[ECS] [request]: More Scheduled Task Metrics

**Tell us about your request**
Adding more built in metrics for scheduled tasks in ECS would help us get a more complete picture of the health of our tasks.

**Which service(s) is this request for?**
ECS

**Tell us about the problem you're trying to solve. What are you trying to do, and why is it hard?**
I want to have metrics for the health of schedule tasks/jobs that run on ECS. Thing that I am interested in are: execution time, the job ran to completion (zero exit code) or the job ended abnormally (non-zero exit code). Similarly the ability to notify a SNS topic when a task has a non-zero exit would help with monitoring the health of tasks.

I can create log metrics that will tell me some of this information, but this requires additional development effort to tweak and tune the log metrics. Also, I may miss some types of problems if my log metric filter doesn't catch them, whereas most failures would be easily found by monitoring non-zero exit of a task.

We're moving away from tasks running on a job server controlled by Jenkins to containerized tasks running on ECS, but we're concerned that we don't have the same level of alerting if something goes wrong.

**Are you currently working around this issue?**
I'm building out log metrics to monitor for errors and exceptions, and other events in the log to work around this.

**Additional context**
n/a

**Attachments**
n/a

No repository files, tests, or implementation entry points are identified in the request. First review the ECS scheduled-task and monitoring capabilities, then define the supported execution, exit-status, duration, and SNS-notification behavior; done means those requirements are specified and implemented with validation.

Written by the indexing model from the issue text.