Arguably, the most important aspect of agile/scrum cadences is the retro. A retro occurs at the end of every iteration. Iterations are typically 1 or 2 weeks. And a retro is a safe, respectful place where the team can share things that went well, not so well, and things to keep an eye on.

Major incidents in technology-land are never pleasant. Mistakes, while never encouraged, will happen, and when they do, how well the team and organization respond to the incident is important. Taking purposeful action to maximize the learning opportunity in each failure.

• Review the Incident. Be as open as possible at the impact. Don't sugar-coat. Quantify if possible.
• Avoid individual blame. Look at improving process - and yes, this does include individual discipline.
• Each person brainstorm things that could have been done to prevent it in the first place
• Each person brainstorm things that could have been done to prevent it in the future
• Actions to prevent the incident from occurring in the future