ExamPark
Course menuSOA-C03

Exams / SOA-C03 / Study guide / Restores, versioning, RPO and RTO

Restores, versioning, RPO and RTO: SOA-C03 study notes and practice questions

What to know about Restores, versioning, RPO and RTO for the SOA-C03 exam, from the Reliability and Business Continuity domain (22% of the exam guide), then 2 free questions to test it. Part of the SOA-C03 study guide.

Last verified against current AWS documentation ·How we write our questions

What to know for the exam

Restore questions ask how to get data back after a mistake or failure, how much can be lost (RPO) and how long it takes (RTO).

RPO and RTO

  • Recovery point objective (RPO): the most data you can afford to lose, measured in time. Backups every 4 hours give an RPO of up to 4 hours.
  • Recovery time objective (RTO): how long you can be down while restoring.

Restore options

Data Restore What to know
RDS Point-in-time recovery from automated backups To any second in the retention window (up to 35 days), typically up to about 5 minutes before now. Creates a new DB instance with a new endpoint
RDS Restore from a snapshot Also creates a new instance; update the application or DNS to use it
Aurora MySQL Backtrack Rewinds the existing cluster in place in minutes, without a new cluster (if turned on)
DynamoDB Point-in-time recovery (PITR) Any second in the last 35 days (configurable); restores to a new table
EBS Create a volume from a snapshot In any AZ in the Region; data loads lazily unless fast snapshot restore is on
S3 Versioning A DELETE adds a delete marker; remove the marker (or copy back an earlier version) to restore
EC2 Launch from an AMI or AWS Backup recovery point

S3 protection features

  • Versioning keeps every version; MFA Delete requires MFA to delete versions or change versioning.
  • S3 Object Lock (governance or compliance mode) blocks deletion for a retention period.
  • Lifecycle rules can expire old noncurrent versions to control cost.
  • Replication gives a second copy in another bucket or Region.

Common traps

  • Expecting point-in-time recovery to overwrite the existing database. It creates a new one.
  • Deleting a delete marker vs deleting a version. Removing the marker brings the object back; deleting a specific version ID removes it permanently.
  • Backups less frequent than the RPO. Daily backups can’t meet a 1-hour RPO.

Try the 2 practice questions

Practice Restores, versioning, RPO and RTOTake the free practice exam

Question 1

Medium ·Link to this question (question 1)

A public library's catalog runs on an Amazon RDS for PostgreSQL DB instance with automated backups kept for 14 days.

At 10:42 this morning, a faulty script deleted thousands of loan records. The last manual snapshot was taken yesterday at midnight.

The library wants to recover the data as it was just before 10:42, with as little data loss as possible.

What should a CloudOps engineer do?

Choose one answer.

  1. A. Restore to 10:41 with point-in-time recovery into a new DB instance, then recover the records from it.
  2. B. Restore yesterday's midnight snapshot to a new DB instance and switch the application to the new instance's endpoint.
  3. C. Perform a point-in-time restore of the existing DB instance in place to 10:41, keeping its current endpoint.
  4. D. Create a read replica of the DB instance and promote it, because the replica still holds the rows the script deleted.
Show the answer and explanation

Answer: A

Explanation

Automated backups give RDS point-in-time recovery to any second within the retention window, typically up to about five minutes before now. A restore always creates a new DB instance, so the team either copies the lost rows back or moves the application to the new endpoint. Snapshots lose more data, and replicas copy the mistake.

Study

Question 2

Easier ·Link to this question (question 2)

A ski resort keeps trail maps and lift-status images in an S3 bucket that has S3 Versioning turned on.

A cleanup script mistakenly ran a simple DELETE request, without a version ID, against 3,000 current objects. The website now returns 404 errors for those images.

What is the quickest way for a CloudOps engineer to bring the objects back?

Choose one answer.

  1. A. Delete the delete markers that the script created, so that the previous versions become the current versions again.
  2. B. Restore the objects from the bucket's S3 Glacier archive tier, which keeps a copy of every object for 30 days after deletion.
  3. C. Turn on S3 Object Lock for the bucket and roll the objects back to the last locked state in governance mode.
  4. D. Open a support case to ask AWS to recover the deleted objects from the bucket's internal backups.
Show the answer and explanation

Answer: A

Explanation

S3 Versioning protects against accidental deletes. A simple DELETE places a delete marker on top of the object, and the earlier versions remain. Deleting the marker restores access to the previous version, so the fix takes minutes and copies no data.

Study

Want more on Restores, versioning, RPO and RTO? The 5 full-length practice sets have 10 more Restores, versioning, RPO and RTO questions, 325 in all, none of them in the free exam. One payment of $14.99, no subscription. See the practice sets

Study Restores, versioning, RPO and RTO

More Reliability and Business Continuity topics