On 14 April 2016 Twitter went down for around one hour. The down-time coincided with the start of basketball superstar Kobe Bryant’s final game with the LA Lakers. Fans and the club both posted quips afterwards about how his popularity and the amount of Twitter activity on this historic day ‘broke’ Twitter.
Of course this was speculation only and was not verified by Twitter. It is however, a somewhat light-hearted example of what happens when enterprise technology platforms – and the applications they support – go down . Many people notice. Rarely in their millions as is the case for Twitter but, depending on what application is affected, the impact could be much greater and much more costly.
Yet not every enterprise platform is truly enterprise ready
Enterprise IT leaders have responsibility for several different technology platforms – from legacy server farms to cloud-based insight tools and global web caches, among others. These platforms can support 100’s of applications, servicing many 1000’s of users. And yet, despite best efforts, the reality is that not every platform and system in such an environment is operating at the ‘enterprise level’ that’s required.
Such scenarios are much more likely in a fast-growing enterprise where the technology investment has not kept up with the rate of business growth, or in particularly large organisations where ‘rogue’ solutions exist on the periphery – developed and implemented by departments that have chosen to by-pass the IT team. Either way, these scenarios represent a significant risk to the organisation – whether it be to reputation, to bottom line results or even to health and safety of employees and clients
Remember – underlying platforms such as your databases or your server instances in the cloud are like the foundations of a skyscraper. If they fail, everything above fails.
Are all your enterprise platforms truly enterprise ready? Consider the following:
Do they all have enterprise level support? It happens. Even in an enterprise setting, certain platforms are operating without formal support arrangements from software vendors and without dedicated support teams . As a result, it is likely that outages are more frequent and that it is the application users themselves who are the first to be aware of a problem, taking the business by surprise.
How confident are you in the outage recovery process for each platform? Having a clear recovery process in place is critical – particularly if the platform down-time was identified by users and the business is already on the back-foot. A reactive recovery process is likely to take longer, involve more resources and ultimately cost more – both in terms of direct recovery costs, the potential loss of business and/or other opportunity costs.
What is the current change management process? Managing change in an enterprise setting is complex and time consuming. Whether it is proactive software patching or the implementation of new functionality, you can’t just do it on a whim. The change management process for each platform needs to be based on the needs of its users, including factors such as down-time tolerance, and peak and low usage times. DevOps is making inroads in supporting greater agility, quality and speed of change implementation, but large complex systems still need careful planning for change.
Bringing all your platforms up to standard
The challenge for the enterprise CTO or COO is to convince the rest of the business of the need to invest the time and the resources to bring all the technology environments up to an enterprise standard. Often at the heart of this is establishing a common appreciation of the risks and balancing this out against any potential budgetary concerns.
In a future article, we explore the true cost of platform downtime – or the false economies of business units trying to save 10’s or 100’s of thousands of dollars in ongoing support costs when one major down-time event can cost ten times that amount.
In the meantime:
Conduct an audit of the technology platforms in your operation that are operating at a sub-standard level. You might find this article a useful reference to some of the elements that you should consider – this includes a review of the original plans and objectives, the design, monitoring set-up and the supporting SLA levels.
Prioritise your list considering the potential impact of an outage event. Which of the above platforms host critical applications and/or those that are relied upon but the greatest number of users? Where is there the biggest gap between the needs of the business and the design implemented to support these?
Implement a program of work and a schedule to review the platform structure and processes with individual business owners.
Regain control of your platforms and applications
Managing the technology environment in an enterprise setting is an increasingly complex task. It is a fine balance between driving innovation, offering new services, new ways of working and new revenue streams, while ensuring optimum uptake and uptime for existing platforms and applications. Having to deal with existing platforms that are not enterprise ready further amplifies the challenge – and the potential risks.
If you need advice on how to bring the management of your key business platforms to an enterprise standard – do get in touch. We specialise in the development and management of technology platforms in the enterprise space – particularly data management, integration and cloud services platforms.
Comments