The digital economy has IT up against the wall. Lines of business are focused on boosting innovation and enhancing the customer experience through digital transformation initiatives, but network complexity can be a significant obstacle for IT. According to one study, 75% of CIOs admit that the network is impacting their organization’s ability to achieve business goals, with an estimated 35% of all network downtime attributed to human error.
This problem has understandably led IT to embrace automation in each technology domain, including the network, as a means of accelerating service delivery. But automating service delivery within each domain alone isn't enough. There must be seamless automation across the entire data center, or the ability to perform consistently and efficiently will suffer serious setbacks. It’s this growing need for agility and operational efficiency that ultimately validates the business case for introducing open, cross-domain automation capabilities.
Simply automating and orchestrating a single domain within the IT services delivery chain fails to address the desired agility and efficiency goals. After all, even when operating with automated, functional silos (including network, compute, storage, and applications), the execution of tasks spanning multiple domains is unacceptably time-intensive. The innovation velocity and efficiency associated with cross-domain automation is what ultimately enables digital transformation.
Two Business Scenarios for Cross-Domain Technology
Consider a business unit that is launching an ad campaign for a new product and expecting a massive influx of online orders. The ability of the website to meet this demand is critical to the success of the campaign, the product launch, and the business itself. However, at the peak of the campaign, a critical network link to the company’s main database server goes down—cutting off the ability to deliver dynamic web content and process orders.
While this would be catastrophic for most organizations, it’s not for an organization equipped with open, cross-domain automation:
- In this scenario, the log system of the impacted network device instantly sends an alert to the monitoring system.
- Using open, cross-domain integration, this alert triggers an automated troubleshooting workflow from the organization’s cross-domain automation platform in just seconds.
- The automation platform then uses the information to locate and log into the failing network device, identify the failed service, and issue a restart command. (At critical junctures, the troubleshooting workflow posts status messages to inform support staff.)
- In less than one to two minutes, and without involving a single support engineer, the database server connectivity is re-established and the website is back online. (After successful resolution, the workflow automation even creates a ticket with all relevant information to streamline future analysis.)
The total time to recovery is approximately one to two minutes.
However, in an organization that lacks open, cross-domain automation the same scenario is riddled with latency and delays. Here’s the same workflow as before, only executed using a manual workflow common to most organizations:
- An on-site network operations center engineer must recognize and acknowledge an alert among many other messages and tasks. Estimated delay: one to five minutes.
- The engineer must analyze the alert information before determining the need to alert the service engineer—usually via a collaboration and/or ticketing system. Estimated delay: one to five minutes.
- The alerted engineer, assuming that person is already on-site and not currently working on another issue, acknowledges the assigned ticket and begins the investigation by logging into different devices and capturing information. Estimated delay (assuming no commands are misspelled): 10 to 15 minutes.
- After identifying the cause of the outage, the engineer begins issuing appropriate commands to re-establish service. Estimated delay (assuming no commands are misspelled): two to 15 minutes.
- After restarting the service, the engineer issues more commands to verify that service is re-established. Estimated delay (assuming no commands are misspelled): one to five minutes.
- Last, the engineer closes the service outage by updating the trouble ticket with the appropriate information. Estimated delay: 10 minutes.
The total time to recovery in this scenario is approximately 25 to 55 minutes.
Although both scenarios involve cross-domain technologies and platforms, in the first scenario critical website services are down for less than two minutes. In contrast, manually re-establishing the same critical services could take nearly an hour. In a time-sensitive campaign like this, each minute of downtime represents significant potential revenue loss. Even worse is the incalculable negative impact to the reputation of the business.
The Cross-Domain Automation Advantage
Open, cross-domain automation works by combining workflow actions with sensors installed as inbound integration points designed to watch for specific events from cross-domain technologies and platforms. With this configuration events will trigger the corresponding workflow based upon predefined rules and “If This, Then That” (IFTTT) logic.
Actions are outbound integration points that execute commands on external systems such as the network device or ticketing system in the previous scenarios. When executed via workflows, cross-domain execution is nearly instantaneous with support staff being made aware and required to act only if the workflow automation is unable to rectify the problem.
With this unique, open, and customizable approach, workflows can respond to events and execute actions in a programmatic way on any network device, cross-domain platform, or application. As a result, IT organizations can respond to requests faster and operate more efficiently to help the business thrive.
For more guidance on modern automation strategies, visit our Network World blog page.
To learn more about Brocade’s advanced automation solutions, visit us here.