Hybrid cloud round table: Openstack, IBM, Rubrik, Scality, Capgemini

Silicon round table

As the hybrid cloud continues to develop, Silicon spoke with several leading experts to gain their insights into current hybrid cloud trends, the pressure points businesses are feeling, and what the future looks like in this fast-changing space

Jonathan Bryce, Executive Director, OpenStack Foundation. [JB]

Jonathan Bryce, Executive Director, OpenStack Foundation
Jonathan Bryce, Executive Director, OpenStack Foundation

Jonathan Bryce has spent his career building the cloud. He started his career working as a web developer for Rackspace, and during his tenure, he and co-worker Todd Morey had the vision to build a sophisticated web hosting environment where users and businesses alike could turn to design, develop and deploy their ideal web site – all without being responsible for procuring the technology, installing it or making sure it is built to be always available. This vision became The Rackspace Cloud. Since then, he has been a significant driver of OpenStack, the open-source cloud software initiative.

John Easton, IBM Distinguished Engineer, Cloud Engagement Hub. [JE]

John Easton, IBM Distinguished Engineer, Cloud Engagement Hub
John Easton, IBM Distinguished Engineer, Cloud Engagement Hub

John is an executive IT Technologist. Skilled at introducing new and emerging IT technologies to the market and running first of a kind client implementation projects. He has many years’ experience with a wide range of different hardware and software technologies across many different industry sectors.

Mark Shaw, UK and Ireland, ME and Africa Sales Engineering Team Lead, Rubrik. [MS]

Mark Shaw, UK and Ireland, ME and Africa Sales Engineering Team Lead, Rubrik.
Mark Shaw, UK and Ireland, ME and Africa Sales Engineering Team Lead, Rubrik.

Mark Shaw has worked in technology for over 20 years, as an IT manager, and in senior business-focused technology positions. Mark and his team’s goals are to drive real business value by delivering exceptional solutions, support and operational excellence to Rubrik’s customers.

Paul Speciale, Chief Product Officer, Scality. [PS]

Paul Speciale, Chief Product Officer, Scality
Paul Speciale, Chief Product Officer, Scality

Paul leads Product Management for Scality, where he is responsible for defining RING functionality, solutions and roadmaps. Before Scality, he was fortunate to have been part of several new cloud computing and early-stage storage companies, including Appcara, where he was focused on cloud application automation solutions.

Ramanan Ramakrisha, Cloud CoE Lead, Capgemini. [RR]

Ramanan Ramakrisha, Cloud CoE Lead, Capgemini
Ramanan Ramakrisha, Cloud CoE Lead, Capgemini

Ramanan leads the Cloud Centre of Excellence at Capgemini. He has over 25 years of experience in technology practice development, innovation and delivery in enterprise technologies, infrastructure, cloud, AI, Big Data, Agile, DevOps, RPA and IoT. Over his career spanning across India, the USA and Europe at end-clients and IT Service Providers, he has worked on various mission-critical environments on everything legacy, contemporary and emerging. He strives to be a practical strategist in bridging the gap between Enterprise IT and the broader society.

How transformative are hybrid cloud technologies for businesses and organizations using these services?

[JB]
Today, most businesses are still in a multi-cloud model where they make use of multiple cloud environments. Often that includes both a public cloud environment and private cloud environment. However, the larger goal for the hybrid cloud of spanning applications seamlessly across various kinds of clouds hasn’t yet been widely realized.

[MS]
I think it depends on how it is implemented. When the right planning, strategy and the reasons to use the hybrid cloud are precise, then this is transformative. Hybrid cloud deployment enables organizations to be more agile, to utilize services not available to them on-premise, quickly realize value, raise security, manage operational costs and critically, free up IT resources to aid transformation within the organization.

[PS]
There was some thinking among customers to even go ‘all in’ on the public cloud and to eliminate all on-premises IT infrastructure in favour of the public cloud. In 2018 and 2019, we see a reversal of this thinking: after trying the public cloud for some applications and services, many businesses have seen the disadvantages in reduced control, risk of lock-in to a specific vendor’s cloud services and API, and finally the cost of cloud.

There have been ample stories published about the ‘sticker shock’ many customers have had from public cloud vendors. This has resulted in much smarter, wiser customers who now look at the public cloud as a very useful and critical asset – but one that should be utilized selectively, and with a mix of their on-premises applications and underlying compute and data storage infrastructure.

[RR]
The ideal hybrid cloud strategy in today’s world would be the perfect mix of SaaS for non-differentiating business processes, PaaS for any service components that can be consumed as a service from the public cloud and a combination of public cloud IaaS and on-premise IaaS type deployment, all integrated seamlessly with cloud integration technologies.
This warrants an honest evaluation of what workloads need to mandatorily stay on-premise, and then migrate everything else to the public cloud. The ones who are left to run within the customer’s datacentre will need to be redeployed into a more cloud portable manner using Container and other such technologies.

How is the hybrid cloud changing the role of the CIO?

[JB]
It comes down to the different operational models of the public cloud versus running your infrastructure in a private cloud. For your private cloud, you need network engineers and sysadmins who are taking care of the cloud. For public cloud, you need a different kind of sysadmin who is maintaining operating systems that are provisioned automatically. Your staff is now diverse than it was ten years ago, so we often hear that finding talent is a challenge.

[JE]
As organizations embrace a hybrid multi-cloud approach, CIOs must also embrace new ways of working while not breaking what is working well. There is much that traditional IT can learn from the cloud. There is also much that cloud can learn from traditional IT. Bridging this gap and making the hybrid organization function well is the crucial challenge for the modern CIO.

There is one other fundamental change here — that of delivering services, rather than products, to the user. The cloud is a service delivery platform. Rather than consuming a product or application as in the past, you now have integrated management. Data is backed up. Services are delivered to meet SLAs in terms of availability and performance. Services are now expected to be always available and delivered on a consumption-priced basis.

Developers used to delivering code are now moving on to focus on how it will run and be supported. What SLAs need to be met, and how? The adoption of new roles such as SRE (System Reliability Engineer) by the business may help address this. This is; however, yet another organizational and cultural change to be managed by the CIO.

[PS]
The CIO now needs to be very informed about how to use clouds most efficiently for their business. This means understanding what services are available, and in which vendor’s clouds, that can help accelerate their business goals and digital transformation the best. Staying informed about leading services from not only the big hyperscaler cloud vendors (AWS, Azure and Google) but also regional and specialized cloud service vendors is now a primary responsibility.

[RR]
The role of the CIO has progressed from being the Chief Information Officer to one, which is a combination of Chief Integration Officer and Chief Innovation Officer. As the Chief Integration Officer, the role essentially demands identifying the best way to procure services, be it as SaaS or PaaS from public cloud providers and integrating those with on-premise workloads which cannot move for various reasons.

As the Chief Innovation Officer, the job entails a constant scan of the horizon for the latest innovation in the market across all industries and looking for ways to abstract the innovation to apply in their industry. Essentially wearing both hats, the focus is on spotting the opportunities from wherever they are to create a digital ecosystem rather than building everything inhouse.

How can CIOs unlock the value in their hybrid cloud deployments?

[JE]
Simply put, follow the data. Many organizations are struggling to take advantage of the data they already have. The cloud is a critical enabler in providing access to many new tools and techniques to help derive insights and value from data.

The challenge for the CIO is doing this while meeting regulatory and business policy requirements around data. CIOs need to ensure their traditional datasets are of sufficient quality, integrated, secure and respectful of the latest privacy standards. Only then can they be exposed to cloud-based tools. Organizations successfully doing this continue to lead the way in delivering business value from the hybrid cloud.

[MS]
The ‘why’ hybrid is the most crucial question. If you know the reasons specifically for your business and have a clear understanding of the use cases, extracting the most value directly aligns with the company. From a technology perspective, I think technologies that are born in the cloud era that have cloud at their heart makes the transition much more manageable.

Legacy architectures are a burden on IT, so limit the effectiveness and change Hybrid cloud might bring. Rubrik built a web-scale platform from day one, the first product natively integrated with AWS, with an operating model that works like the hybrid cloud. All customers benefit when legacy or ‘technical debt’ is removed.

[PS]
One way is by understanding that leveraging hybrid cloud is a crucial way to unlock the value in content (data), there are a few examples of how to leverage hybrid clouds for data: Replicate data from on-premises to a public cloud for backup and D/R. Replicate data from on-premises to public cloud for compute-intensive processes, where on-demand instances can be used in a ‘bursting’ manner for things like analytics, simulations, transcoding and many other on-demand jobs Bring data ‘back’ from public clouds to an on-premises object store, since the TCO can be reduced by 30-50% compared to long-term retention in services like AWS S3.

How will 5G impact on the development and deployment of hybrid cloud services?

[RR]
5G is based on software-defined networking providing extremely low-latency high bandwidth at low costs and available for access from pretty much anywhere. The next-generation network enhances the business case for hybrid cloud adoption due to reduced latency in bridging between on-premise and public cloud environments and allowing computing to further extend to the edge.
High-speed low-latency connectivity between cloud providers without utilizing physical networks will aid faster adoption of hybrid cloud among all industries. 5G combined with the increased computing power in the IoT, and other Edge devices will also result in much more immersive and data-intensive analytical applications that can use the full breadth of what hybrid cloud has to offer.

[JB]
5G will make it easier to have bits of infrastructure everywhere. That proliferation can create massively distributed IT systems, bringing the benefit of low latency and fast access to computer services close to the end-use case. But it will also bring management challenges from having a more distributed footprint. 5G is expected to drive the distribution of cloud resources and the creation of tens of thousands of miniature clouds that are close to the end-user.

We’ve seen open-source projects like Airship emerge from AT&T to solve this highly distributed cloud environment problem in a declarative way, so that environments can be deployed, replicated, and upgraded with zero human intervention. This approach will be essential when dealing with the complexity of 5G and edge.

[PS]
The bandwidth breakthroughs from 5G will be monumental, especially for the content-rich applications. Today, we speak about the issue of keeping data where it originates due to concerns known as ‘data gravity’ – data has a weight in terms of time and cost to move it around. Today this precludes some use-cases from being practical, such as moving data to the cloud for D/R or processing purposes. With 5G, the time issues with moving petabytes of data from on-premises to public cloud (or vice versa!) will be significantly diminished, opening up practical business uses.

What does the future of the hybrid cloud look like?

[PS]
A significant inflection point affecting the future of hybrid cloud comes from connected edge devices and edge data. The primary industry analyst firms have published research studies that estimate between 25-31 billion connected edge devices will be online by 2021. About 50% of new data will be created by these devices, which include not only our mobile devices, but also smart sensors, appliances, medical scanners, shipping scanners, aeroplanes, cars – you name it, and it will be connected.

This will result in the emergence of a local ‘edge cloud tier’ that services dedicated communities of edge users. For example, a sports stadium or hospital may become the on-premises cloud in a hybrid cloud setting, or even imagine that data stored in a public cloud, needs to be replicated to an ‘edge cloud’ such as a military base, a ship, or an offshore oil well. These scenarios requiring edge access to data will dramatically transform the hybrid cloud from what we know it as today.

[JB]
In the future, some of the open-source approaches to networking like service mesh will mature, allowing applications and workloads to span a variety of clouds. That unlocks a lot of additional potentials that people are not achieving right now with the current state of the hybrid cloud.
The future is a state where the underlying resources for clouds are more interchangeable, bringing choice and flexibility so you can lower costs and get better performance.

Also, if we start distributing compute resources to smaller clouds (i.e. edge computing), we will have more diverse configurations tailored to specifics workload. This future would consist of smaller, more varied combinations of multiple clouds that are local or spread geographically based on the workload.

[JE]
Looking ahead, we can be sure that organizations will look to embrace hybrid models as they focus on shifting their cloud strategies from the low-end infrastructure-as-a-service (IaaS) opportunity to instead extracting valuable data from their business processes; integrating data across the enterprise and with external data sets; and applying new, innovative services such as AI, blockchain and analytics to that data.

Organizations will need hybrid and multi-cloud environments comprised of the right tools and infrastructure to optimize business applications across different contexts, drive efficiencies and launch new services, with levels of speed and cognitive innovation previously impossible.
This, in turn, will drive new organizational constructs and new ways of working for the IT department of the future, which will be the responsibility of CIO’s to manage.

[RR]
The future of the hybrid cloud will be marked by the disappearance of the word hybrid altogether. Due to low-latency connectivity, workload portability and the ability to manage both on-premise and cloud in an identical manner, people will no longer realize they are on hybrid. A parallel can be drawn with the early days of virtualization, and when it was evident if you were working on a virtual or physical server, Nowadays, it’s incomprehensible whether anything is running on a virtual server or not, and this level of osmosis will be replicated with hybrid cloud.