A developer is monitoring an application on an EC2 instance and has created a custom CloudWatch metric with 1-second granularity. The developer must receive an Amazon SNS notification within 30 seconds if a problem occurs. What should the developer do to meet this requirement?

Creating a high-resolution CloudWatch alarm for the custom metric is the correct approach. High-resolution alarms can be configured with a period of 10 seconds or 30 seconds, allowing for rapid evaluation and notification within the required 30-second timeframe. Standard CloudWatch alarms have a minimum period of 1 minute, which would not meet the 30-second notification requirement. Building a custom CloudWatch dashboard is for visualization, not real-time alerting. CloudWatch Logs Insights is for analyzing log data, not directly for alerting on custom metrics with low latency. Switching to a standard CloudWatch metric would not allow for 1-second granularity or the required rapid alerting.