Inside HCL:VDI and more

Pratima H

INDIA: Observing and acknowledging actual needs, gaps and possibilities is just one of the many wise approaches that this organisation’s IT folks wield when it comes to embracing something new and big. After all, they breathe the same air as their brethren, who delivers some really cutting-edge IT to the world, does.

Well, smart people seldom look for the ointment; they look for the flies; lest those unsolicited creatures reach their tea-cups one day. But how? Ashu Kakkar, Associate Vice President, IT, HCL Technologies gives us a peek of that alloy of sweat and magic that condenses strategy into execution smoothly and that matters a lot when scale and stakes are high with something as seemingly forklift-ish and ambitious as VDI.

Why VDI (Virtual Desktop Infrastructure) and not other alternatives? What did you read in a market where options like RDS (Remote Desktop Services) dominate both hopes and worries?

Traditionally we were using desktops and high-end workstations. We also had standard OS requirements. With VDI, provisioning turned to be easy and the matter shrunk from hours to minutes. Turnaround time, too, was pretty fast. For us, security and patching etc. assume different proportions of priority and gravity. So VDI became a reasonable answer.

Also, cost angle was an important one. Although costs were a tad high initially, they are now equivalent to thin-clients, desktops etc. specially as end-point pricing is going down. This option is also compatible with Office 365 and OneDrive. In terms of power consumption, when we consider a 50 per cent gain, this option is particularly attractive. So there were a lot of advantages with VDI.

But why VMware?

We started our journey with VMware early in 2008 with server virtualisation (which moved from zero base to 6o/70 per cent of server workloads as we moved forth). We enabled a lot with VMware including DR, as well as a lot of enterprise workloads and private cloud footprint with underlying VMware technologies. Somewhere around 2013-14 we started with desktop virtualisation. We initially did it with Citrix but due to some considerations and challenges we contemplated new scenarios. Factors like skill-sets, app-volume etc were well-aligned with VMware from a hypervisor point of view and that gave us an edge of sorts. Since with Citrix too, hypervisor was from VMware, the transition proved to be quite smooth.

How was the experience, warts and all?

It was a bit of a bumpy start, specially when it comes to convincing users and handling complex scenarios. There was the part of dealing with different levels of architectures too but we sorted it all out well in due course of time.

Infrastructure and skills –just how tough were those parts?

Initially a little tricky; but now we have got 80 to 85 per cent of the ground all smooth. Some push-backs from project and client-side notwithstanding, we had, overall, a good journey here. We tested all applications and ensured with our teams that all applications were compatible in the new VDI environment. We also had a lot of use-cases, adequate testing, and certifications etc which were shared with all project teams.

Ashu K, HCL

Do protocols and application-diversity bring unexpected complexity somewhere?

Some applications are actually not exactly successful and we are trying to cover those areas. We are looking at solutions from VMware for different kinds of workloads. We are in the process of sorting some stuff out.

Are the results satisfactory?

Yes, with 20,000 plus on VDI in our set-up, we are now scaling up already. Availability metrics, performance etc are very good. We have built in automation inside provisioning and the application control parts are quite strong.

Weren’t there any scale, sprawl or storage issues at this level of usage?

Yes, but we addressed those well. We have invested in good storage options with EMC and also took care of issues like boot-storms. We are quite in control. The centralisation is region-wise and we have planned the set-up across various regions; and now we are looking at hyper-converged infrastructure.

Ok, so how else will this move and grow further?

We are taking VDI deeply with BYOD (Bring Your Own Device) and work-from-home formats. There’s MDM integration next as well as public cloud possibility. We are already doing some POCs (Proofs-of-Concepts) there.

What else can be done to amplify the outcomes and your needs with VDI? AI?

We may move into more automation, DevOps and borderless ODCs (Offshore-Development-Centres) soon. Static machines would not be mapped down to a project to a large extent then. That will bring in a lot of transparency and security.

Anything else brave and huge that you would be considering from an internal IT perspective?

We are serious on Cloud-first and Mobile-first journeys. Also, new campuses would be leveraging IoT (Internet of Things) and digital transformation in a major way. From Office 365 roll-outs, giving a digital experience to delivery teams and users, augmenting mobility levers; we are accelerating the whole digital wheel side. That said, security would stay paramount to all our investments and strategies.

For lack of a better term, dog-fooding - does that happen at HCL? Do you try out new stuff internally first before bringing it out for clients?

Yes. We deploy a lot of automation tools infrastructure etc for users and that covers self-healing, automation-back-end tools etc. There is a high focus on automation. There are quite a few POCs that we undertake before delivery teams take something to customers.