VMs and Containers: Ironman and Spiderman

And by that analogy, virtual machines and hyperscalers could be the Marvel and DC Universes. But what if these heroes and universes joined hands instead of being busy with their own super-powers and battles? Are they? We ask Guru Venkatachalam, Vice President and Chief Technology Officer, Asia-Pacific & Japan, VMware about some recent and future super-forces that are shaping this world

Hardware-assisted virtualisation, hardware virtualisation and bare metal virtualisation. Are they similar or not? Tomaahtoes and Tomayyytoes?

Virtualisation as a concept of abstraction of resources is the same. People can call it various things. On our part, we are excited about DPUs – and it’s the place where the next-generation of data centres will evolve towards. Decades back, sharing compute and handling specific workloads were problems to contend with. Then GPUs came along. A similar trend is being seen with DPUs. CPU, as a premium resource, can be managed very well when a lot of stuff can be redirected to DPU. It requires apt hardware and chip-level support. The next generation of data centres will see a lot of demand on this front.

Can containers and virtual machines coexist?

In 2019 we announced a specific architecture and solution. People in IT were used to watching VM workloads. Developers were used to watching containers as the side they see the most. We brought in an apt solution. We also brought in Kubernetes. Since the announcement, both sides have been good, and this move has resonated very well.

What about all the complexity that comes along with such co-existences? Like cloud-native and other ‘chaos’ environments? Is that what you are trying to address with ‘cloud smart’ as an approach?

Cloud-native environments cannot happen with the strike of one button. There are some customers that still run on mainframes, there are some that depend on ISVs and so on. That’s exactly where VMware comes in. We can enable all

kinds of environments. We provide the necessary guardrails for the cloud. We continue to optimize IT with the power of ‘and’. With the right management frameworks, and appropriate governance models, complexity can be addressed—and very confidently. Be it for legacy environments or cloud environments – it can all be simplified.

Has the issue of memory management been addressed well?

The boundaries for specific set of applications will keep stretching. We have been trying our best as we get challenged with new use-cases. It’s a consistent effort.

What’s the equation between virtualisation players and hyper-scalers?

Conceptually, virtualization enables Cloud. It’s about sharing resources. We come in where we help enterprises with their various workload-issues. This can happen with a public cloud stack or a tech stack for own Cloud. We are doing partnerships with Hyper-scalers so that customers find answers for all kinds of workloads. It’s an interesting equation—of both co-operation and co-op-tition.

If we go back to Popek and Goldberg’s principles of what defines a VM – being identical, being in control and being efficient- have these factors become challenging to juggle today?

I used to be on the customer side some years back. It always comes down to management. It’s not about technology but how you manage it. Once technology is seen as something good, it gets deployed, and with that roll-out moment, complexity starts to take shape. People start facing issues like scale and speed. Questions like ‘is it resilient’ become paramount. Specially in the last two to three years where everyone needs a stable, secure, 24/7-on and resilient IT.

So how can performance, availability and sprawl be balanced well? Is over-provisioning a CIO’s problem or a vendor’s problem?

The conflict is in the mind. It’s not easy, of course. It’s a complex problem. It goes back to the formula of being stable, secure and scalable. Cost comes in as an added parameter. That’s where it can turn into a CIO’s job. Technology needs proper frameworks and governance to deliver the results we expect it to. It all comes back to priorities.

What would you say to CIOs?

We are in the midst of a lot of new and exciting opportunities. But prioritising and defining focus can become a challenge to reckon with. It’s not easy, but we need to opt for the right management tools in place before we opt for big endeavours like a multi-cloud journey. ‘Business driving projects’ always wins over ‘technology driving projects’. A good technology endeavour is a big chance to reengineer one’s processes and operating models. You cannot be doing the same things the same way. Technology is an opportunity to strike a change.

Any comments on the Broadcom-VMWare deal buzz from a technology lens?

Broadcom is giving all the support to our vision. Our three pillars – multi-cloud, application modernisation and end-user computing are well-supported here. The exact integration of tech pillars is something too early to comment on, but they are in alignment with our vision.

And what are you excited about the most when you look ahead?

Scale. Be it BFSI, manufacturing or government – everybody is going to be keen to leverage features and functions from different clouds. Every aspect of every industry is going through digitisation. And virtualisation is one important piece of that puzzle. It will be all about scale and cost-optimisation for various use-cases.

Guru Venkatachalam

Vice President and Chief Technology Officer, Asia-Pacific & Japan, VMware

By Pratima H

pratimah@cybermedia.co.in