3 keys to navigating the hybrid data centre

Efficiently managing a scalable and flexible hybrid data centre can be difficult in the age of software-defined resources, particularly when trying to ensure that application Quality of Service (QoS) isn’t compromised.

To guide a hybrid data centre through its entire lifecycle, you need to perform the following three “IT OAR” actions.

1. Optimisation
2. Automation
3. Reporting

Together, these OAR processes will help the IT team navigate the digital transformation era and the burden many IT professionals are beginning to feel.
So, how do you get the most of the IT OAR?

>See also: Understanding data centre workload quality of service

Without experience and expertise, these processes can leave you running around in circles. Let’s explore how OAR can help to ensure successful service integration, delivery, and managed scalability in the hybrid data centre.

1. Optimisation

Optimisation in the virtual data centre comes down to optimising resources—essentially, doing more with less. It spans virtual data centre health across resource utilisation and saturation, while also encompassing resource capacity planning.

Neglecting to optimise often-limited resources usually results in resource over-commitment and eventually contention issues due to the saturated state. If this contention lasts a long time, then errors often follow.

>See also: Enterprise storage in 2017: trends and challenges

Resource optimisation starts with tuning compute (vCPUs), memory (vRAM), network, and storage. It extends to the application and its tuneable properties through the hypervisor to the host and cluster.

vCPU and vRAM penalties manifest in saturation and errors, which lead to slow application performance and tickets. There are definite consequences when oversizing and undersizing virtual machines (VMs). Therefore, the role of optimisation is to find the fine line that factors in and respects the entire data centre environment.

2. Automation

Automation in the virtual data centre spans across workflows such as management actions, setting up profiles and configurations, reflecting best practice in policies, and provisioning or reclaiming virtual resources. It is made up of scripts, templates, and blueprints, which are developed from your best practice methodology.

However, the downside of this is that automation cannot distinguish between good and bad — before you know it, bad IT practices can be automated across the enterprise and lead to a multitude of errors across the entire virtual data centre. For instance, think about the Amazon S3 fiasco, which took down US East. That was caused by an automation script.

>See also: What does 2017 have in store for the data centre industry?

The good news is that there is a way to avoid this: keep it simple. First, automate on a smaller, controlled scale by following the mantra: “Don’t harm your production data centre environment.”

Second, monitor automation processes from start to finish so that you can identify any unexpected executions or errors in the automation workflow. Finally, keep an eye on the results and reporting, and use this data to optimise automation as an iterative process.

3. Reporting

Reporting shouldn’t be overcomplicated. It should show exactly what has happened—your journey in the virtual data centre. This begins with details of the data centre, including metrics and key performance indicators. Then, it should frame the story, backed with data, of how to efficiently and effectively expand the delivery capabilities of the virtual data centre.

Throughout the entire application journey, agility, availability, and scalability need to be at the centre of the virtual data centre. Reporting provides justification for optimisation and automation success. Ultimately, reporting’s goal is to seek a decision that needs to be made—no more, no less.

>See also: Data dependence: the need for a reliable data centre

It’s important to remember that OAR actions are all related. Automation without optimisation will lead to work being carried out without purpose and will therefore not make the most of resources.

Optimisation and automation without reporting will lead to a lack of credit for the work, and won’t help the IT team take the correct course of action going forward.

These three OAR actions will continue to play a role in a scalable hybrid data centre for the near future. The better they are used in conjunction with each other, the better the journey to the hybrid data centre for the IT team and wider organisation. And as the digital transformation presses on for IT professionals, these OAR actions serve as foundational guides for organisations to successfully navigate the change storms.

 

Sourced by Kong Yang, Head Geek™, SolarWinds

Avatar photo

Nick Ismail

Nick Ismail is the editor for Information Age. He has a particular interest in smart technologies, AI and cyber security.