Exoscale Certified Cloud Professional study plan

Map the Compute domain objectives into a personal checklist and mark each item as known, partial, or unknown before studying anything else.

Create an Exoscale account, generate API keys, and install the CLI so you can launch and inspect instances from the command line.

Launch instances across at least two instance families and record how vCPU, memory, and CPU model differ in practice.

Practice start, stop, reboot, and destroy lifecycle actions and note which actions preserve the disk versus which bill differently.

Write down the difference between a compute instance and the platform's managed Kubernetes offering in terms of who owns the control plane.

Attach a block storage volume to a running instance, format and mount it, then detach and reattach it to a second instance to confirm data persistence.

Resize an instance and predict in advance whether the resize requires a stop, then verify your prediction against the actual behavior.

Create an anti-affinity group and prove that two instances in the same group are placed on different physical hosts.

Compare object storage buckets with block volumes on durability, access pattern, and typical use case, and state when each is the wrong choice.

Configure a bucket policy or ACL that makes one object publicly readable while keeping the rest private, then test both paths.

Enable versioning or lifecycle rules on an object storage bucket and verify that an expired object is actually removed on schedule.

Explain the platform's snapshot behavior for block volumes and demonstrate restoring a volume from a snapshot into a new volume.

Provision a managed database service and connect to it from an instance over the private network only, with public access disabled.

Configure an automated database backup and perform a point-in-time or snapshot restore into a separate service instance.

Document which storage and database services are zone-scoped versus regional, and what that means for a failure scenario.

Compare the managed database engines offered and list one workload where each engine is the correct pick and one where it is not.

Set up a private network, place two instances and a database in it, and confirm that the database is unreachable from the internet.

Create a network load balancer in front of two instances and verify health-check failover by stopping one backend.

Configure security groups so that only the load balancer can reach the application port, then prove the direct path is blocked.

Draft the questions you still cannot answer about database scaling, replicas, and maintenance windows, and resolve at least half of them from documentation.

Set up a zone-aware architecture with resources in two zones and describe exactly which failure each zone protects against.

Configure an elastic IP and move it between two instances to simulate a failover with minimal DNS change.

Build a DNS record set for a domain you control, including an A record, a CNAME, and a short TTL used for failover testing.

Use the platform's private connectivity option to reach an external service or your on-premises network and document the routing involved.

Measure and record latency between two instances in the same private network and between zones to build intuition for placement decisions.

Create a non-root IAM user or API key with least-privilege permissions and prove that an over-broad key is rejected for a forbidden action.

Enable key rotation practice by issuing a new API key, updating automation to use it, and revoking the old key without downtime.

Harden an instance: disable password SSH, enforce key-based login, and restrict the SSH port to your own address via security groups.

Encrypt data at rest and in transit for a database connection and verify the TLS handshake from a client.

Review audit or activity logs for your account and identify at least three events that would matter in an incident investigation.

Describe what the platform's AI offering provides, how you authenticate to it, and which deployment or data-handling model applies to prompts you send.

Run a minimal working example against the AI service and record latency, token limits, and error responses you encounter.

Map every service you have used to the zone or region where its data resides and flag any service that stores data outside your intended jurisdiction.

Summarize the platform's compliance posture and certifications relevant to a regulated customer, citing the categories of attestation rather than inventing specifics.

Write the data residency questions you would ask a vendor and answer them for your own demo architecture.

Rebuild your full demo architecture from scratch in under two hours using only your own notes, to expose gaps in documentation.

Take two timed practice sets weighted by domain percentage and log every missed question by domain and root cause.

Re-study only the domains where your practice accuracy fell below your target, using your earlier checklists.

Write a one-page summary of the five facts you most often got wrong and the correct version of each.

Do a final pass over the trust, compliance, and residency material since it is easy to under-prepare relative to its weight.