‘Business Ready’ IT Infrastructure

Chandra Mouleswaran S,
VP- Infrastructure Management 


Hariharan Madhavan,
Manager-Information Security


One thing is common across baseball, soccer and the business – the technology that is deployed to analyze the past and future of these sports and businesses, keeps evolving and has the stakeholders on the run always. The advanced metrics of ‘swing’ in baseball, the goal line technology in soccer, hot spot in cricket and the artificial intelligence in business are the newest in the market today, but will soon be buried with something newer. While the coaches and CIOs welcome the barrage of new technologies on one hand, they realize that the internal process and structure are not agile enough to adopt them as quick as they come and dump as soon as they get old.

‘Time to Market’ a new technology or tool across the organization is much higher than ‘time to invent’. It is important to have the base IT infrastructure ‘business ready’ all the time. CIOs are expected to build the IT infrastructure that can handle the new technologies and products that business may need tomorrow.

This article is written to help the CIO/CTOs with few directions and recommendations to build a ‘future proof’ and ‘business ready’ IT infrastructure. The recommendations are as detailed below:

1. Compute, Storage, Backup & DR:

Most of the organizations are not able to keep pace with the speed of the demands for additional compute & storage resources. Even if the organizations have enough budget and the wisdom to buy the upgrades at the right time, the efforts required to upgrade and the risks involved while upgrading are higher. The process of maintaining and managing the support contracts for the compute & Storage hardware with the vendors are even more complex. The backup of the data is even tougher due to the huge maintenance required in adhering to backup schedule, looking into each backup failure, sample restoration checks etc. Disaster Recovery does not exist in some, does not work in many organizations. The cost and efforts required to maintain and manage these resources are very high and often not defect free. The best and affordable solution for all the woes of compute and storage is migrating he entire workload to cloud which has abundant capacity to scale up or down, built in HA, backup and cost effective DR.

There are Cloud migration tools available to migrate the ‘on premises’ applications with ‘zero down time, zero data loss’ feature. There are cloud orchestration tools available to provision and de-provision VMs for running just OS or a stack of software along with OS with or without HA, in one single click. These tools also provide ‘Environment as a Service (EaaS)’ using which every HW and SW configurations used in setting up VMs are stored as meta data and they could be used to recreate those VMs at any point of time.

2. Network:

The availability of network may not be a problem in these days for most of the organizations. But the ‘reliability’ of those networks are still a concern. The HA feature of networks is not intelligent enough to detect the intermittent issues like errors, packet loss, latency etc. in one circuit and switch to other circuit. The standby gets kicked only when the primary circuit is fully down. The unpredictable Bandwidth consumption is another problem that Network managers need to cope up with. The problems of maintenance, management and standardization more specifically for organizations having multiple sites are large enough to keep the entire network team fire fight on transactional issues every day. The virtualization of Network like the way servers, storage and desktops are virtualized, is the solution going forward.

Network Virtualization (also called Software Defined Network) consolidates all the software and configurations of switching and routing functionalities of Network devices at a central server. This enables control of all network hardware installed at various locations in the offices / remote offices from the central server. We may have to replace few devices that are not compatible with SDN, but the spend will be worth the efforts that it would save and the qualitative benefits it would bring.

3. Telecom:

‘On Premises’ hosted voice solutions lacks agility due to heavy investment made on PBX. Further, integration of voice systems with Office automation tools like MS Outlook for the better collaboration is a complex one and needs lot of efforts from experts. Maintaining and managing an ‘on premises’ comes with its own complexity in managing the circuits and PBX systems. Building a HA and DR solutions for voice systems are cost prohibitive. So the best solution is to move the PBX functionality also into cloud with a tight integration with MS Outlook or whatever the collaboration tool is.

The cloud based voice solution integrated with SD WAN guarantees best performance and consistent user experience for all users including remote and mobile users. The burden of capex investment and the associated pressure of choosing the right and future proof solution goes away. The Total cost of ownership for an opex based, cloud centric voice solution with full scalability, HA and DR will be much lesser than that of an ‘on premises’ based solution.

4. End User Computing:

After the virtualizations of server, storage and Network, desktop cannot be left. The benefits of desktop virtualization are known to everyone and it is proven. It brings in standardization, improves information security, enables desktop software license compliance and reduces the incidents. Since the entire workload is moved to cloud, the VDI infrastructure should also be hosted in the same cloud at same location.

Automation of any non-interactive tasks like restarting a service, initiating another job after completion of previous job etc. are easy ones, but the difficult ones are the non-interactive tasks where the user has to give some input for the automation to kick off. The users are used to certain kind of forms or methods to initiate non-interactive tasks like creating an ID for a new employee or application of leave, but the automation requires a different form or method to initiate the automation. The CIOs do not want their users to go through any change due to perceived discomfort to the users. That’s where there is a need to come up with a solution that enables automation as well as enhance the user experience.

Digital Assistants (also called ‘personal assistants or ‘Virtual assistants’) takes instructions from humans orally and executes a task or initiate a workflow. The users instead of writing a mail or filling up a form, can just speak to a DA like “can you please create an ID for James as the first name, Samuel his last name, DB COE team, Victor as his manager, full time employee, default access to filers…”. The DA will initiate the work flow, send the request to the manager for the approval, follow up, remind and get the activity done. DAs are going to make the life lot easier for employees and managers and enhance the user experience by 10x.

5. Data Framework:

The growing habit of introversion among literates, aided by the comfort and convenience of smartphones, helped not only increase the volume of ‘unstructured data’, but also the true value of those data. Not only human, but also the machines (servers, storage, network appliances) join the race to generate unstructured data. In the years to come, the rate of growth of unstructured data is going to outpace the structured data. An open standard framework customized to the needs of an organization can consume any data and any dynamics at the data sources effortlessly is in the future.

NoSQL is preferred to RDBMS for enterprise wide storage of unstructured data due to its superiority in flexibility and performance though latency is relatively higher in NoSQL. The design principle should be that any information or data that needs to be notified / viewed as part of an information or warning or alert or report or dashboard, needs to be routed thru single data collection framework as it is capable of reading, indexing and storing any structured unstructured data from any sources of any format. The best cost effective solution would be to deploy a framework built on open source tools and NoSQL database, on the cloud.

6. Information Security:

Managing information security on a cloud platform is little different from the way its managed on-premise. An insecurely configured cloud setup will render the organization more vulnerable than on a traditional datacenter. Devops for Infrastructure has quickened infrastructure provisioning, but enforcing security in Devops environment needs more Devops knowledge for the information security professionals in the team. With Antivirus now become just an additional control, application whitelisting and privileged access management has become more of a norm in today’s endpoint environment. Having piles of classified data on repositories may not be new for organizations but in the event of a data breach the implications are more severe than before. Companies need to migrate from a checklist based approach to a practical review of their risks.

Key directions to consider for ‘business ready’ security:

  • Leverage existing cloud vendor’s cloud security tools to assess and baseline security and identify gaps.
  • Leverage VDI to quickly recover from endpoint infections and better patched endpoint infrastructure.
  • Automation of security incident remediation to expedite first level incident remediation.
  • Integrate existing log management and SIEM with other tools to aid automation and information sharing.
  • Strengthen Security awareness still continues to be the best defense against bad threat actors.
  • Extend Vulnerability management from just workstations and servers to cameras, IOT’s, printers and any SMART devices which have access to sensitive data or in critical environment.

7. Incident Free Operations:

The systems that are getting built for tomorrow needs to be monitored and managed by tools that would tell us what would happen tomorrow. The ‘business ready’ infrastructure should be smart and intelligent enough to predict the events that would happen in future. The IT organization should be working for the problems that would come next day or week and drive the organization towards ‘Zero Incident Enterprise™’.

The tools or the framework that would drive the organization towards ‘Zero incidents’ consist of a predictive engine that uses Machine Learning, Artificial Intelligence and NLP, an Instrumentation engine to diagnose the health of the application in production environment in real time, A Real User Monitoring system to capture the users experience while accessing the applications, IT process and robotic automation platform, Virtual end user computing platform, Cloud orchestration cum migration platform.

GAVS Technologies is the one and only proponent of Zero Incident Enterprise™ through its Zero Incident Framework™ which consist of tools and platforms detailed above.

This article is reproduced from GAVS’ enGAge magazine, Mar 2018 edition.