Use this for last-mile review. Keep it open during mixed OCI networking questions and pair it with the Resources when you want Oracle’s exact networking wording.
1Z0-1124-25 usually gets easier when you read the stem in this order:
- Path lane: where should traffic go?
- Filtering lane: what should be allowed or denied?
- Private-connectivity lane: DRG, VPN, FastConnect, service gateway, or NAT?
- Troubleshooting lane: route, filter, DNS, then gateway or attachment state.
OCI answer sequence
Use this when the stem mixes ingress, async delivery, reliability, security, or operations.
flowchart TD
S["Scenario"] --> I["Classify the interaction mode"]
I --> E["Pick API Gateway, Events, Notifications, Streaming, or Functions"]
E --> R["Check retry, idempotency, ordering, and dead-letter behavior"]
R --> S2["Check Vault, IAM, private exposure, logs, and auditability"]
Fast lane picker
| If the question is mainly about… |
Start with… |
Usual winning move |
| where traffic should go |
route and gateway choice |
path first |
| what traffic should be allowed |
NSGs or security lists |
filter second |
| private connectivity |
DRG, VPN, FastConnect, or service gateway |
classify path type first |
| public exposure |
internet gateway and edge design |
keep the edge deliberate |
| a broken connection |
route, filter, DNS, then gateway state |
do not guess randomly |
| subnet design |
public/private placement and route intent |
subnet purpose before instance details |
| traffic distribution |
load balancer shape and backend health |
distribution before security fine-tuning |
| name resolution |
private/public DNS view and resolver path |
names must match the chosen path |
Read every networking scenario in this order
- Identify the communicating endpoints and whether traffic is north-south, east-west, or hybrid.
- Decide whether the path should be public, private internet egress, private Oracle-services access, or private hybrid transport.
- Confirm the route target and route propagation story.
- Confirm filtering with NSGs, security lists, or upstream inspection.
- Check DNS resolution and resolved endpoint only after the path exists.
Network boundary table
| Boundary |
What it really answers |
Do not confuse it with |
| route table |
where traffic is sent |
whether traffic is permitted |
| NSG or security list |
whether traffic is allowed |
how traffic reaches a destination |
| gateway |
which network boundary is crossed |
DNS or name resolution |
| DNS |
which target name resolves to |
route selection or filtering |
| subnet |
where resources live and which route table/security context applies |
IAM or compartment policy |
| DRG attachment |
which network joins the private routing hub |
end-to-end firewall logic |
Gateway and path map
flowchart TD
Internet["Internet"] --> IGW["Internet Gateway"]
Private["Private Subnet"] --> NAT["NAT Gateway"]
Private --> SGW["Service Gateway"]
OnPrem["On-Premises or Other Network"] --> DRG["DRG"]
DRG --> Private
Fast rule: choose the traffic path first, then decide which control filters it.
Gateway chooser
| You need… |
Strongest first fit |
Why |
| public internet ingress or egress |
Internet Gateway |
direct public edge path |
| private subnet outbound internet only |
NAT Gateway |
egress without public IP on the instance |
| private access to Oracle services |
Service Gateway |
stay off the public internet for Oracle services |
| on-premises or cross-VCN routing hub |
DRG |
attachment-based private routing control |
| private subnet outbound internet only |
NAT Gateway |
no inbound internet exposure |
| public edge ingress to a service |
Internet Gateway plus edge/load-balancing design |
explicit public entry point |
Subnet and route reminders
| Topic |
Fast rule |
| public subnet |
route to Internet Gateway plus public-IP-bearing resources where appropriate |
| private subnet |
no direct internet ingress path by design |
| route table |
chooses next hop; wrong route can break a valid architecture |
| multiple route tables |
useful when tiers need different network behavior |
| overlapping CIDRs |
create peering, routing, and migration pain |
| compartment |
governance boundary, not packet path |
NSGs versus Security Lists
| You need… |
Strongest first fit |
Why |
| simple subnet-wide filtering |
Security Lists |
broader subnet boundary |
| granular workload-level control |
NSGs |
narrower app or workload scope |
| least-privilege tiering |
NSGs |
easier to model app-specific traffic |
Filtering and inspection traps
| Trap |
Better reading |
| using security lists for every tier because they are already there |
NSGs often express workload boundaries more cleanly |
| changing both NSGs and security lists blindly |
change the narrowest layer that owns the traffic rule |
| using routes as a security boundary |
routes send packets; filters and inspection decide what is allowed |
| forgetting return traffic logic |
both directions of the flow still matter |
| assuming private path means safe path |
private reachability is not least privilege by itself |
Path and control order
| Ask this first |
Then ask this |
| does a valid path exist to the target |
do rules permit the traffic on that path |
| is the path private or public by design |
is the workload attached to the correct gateway or DRG route |
| does the name resolve to the intended endpoint |
do filters permit the resolved destination |
Connectivity chooser
| Requirement |
Strongest first fit |
Why |
| lower-cost quick private setup |
IPSec VPN |
simpler hybrid path |
| steadier bandwidth and lower-latency expectations |
FastConnect |
dedicated private connectivity |
| centralized routing for many attachments |
DRG hub-and-spoke |
cleaner route control |
| many VCNs with centralized inspection or shared services |
DRG-centered architecture |
clearer network governance |
| OCI service access without internet traversal |
Service Gateway |
private Oracle-services path |
Hybrid and private-path traps
| Trap |
Better reading |
| choosing FastConnect automatically because it sounds more professional |
choose it for throughput, predictability, and hybrid scale needs |
| treating DRG like the private link itself |
DRG is the routing hub, not every underlying path |
| debugging application firewalls before proving the private route exists |
path first, filter second |
| assuming VPN and FastConnect are interchangeable except for price |
path stability, throughput, and enterprise expectations differ |
| forgetting route propagation and attachment state |
DRG design is more than creating the object |
DRG and hybrid reminders
| Topic |
Fast rule |
| DRG |
private-routing hub for attachments and transit patterns |
| attachment |
how a VCN or external network participates in DRG routing |
| propagation/import/export |
route learning and distribution logic matter to reachability |
| on-prem path |
transport plus route plus filter must all align |
| transit design |
central hub patterns simplify many-network connectivity but raise design responsibility |
Path and control traps
| Trap |
Better reading |
| using route tables as if they filter traffic |
routes choose path; NSGs and Security Lists filter traffic |
| treating DRG like a generic security control |
DRG is a routing hub, not a firewall |
| exposing resources publicly because private connectivity is “harder” |
strong answers keep exposure deliberate |
| choosing Security Lists automatically for all workloads |
workload-level control often points to NSGs |
| mixing subnet placement with IAM policy scope |
network placement and authorization are separate concerns |
| forgetting service dependencies such as DNS or load balancer health |
path success depends on multiple layers |
Load balancing and edge reminders
| Need |
Stronger first fit |
| public app entry point |
public load balancer with private backends |
| internal tier distribution |
private load balancer |
| backend availability |
health checks plus multiple backends |
| TLS or edge control |
explicit edge or load-balancing/security design |
| avoid single-instance exposure |
place the load balancer at the edge, not the app host |
DNS and name-resolution reminders
| Topic |
Fast rule |
| private endpoint |
private DNS must resolve to the private target |
| public endpoint |
public DNS and public path must align |
| split-horizon behavior |
internal and external clients may need different answers |
| hostname success |
does not prove the route or filters are correct |
| service discovery |
naming choices can hide the real path if you do not test resolution |
DNS and troubleshooting order
| Step |
What to verify |
| route |
route tables, attachments, and intended next hop |
| filter |
NSGs and Security Lists |
| name resolution |
DNS records and resolved targets |
| gateway state |
IGW, NAT, Service Gateway, or DRG configuration |
| backend health |
load balancer target status if traffic should be distributed |
Private-path troubleshooting table
| Symptom |
First checks |
| private instance cannot reach the internet |
route to NAT, subnet type, and egress policy |
| private workload cannot reach Oracle service |
service gateway path and relevant route attachment |
| on-premises cannot reach OCI subnet |
DRG attachment, route propagation, and hybrid path state |
| hostname works publicly but not privately |
DNS view and resolved target versus intended private endpoint |
| load balancer listener is up but app fails |
backend health, NSG/security list, route, and app port reachability |
| one subnet works and another fails |
subnet-specific route table and security differences |
Troubleshooting traps
| Trap |
Better reading |
| restarting instances before checking routes and filters |
path logic should be verified first |
| checking only one control plane |
route, security, DNS, and gateway state all interact |
| blaming DNS when the private path is missing |
confirm the network path exists first |
| blaming gateways when the resolved endpoint is wrong |
DNS can send traffic to the wrong target |
| forgetting health checks and backend sets |
traffic distribution failures are not always route failures |
High-confusion pairs
| Pair |
Keep this distinction clear |
| route table vs NSG |
path selection versus packet filtering |
| NSG vs security list |
workload-scoped control versus subnet-wide control |
| NAT Gateway vs Internet Gateway |
outbound-only access for private subnets versus public edge access |
| Service Gateway vs NAT Gateway |
private Oracle-service access versus general outbound internet egress |
| DRG vs VPN or FastConnect |
routing hub versus private transport mechanism |
| DNS failure vs path failure |
name resolution issue versus missing route or blocked traffic |
| subnet vs compartment |
network placement versus governance scope |
| load balancer vs gateway |
traffic distribution/control point versus path boundary crossing |
Decision order that usually wins
- Classify the issue as subnet placement, route/path, filter, DNS, hybrid transport, or load balancing.
- Choose the intended public/private path before touching security rules.
- Verify route target, DRG attachment state, or gateway type.
- Verify filters and only then verify DNS resolution to the intended endpoint.
- For distributed traffic, check backend health and listener/backend mapping before changing the network design.
Last 15-minute review
| If you only keep one thing from each lane… |
Remember this |
| gateways |
IGW public, NAT outbound only, SGW Oracle services, DRG private routing hub |
| filtering |
NSGs narrow, Security Lists broad |
| connectivity |
VPN and FastConnect solve different private-path needs |
| troubleshooting |
route, filter, DNS, gateway |
| architecture |
keep public edge and private lanes deliberate |
| subnet vs compartment |
placement boundary versus governance boundary |
| load balancing |
a healthy path still needs healthy backends |
What strong 1Z0-1124-25 answers usually do
- classify the issue first as path construction, filtering, private connectivity, or name resolution
- separate routing decisions from security-control decisions
- choose the narrowest network control that still solves the problem cleanly
- keep public exposure deliberate instead of mixing public and private responsibilities casually
- verify DRG, route propagation, DNS, and backend health instead of guessing at the failing layer
- keep hybrid transport choice tied to reliability and throughput needs, not branding