HNI, a leading manufacturer of workplace furnishings and residential building products, was challenged with error-prone, manual, and scripted processes. The lack of visibility and a cumbersome reporting process for job validation exacerbated the situation, causing outages in production, an inability to give a good picture of their execution timeline, and an inability to deliver at the speed of business. However, that changed when HNI happened upon Flexagon at an Oracle conference.
Running manual processes put HNI’s IT team at a disadvantage in terms of visibility. Where were artifacts across the lifecycle? What deployments had been executed to which environments? What errors occurred post-deployment? The team had to generate specific log files to validate what they did five minutes prior. Yet these logs could be—and were—accidentally deleted. Running manual reports was an extra, time-consuming, and non-value-add step.
Manually writing shell scripts to deploy code, customizations, and configurations was equally challenging. Although scripts appear to offer automation, they still require a great deal of management and time, i.e., they are touched by multiple people who must remember to elevate to proper credentials or run on appropriate systems. The result was continual errors. If someone ran the script on the wrong server, with the incorrect permissions, or at the wrong time, this would cause an error that could lead to not only that execution failing but subsequent dependent deployments failing. “Fat fingering” was also a problem, as “DROP” often became “DORP.” The Apps Manager said, “Each time we saw a release or an event was bigger than one object, we exponentially grew in our risk to be able to do that successfully and do it in a timely fashion.”
HNI wanted one tool that provided support for their landscape, visibility across the pipeline, and allowed them to categorize all their artifacts; one source of the truth for validations—where it’s been, where it went, and how things went when it did, with the confidence that the execution ran accurately and when it was supposed to run.
The team initially found Flexagon at an Oracle conference, which spurred them to investigate the Continuous Delivery and Release Orchestration space. HNI specifically liked the holistic approach as they didn’t want to implement multiple tools, such as Jenkins for CI and Octopus Deploy for CD, and run them in tandem. They also found FlexDeploy’s out-of-the-box connectivity and functions for all their critical systems attractive. Rather than write scripts to perform a process and utilize a tool to run it, the team could create and execute an automated workflow in FlexDeploy using pre-built plugins.
In addition, the pipeline and release features would also allow the team an orchestration component they couldn’t find in other tools. The ability to define one construct that could be used with all types of customizations for EBS was a big selling point.
The team spent the beginning of 2019 getting the pieces in place and went live with their implementation in April 2019. On day one, they could go live with support for Oracle EBS running on OCI, Oracle SOA Suite, and Enterprise Data Warehouse. “The first weekend following our go-live, I was deploying to our production environments from FlexDeploy for all of those major systems.”
HNI has since added support for OIC, and their custom deployment of data fixes pre-126.96.36.199. The company added OBIEE, Informatica, and BizTalk deployments.
HNI immediately realized significant benefits from FlexDeploy. Manual steps have been eliminated, and large deployment events that took six hours were reduced to under an hour. Off-hours support was reduced by two hours a night, and the time from a request being raised to deployment has significantly improved.
Today, there is clarity around deployment times; when an approval gate is met, it’s automatically in a queue to go to the next deployment window. The team is now immediately aware of what did or didn’t succeed in deployments. “We’ve gained a much stronger position to say what’s going to happen when we deploy, and we can guarantee it’s going to be successful, which we didn’t have before. So now, when things are being pulled into the development lifecycle, we’re able to show that what we did in our first round of testing is what’s going to happen in production.”
The team had previously struggled with early mornings, outages, user errors, tracking and visibility, and extensive manual processes. FlexDeploy offered them a total solution, simplifying their processes and their lives. What one person previously completed in a day can now be finished in 15 minutes. Plus, there are fewer errors, more visibility, and immediate feedback. “The majority of our business partners raised their eyebrows, ‘You’re really done within an hour, I can really get back to work?’”