1Z0-997-25 Question 91
Single answerYou are managing a DB System in Oracle Cloud Infrastructure that is experiencing high CPU utilization and long query response times. The on-premises team complains about applications timing out during peak hours. What is the most appropriate first step to troubleshoot the database performance issues?
- A
Manually scale up OCPUs and memory right away to reduce resource contention.
- B
Shut down and re-create the DB System from a backup to eliminate potential corruption.
- C
Collect performance metrics using AWR or ASH reports and OS-level monitoring to identify top wait events.
- D
Perform a rolling patch upgrade of the database software to eliminate any known bugs.
Show answer and explanation
Correct answer: C
Explanation
When troubleshooting performance issues on an Oracle DB System in OCI, it is best practice to identify the underlying cause rather than taking immediate remedial actions. Oracle� AWR (Automatic Workload Repository) and ASH (Active Session History) reports, along with OS-level metrics available in the OCI Console and the database, can guide you to the main wait events and resource-intensive queries. Once you have this data, you can take targeted action to address the root cause. Refer to Oracle Cloud Infrastructure documentation on 'Monitoring and Tuning Databases' and Oracle DB performance tuning guides for best practices.
- A. Incorrect.
Incorrect. While scaling up may temporarily help by providing more resources, doing so without first diagnosing the root cause can lead to unnecessary costs and potentially mask the real issue.
- B. Incorrect.
Incorrect. Re-creating the entire DB System from a backup is a drastic measure. It could introduce downtime and may not address the underlying performance bottleneck if the issue isn�t data corruption.
- C. Correct.
Correct. Oracle� AWR and ASH reports, in addition to OS-level metrics on CPU, I/O, and memory, are key diagnostic tools for pinpointing bottlenecks such as high wait events or inefficient queries.
- D. Incorrect.
Incorrect. Applying patches or upgrading without root-cause analysis can be risky. While a patch might fix known bugs, it does not necessarily address performance if the bottleneck is due to query design or configuration.