Getting onboard the automation bandwagon
(Above pic L-R): Kingsley Wood and Richard Koh
When the Director of Red Hat’s Infrastructure Business Group in APAC, Kingsley Wood, was in KL recently, he shared about the levels of automation that could typically exist in an IT environment.
Forget what you may have seen in the movies, automation can be about something as simple as sending instructions to us, and then us manually executing it. For example, Waze. Instead of struggling with paper or online maps to know the route to take, we key in our destination, and the mobile app does everything, except drive us there.
Organisations in Malaysia, are still familiar with doing things themselves, however. Woods said, “Automation requires specialised skills, and the rest will astay away.” That said, he does believe that with automation tools like Red Hat’s Ansible, individuals can start using automation much more easily.
The next level of automation according to Woods, is when data is collected, insights are derived from it, and humans do the rest. So for example, the system is able to react by itself. “A corporate IT set up should be able to handle the worst case scenario. So, systems should be elastic and able to instruct cloud services to scale up or down (depending on demand for resources),” Wood described.
The third level, is when automation should be able to warn when something is about to go wrong. Self-healing networks which fix themselves, is an example of this kind of automation.
The thing about Ansible is that there is less focus on script, according to Woods, and instead there is a common set of rules that can be used across the organisation, regardless of boundaries and siloes that are usually created by departments.
“It used to be that individuals do tasks manually, or find tools to help automate it, within their own domain. So, the process isn’t integrated, there is a lot of lapse between tasks that are automated and tasks that are not. This introduced a higher chance of error.” Woods pointed out.
Why automation helps?
Red Hat’s Director of Enterprise Sales and Strategic Business Development, Richard Koh said, “With limited amount of resources, IT is still expected to maintain operations and incorporate changes to deliver new services, as well as also firefight.
“The amount of work that goes on to the operations team is massive. To also become agile and innovate is almost impossible.”
The proposal is that automation can take over manual and repetitive tasks. When popular ride-hailing service, Grab automated mobile app development, Koh said 20 minutes to repair systems reduced to 20 seconds.
Besides the purported increase in reliability and speed, there is also the allure of having an end-to-end view of a whole automated process. Koh said, “You can have good understanding of where resources are, what and where went wrong, and fix it. Governance becomes high this way.”
Automation use cases
Wood pointed out two other scenarios that would benefit from automation – during DDoS attacks and when managing IoT sensors.
He explained that when a Distributed Denial of Service (DDoS) attack is detected, Ansible would shut down certain network equipment. “Automation helps pull different pieces of information together faster, so the system can react faster. Ansible automatically takes all the remediation steps, via APIs that enable it to talk to different systems.
In the case of IoT, Ansible could identify, check the fix and with the push of a button, deploy updates to thousands of endpoint devices.
According to Wood, Ansible can help automate reaction to inventory changes, for example if something extra has been plugged in or removed.
Koh concluded, “We have seen partners come up with solutions that can benefit from automation, that we never thought of before.
“Down the road, we will discover more use cases that can he helped by automation.”