eWeekChat

A3: Cost is the biggest challenge on any single cloud and gets exponential dealing across multiple clouds. Leveraging technologies proprietary to a cloud makes it very difficult to go multi-cloud. Though the capability gap has minimized not all clouds are still equal.

I'd say Cost is the effect - the challenges are mostly around either speed of operations, the NxM matrix of capabilities that someone needs to dig into, skill-set gap especially in Enterprise.

A3 Cost is something to be managed, whether on one cloud or multiple clouds. Most organizations confront higher-than-expected bills when their traditionally-designed apps confront a hosting environment designed for cloud-native architectures.

A3: Increased architectural complexity, disaster recovery planning, and security policies are just a few #multicloud challenges. #CostOptimization is absolutely a key challenge especially when cloud resources are over-provisioned, leading to wasted operational spend.

(edited)

Early to cloud customers are trying to get those apps running - so they don't feel the pain of cost as much. Minute you go past 20-30 VPCs etc - the cost suddenly shoots to the forefront.

A3: "is it worth it" This is a hard question. There is value no doubt, and often a lot of it. But is it right for you? Cost is part of it, but the opportunity cost is another. Yes it might be cheaper in my bill, but is it cheaper in my time?

A3: cost is a big one, especially with cloud application sprawl inside the organization but security is also a big one

That said - we've seen enough enterprises focus on their careabouts well. If they care about speed of operations, and getting apps to deliver on their digital experience, then cost is the price you pay for that... given current macro climate, it certainly is a big factor

@bmcsoftware agree on the complexity of the architecture, requires lots of due diligence and planning

@arnlopez - completely agree. Cost management cannot be an afterthought. Need to cost mgmt as part of design.

Q3: Cost and parity of features for products and services. e.g (1) Take Serverless, or Cloud Functions. the Dev Frameworks and tooling are all not at parity, (2) the CI/CD integrations are not consistent, (3) e.g Marketplace options have different integration models / providers

@bernardgolden Seems like there is a spectrum of responses to this that vary over the timeline of a cloud migration. Early on, the response might be "this too shall pass" but after multiple quarters of unexpectedly higher costs the patience begins to wear thin.

@bernardgolden Good point - if you're not Cloud Native, cost is likely to be an issue, regardless of multi-cloud or not

@bmcsoftware agree, we hear a lot about how Medium and small size enterprises do not want multi-cloud because of the complexity, For multi-CDN/Edge complexity outweighs the gains from reliability if one provider can offer performance and reliability the customer needs.

A4 Also understand that running an app across multiple clouds requires work beyond the app itself -- cloud infrastructure, use/rejection of cloud-specific services that require mapping/encapsulation (e.g., SQS from AWS), and storage.

@bernardgolden Plan to address all of these areas as part of an app-centric multi-cloud initiative.

@bernardgolden If the business goal is resilience, is running the app on multiple clouds really the "best" way to accomplish that outcome? Are you managing the risk of a major cloud provider losing a region or something less dire? Can you really afford to mitigate against that?

The biggest issue we see is data gravity locking into particular cloud providers. This gravity is often created before an application scales and leads to unwanted spread across clouds.

@mdcarlson Not always. this is a question of size of the enterprise as well. An S small to medium size company can get most benefits with one providers , at least in Edge Cloud Space.

Another major challenge is security. Clouds bring appealing modern identity and role-based security concepts but they break down outside of the context of a single cloud, leaving SecOps teams to solve security challenges with rudimentary security tools at cloud scale.

Q2: (1) Digital Transformation, (2) Leveraging the power of Software and (3) Developer ecosystem

@iLakshmiS What about the dev ecosystem? How is that driving multicloud?

A2: a mix of things from my perspective - avoiding vendor lock-in, need for features that a specific cloud provider excels at, geo needs, etc.

A2: Cloud competition is catching up. The "best" cloud is not as clear of an answer so putting everything in one place isnt as simple as it was a few years ago. "our software is a service AND multi-cloud" was THE Re:invent pitch from vendors.

A2: Containerization of workloads lends itself to ease of deployment in any k8s environment like GKE,EKS or Tanzu and is a key driver to growth of multi-cloud apps. Enterprises are derisking from being locked in to one cloud and are using multicloud for DR,edge and data residency

A2. Prior trends continue - industry cloud, workload specific choices (e.g. infra heavy vs analytics etc). Now increasingly cost leverage is a factor...and I think we'll see more of this in the coming months.

@codyhosterman Interesting.

A2: mix for sure but hybrid work is still very top of mind for CIOs, cloud app usage thru the roof, hard to manage and secure

A2 the ongoing shift from on-prem to cloud computing drives overall adoption and inevitably, in large enterprises, results in deliberate or inadvertent use of multiple clouds.

A2: It comes down to services that can offer speed + efficiency at optimum cost. We’ll also see data sovereignty issues becoming important & cloud providers differentiating based on geographical support.  Having options & flexibility for is key.

@ramsba - fully agree. Cost leverage is a big factor. More commitments to a single cloud means more discounts - but less choices and less efficiency. It is a balance.

i thought so! I visited a lot of booths and got the same conceptual pitch over and over. Multi-cloud was not uttered at reinvent a few years ago at all

Apps and data, as we know lead the way. As much as I'd love to say infrastructure choices (e.g. cloud networking, security) weigh in - its generally the apps and data that rule the choices

Q2: Digital Transformation: COVID has accelerated the move to Digital to every country and every industry even Cities and Public sector.

@jwittich hear the same thing, avoid vendor lock in or dependance, more flexibility

@BMCSoftware Good one - many Enterprises grappling with laws, regulations and data sovereignty. This is one that certainly comes to the forefront as it becomes a make / break decision.

@arnlopez @jwittich But can vendor lock-in even be avoided in multicloud? Once you're set up, hard to move.

speed of innovation also requires elastic use of the cloud, have to move fast to stay ahead of the competition

It is about Digital Transformation and Digital transformation is about Applications on the Internet . The differentiation comes by building the best user experience of the app or the website you have to connect with your users. This is lead by developers today.

- hard but doable. Needs to be built ground up. Cannot be an after thought.

@iLakshmiS Got it, makes sense

@BMCSoftware Yes, speed / timer to market the services and also the speed to build and deploy those services

Q1 yes, those who have adopted multi-cloud (let us call public cloud providers, are seeing the gains. However, i would not it at maturity yet.

A1. "Chaotic". Enterprises are grappling with choices. Choices seem to be (a) choose the right cloud for the right workload - great for apps, not so much for infrastructure or operations

(b) go dominantly on one cloud and build apps, infra, tooling for that ... the secondary cloud suffers ... and a few other choices in between.

@ramsba That's what I hear!

A1: Taking a ‘horses for courses’ approach to deploy workloads across multiple clouds is a reality today. The multi-cloud market has grown out of infancy but is not fully mature yet in terms of consistency of security, cost and an overall cloud experience with interoperability.

Q1, its a work in progress, challenge to keep up with so many services moving to IAAS, PAAS and SAAS, requires multicloud arch

A1: For the most part, I think organizations are pleased with their #cloud adoption. With the competitive pressure orgs of all types are under, delivering services with speed, efficiency & now cost is even more important. There is a greater need for cost optimization.

A1: I think overall we are making progress on this as an industry and weve learned a lot. Multi-cloud doesnt mean move things between clouds daily. But the right cloud/service for the use case, while also derisking cloud choice with 3rd party services being available consistently

@codyhosterman In fact things are not moved between clouds, true? Or not much.

@GaneshJKRam Good one. Interesting times - with some still dipping their toes, while some have embraced whole heartedly (and admittedly seeing the pains associated)

Q1: Multi-cloud includes an extension of ecosystem with Devtooling to build / deploy applications that works across all clouds , that is an area that still needs a lot of work

A1: There is a shift in mindset from annual and periodic #capex optimization to one of continuous #opex optimization – the cloud providers have the resources you need,  there is a minor issue of having to pay for it!! :)

@ramsba Agree, different cloud providers are chosen for specific use cases. and hence companies that help with multi-cloud deployments and also standards or frameworks such as Infra as a Code through Terraform and software that can deploy that simply / faster are important

A1: We're seeing some interesting multi-cloud deployments where latency is a major concern and workloads are running across both large central data centers and more dispersed regional runs, operating by different cloud providers

I think there's a definition of multi-cloud thats important to articulate here. Is it independent parallel clouds (with diff workloads in each and not inter-related) or is it different workloads all talking to each other across clouds. The former is manageable but latter is mess

@jwittich Seems like latency will be a nagging problem for multicloud

It happens of course (migrations, DR, etc) but the concept of things truly bouncing around regularly is just not a reality.

A1: Customers want to consolidate but flexibility is needed depending on the application and use case moving to the cloud

@BMCSoftware agree, it is not more a CAPEX versus OPEX.. Enterprises who have adopted cloud , are now asking for price compression in common features and capabilities that are core across cloud providers.

@iLakshmiS Absolutely - this is the right way to go about it if you think apps and app services. But the minute you bring infra or shared services it becomes a tough choice

@jwittich - latency is definitely a factor. We need to be in a cloud closer to where the customers are.

@codyhosterman good point Cody on de-risking cloud choice.  As a #SaaS Provider, we deliver our service across multiple #cloud platforms worldwide for that reason - and it gives us flexibility as well.

@jwittich Yes - but latency is often under-estimated in terms of impact though. Its only after things come together that most realize they have screwed up on app experience

@codyhosterman We do the same, keep the service close to the user for best experience