decision deadline
- Rollback plan includes a tested path to source, reconciliation for post-cutover writes, and a decision deadline such as four hours before rollback cost becomes prohibitive.
- Post-migration validation checks
- Compare row counts, checksums, application smoke tests, performance baselines, and security posture against pre-migration metrics. Document deviations and remediate before decommissioning source.