[improvement] custom metrics and alarm support in apps functions

## Background

There are currently no custom metrics implemented in the apps functions. I want to utilize this issue to track potential custom metrics and metric alarms for these functions.

### Potential Metrics

* Log a metric when the `App is already running` error occurs
* Number of logs collected/sent to classifier lambda

### Potential Alarms

* Create alarm using the `App is already running` error that will alert if this occurs more than N times in NN period.

No files or tests are named. Start by locating the apps functions, the `App is already running` error, and the classifier lambda path; define the metric and alarm scope before implementing. Done should include agreed metrics and an alarm for the specified error threshold and evaluation period.

Written by the indexing model from the issue text.