The age of automation is well and truly upon us. We already have autonomous cars, but what about autonomous data centres? Like self-driving vehicles, the data centre that runs itself, manages itself and calls for help when needed is closer than we think. But are we there yet?
As data environments scale in both size and complexity, many organisations are finding that varying levels of autonomy are necessary just to keep basic services afloat, making these self-sufficient facilities not just a helpful convenience, but a necessity.
Key areas for data centre automation
By definition, data centre automation enables the automated and unattended execution of critical workflows and processes on servers and other data centre equipment, including:
- Scheduling routine processes such as backups, replication, downloads/uploads, application events, and other items that previously required manual effort.
- Monitoring the status of components and automatically alerting key responders when a problem occurs.
- Maintenance functions such as patching and updating equipment.
- Application service delivery within minutes of a user request, fulfilling approved user requests without manual intervention.
- Delivery of additional application workload on demand, such as additional Web servers, app servers, or load balancing capability, or for automatically moving network traffic to a data centre with spare capacity.
- Helping to ensure compliance with corporate policies and external regulations.
Benefits of autonomous data centres
On the whole, data centre automation provides efficiencies in faster service delivery, reducing the IT time needed for mundane tasks, thusly freeing up IT staff to concentrate on more strategic projects. Automated functions also provide consistency, being performed exactly the same way each time, reducing IT costs and increasing service quality.
Decreased staffing needs
Larger companies can increase profitability by reducing their workforce through automation. New business startups and entrepreneurs can use automated systems to fill roles that would have otherwise gone unfulfilled. Decreased staffing needs subsequently lowers the cost of doing business too. With fewer staff members to keep on the payroll and the ability to assign remaining employees to more-meaningful tasks, the potential for reducing day-to-day costs are tremendous.
Reduced human error
It’s a fact; humans make mistakes. Even the most skilled and diligent workers are bound to trip up every once in a while. Depending on the scenario, errors could be costly to business. An automated system will only make a mistake if it’s been incorrectly programmed (by a human) in the first place. A system that’s been coded with 100% accuracy will complete its assignments to perfection every single time.
Skilled staff members regularly adapt to new situations and projects on the fly, but going from one task to another still takes time. Robots and AI-driven systems can multitask, assume new activities and halt completed operations much faster than their human counterparts and with minimal worker intervention.
Increased data security
Automated security systems can minimise the impact of a data breach or similar incident. Currently, manual systems require a human response and solution to patch the problem. This process can take hours, days or even weeks. Automated systems address such issues immediately and work in real time to fix security holes.
Disadvantages to consider
Less opportunity for employment
This directly opposes the much cited ‘less need for staff’ plus point, but does automation mean jobs are under threat? As a result, IT professionals may well face a tougher time finding opportunities in their area of expertise. This reality is especially true for those who are trying to start their career with a local data centre, as robots are taking over many entry-level jobs. That said, humans can work alongside these technologies. Yes, some menial tasks will indeed fall by the wayside, but roles in programming, maintenance and supervision will always be available for those who specialise in AI, robotics and similar technologies.
Lower consumer confidence
Think back to the last time you had to make a complaint over the phone, the barrage of key pressing, the automated options, the failed voice recognition. In a word, frustrating. Although automation is catching on in all areas, including among mainstream consumers, this mindset is generally echoed when it comes to automation; we just don’t like the idea of putting all our faith a machine. However, over the coming years this may well become less of an issue, as technology improves, and younger generations become more prominent. But as it stands, a lack of consumer confidence still remains a considerable roadblock to full-scale data centre automation.
Possible system failure
Although a seamlessly programmed machine is highly unlikely to make an operational error, it is far more likely to malfunction than its human counterpart. It might be years or decades down the road, and some might never experience a serious hardware failure at all. But those who do will find themselves relying on their human staff - and human response time - to put the system back online. This point alone underscores the fact that a fully automated data centre is at this stage, still an abstract concept. Humans will always be needed, at least for now, to monitor and manage these systems.
Only fools rush in
Despite a fully automated data centre not quite being feasible, automation is still a necessity required to stay ahead of the curve, and the benefits generally seem to outweigh the pitfalls. That said, IT professionals might be tempted to automate everything within their data centre. However, the urge to automate everything on an ad hoc basis should be resisted. Ad hoc implementations can be dangerous, simply due to the fact they can introduce risk into a data centre, which can result in expensive mistakes.
Automating intelligently via an incremental approach is key. Performing selective automation only on the processes where substantial business value can be realised, will ensure maximum benefit and minimal risk. Going forward, it seems that the biggest question for IT executives is not whether to embrace autonomy, but by how much.