Automated Production Release Validation
The Company
A large insurance company deploys automated weekly releases across a broad range of software, supporting a wide range of complex distributed applications in geographically diverse locations. The Application Development team is particularly concerned with incorrect or incomplete software deployment to Production.
The Problem
There are complex interdependencies across multiple teams, each using their own processes and tools to deploy software to multiple environments, in which almost every device has its own unique configuration. The teams have no visibility into the Production environment once a release is deployed, and determining correct deployment is a manual process that is only instigated in a reactionary manner once they receive negative feedback from an end user or experience an outage.
This company was facing a Build vs. Buy decision at this point. They had already hired a consultant to come in to evaluate the situation, and give them an estimate as to how much time and money it would take to implement some rudimentary verification capabilities. The final estimate: six months to provide basic verification capabilities.
The Solution
The SignaCert team came in and demonstrated superior verification capabilities, and the ease of implementation that SignaCert Enterprise Trust Server™ (ETS) offers right out of the box for an initial 30 servers. This company chose Buy over Build after evaluating the SignaCert solution. The teams now use SignaCert's ETS to validate their Production code against a known and tested reference (trusted source), and have eliminated the guesswork and manually intensive process involved in verifying their software deployment. They now automatically validate production systems against the trusted reference and have plans to use this approach of measurement for all 800 of their enterprise production systems.
Results
The company integrated SignaCert ETS into their highly complex, custom release process. They now monitor the state of their production servers nightly and run comparisons so they can see drift (cumulative changes) over time. What was once a post-push, manual process that only took place after a problem was identified in production, is now an automated process that proactively identifies potential problems before they surface in Production.
- Time Savings—Reduced post-release build qualification time from 4 hours to 4 minutes per server.
- Process Assurance—Provided more verification capabilities out of the box in a matter of days than an expensive custom solution could provide in six months.
- Reliability—Drove up service levels while increasing capacity.
.