Consider implementing tighter anomaly thresholds for automated alerting immediately after deployment events
As a result of a late (5:25pm ET) daily deployment which introduced a PR that impacted submissions and access to the 10-10EZ, authenticated Veterans were impacted overnight and into the next morning with not being able to access the form.
Since this occurred outside of east coast working hours, it brought to light that we should look at creating a monitor and alert that may indicate that something in the deployment has impacted our form, so that there is a smaller window of time that issues will go unaddressed.
Background
As a result of a late (5:25pm ET) daily deployment which introduced a PR that impacted submissions and access to the 10-10EZ, authenticated Veterans were impacted overnight and into the next morning with not being able to access the form.
Since this occurred outside of east coast working hours, it brought to light that we should look at creating a monitor and alert that may indicate that something in the deployment has impacted our form, so that there is a smaller window of time that issues will go unaddressed.