Risk is something that all businesses are faced with and almost all elements of business operations create some level of risk. Often however, many risks will have a low level of probability (likelihood of something happening) and limited potential impact on the business and so are mitigated relatively easily enabling operations to continue. But what about business risks where the probability and impact levels are high such as those in a data centre relocation.
Risk in Data Centre Relocation
“Without proper planning and processes, data centre migrations pose huge financial and operational risks to organisations. Understanding the primary steps required in a migration project is key to identifying what you don’t know and what resources will be required to prepare your project for success.” Gartner 2015.
Knowing how to identify the key risks when relocating a data centre starts by asking some questions at both a business level: How will business critical applications and services be affected?, What impact will a move have on service levels? and at a network level: Have we considered bandwidth, jittery links, increased latency and fluctuating service quality?
By asking these questions you start to quickly identify what you do and don’t know and consequently what you need to be able to find out and test before embarking on a relocation project.
To put into context the impact of some of the network level considerations and the potential associated risk, we considered latency.
A spotlight on Latency
How much time it takes for a packet of data to get from one designated point to another.
Network related delays can be changed dramatically when a data centre is moved. Geographical distance contributes significantly to the network latency between the client and server. Just moving the server physically further away from the user can double the latency. However it won’t just slow down operations, it will also reduce the throughput of applications further impacting the end users. To put into context the relationship between network delay and an application response time is far from one-to-one.
The following example provides a relatively simple overview of the relationship between network latency and application latency. For a local user with one millisecond of network delay between the client and the server, 150 transactions would be completed in 0.15 seconds. When just 50 milliseconds of network delay is introduced (representing a typical cross-country WAN connection), the performance of these transactions did not slow down by only 50 milliseconds (0.05 seconds). Instead, the same 150 transactions took 7.5 seconds.
This illustrates how small changes in network latency result in major problems with application performance. In a complex project such as a data centre move, the challenge of assessing the network impact on application performance becomes one of large scale as hundreds of servers hosting thousands of applications need to be tested to understand the full impact of the data centre move.
So what can be done to mitigate the risk?
There are many issues that can arise during a data centre move which can threaten business operations. Many of these issues cannot be predicted by organisations that operate in traditional IT silos, because they’re caused by subtle, complex interactions between networks, servers and applications. In fact, after these issues manifest themselves in the post-move production environment, many are left struggling to understand and resolve them.
The solution to these problems is to carry out testing and gain an understanding of where the biggest risks are pre-move. These systems must be rigorously tested under the exact network conditions under which they must perform – before they are deployed to those that rely on them. There are a variety of solutions available to carry out testing but when it comes to data centre relocation, choosing the right option is critical to business success.
Take a look at our infographic showing the testing and delivery strategy options available for data centre relocation. and how risks can be mitigated by introducing Network Emulation.