Measure the Actual vs. Planned Duration to Achieve Deployment Plans that Model Reality

A deployment plan can model anything: infrastructure changes, software releases, or configuration changes that span multiple departments.  While many of the core software deployment tasks that enterprises face today can be automated most systems still require a plan.   Most deployments still involve a coordinator to make sure that a large team is synchronized and executing according to a plan.

The key word in deployment plan is "plan."  It’s just a plan, and the start and end dates for various tasks in that plan often don’t match reality for a number of reasons.  Here are a few examples:

1. Step 32a of the database migration was delayed by two hours due to a problem with a storage controller.

2. A critical step in a plan for conducting a performance test had to wait for approval from a senior manager.

3. One department involved in a multi-project release failed to get the proper approval from production control.

4. A power outage in Sunnyvale, CA took the majority of the team offline and the release manager made the call to block until the team could be reengaged.

Systems fail, people become unavailable, and the best-laid plans are always interrupted by reality.  Some enterprises take the time to analyze past deployment plans and account for uncertainty, but most software organizations tend to assume the best and proceed forward without much thought to contingency planning.

In Plutora we measure everything about your deployment plan because we understand that you are going to have to do it again.  We want you to be able to look back on the last time you upgraded that firewall or the last time you did a multi-project release so that you can account for the risks you missed the first time.

Plutora measures actual time vs. planned time and gives you insight into exactly where problems happen.  If you work in a large organization that maintains playbooks or runbooks for complex, coordinated tasks use Plutora to perform after action analysis and review and use this data to iterate toward more realistic estimates.

The next time someone asks you how long something should take don’t stop and think about it and model out your estimate.  Look at Plutora and ask it how long it took the last time.

With Plutora you can stop estimating and start understanding.