Having had the questionable fortune of following the cloud space ever since… well, ever since Amazon Web Services was just a twinkle in the eye of Werner Vogels, I’ve seen a bunch of things come and go. From the early days when traditional vendors (yes, I’m looking at you IBM, HP and Oracle) decried the cloud as dangerous and not fit for enterprise workloads to only a few short years later when they miraculously renamed all of their legacy product as somehow ”cloudified” (who remembers Oracle’s ironic cloud in a box?) Fast forward to today and most people accept a fairly standard (ish) definition of what cloud is and have moved on to actually delivering value to customers.
One thing we saw a lot of in the early years of this decade (and, for that matter, in the latter years of the last decade) was vendors offering the apparent holy grail of a cloud migration service. The allure was obvious, the value proposition these vendors offered was that, miraculously, legacy applications could be refactored in the blink of an eye into cloud applications. No matter that the cloud works fundamentally differently from on-premises infrastructure and that cloud applications, by their very nature, require a different type of architecture – these vendors were on a roll.
Luckily most of these vendors failed and, in their demise arose a more mature approach towards cloud applications – a realization that “cloud native” means something completely different from simply stuffing existing applications into cloud VMs. It means leveraging modularity, composability and a completely different approach towards development and operations of those applications. Where cloud migration failed in a sea of incredulity, cloud-native approaches arose.
Until now
I thought Gmail was playing tricks on me and sending me missed emails from 2010 the other day when I received a pitch from Virtual Instruments trumpeting its new cloud migration readiness service. The offering is apparently aimed at helping enterprises intelligently move to the cloud and delivers an unheralded number of supposed benefits. According to the release, enterprises can:
- De-risk their upcoming cloud migrations
- Validate the suitability of the targeted applications based on their on-premises performance SLAs
- Understand application dependencies
- Preserve performance in the cloud
- Compare and contrast estimated costs of various cloud platforms
Unicorns and rainbows emanate from this particular vendor, it would seem.
But what’s wrong with what they’re offering?
Two things. Firstly, simply repackaging applications and shoving them in the cloud doesn’t work – it’s inefficient, expensive and prone to untimely death. Secondly, even if it was viable to do what Virtual seem to think is the norm, it would be an approach that fails to leverage all the benefits that the cloud can bring. Looking at all of the purported value proposition that Virtual is delivering:
- De-risk their upcoming cloud migrations – how? By relying on an archaic approach towards application development and operations? That’s not a sort of de-risking that I like
- Validate the suitability of the targeted applications based on their on-premises performance SLAs – This is consultant double-speak. SLAs are different in the cloud from on-premises. Pretty much there’s an easy answer, if your application was created prior to the advent of the cloud, it is unlikely to be a prime candidate for cloud migration (even if you could achieve that holy grail)
- Understand application dependencies – as above, dependencies work differently in the cloud and Netflix has spawned a gazillion technical articles with its thought leadership on just how different this looks
- Preserve performance in the cloud – actually, quite the opposite. Not creating an application with a view to the cloud’s characteristic pretty much guarantees it won’t perform
- Compare and contrast estimated costs of various cloud platforms – this is just dumb. Like for like cloud is cheaper, but this is the wrong answer to the wrong question
Gilding the already golden lily
In justifying the reason for its own existence, Virtual suggests the pain that enterprises are facing:
For many enterprises, moving certain applications to the cloud is a core pillar of their digital transformation strategy, as it offers significant value in terms of business agility, faster innovation and scalability. However, migrating legacy applications to the cloud is a significant initiative for most enterprises, and they must answer critical questions before doing so – such as whether their business-critical applications will perform as expected once in the cloud, how much it will cost to run those applications in the cloud, and which cloud provider makes the most sense. Since most enterprises’ internal IT organizations lack deep cloud expertise, they need to partner with a services provider that offers extensive workload behavior knowledge and supporting cloud migration technologies.
Or, to paraphrase “we can help you move to the cloud without actually having to put in the effort to think what a cloud application should look like. Also, we know you just want to be able to say ‘yeah, we’re cloud’ to your CEO so we can get you there without any effort on your part and just via a sweet consulting gig for us.” It’s a business model and a product offering which was flawed a decade ago and, to be frank, feels pretty obscene today.
I mean, really, even those snake oil merchants who sell hollow dreams via lucrative consulting arrangements (and, yes, I’m thinking of the likes of Deloitte, Accenture, PWC and their ilk) have moved beyond such shameless attempts at avoiding a reality that 95% of the technology already regards as a given.
I can’t fathom what these people are thinking, and how anyone could have imagined that what they’re offering is really needed, or wanted, in-market. Welcome to the future, folks, get into your DeLorean and buckle up.
Ben – after seeing your post, we really enjoyed our discussion earlier this week about Virtual Instruments’ rationale behind developing the Cloud Migration Readiness service. To re-iterate, that rationale is predominantly based on our existing VirtualWisdom customers’ feedback on their cloud migration needs. While cloud native is the ideal approach for most enterprises’ longer-term cloud initiatives, the reality is that an overwhelming number of enterprises need to take a staggered approach to the cloud that leverages the ‘lift and shift’ strategy for existing workloads. This new service, which leverages our unique expertise in workload performance analytics, workload simulation, and infrastructure performance monitoring, was built with that reality in mind, as it gives customers the flexibility they demand for their cloud migration projects.