Business · Cloud

Oracle works toward capturing enterprise Cloud IaaS demand

4

The enterprise cloud market is still shows a widely untapped potential. A significant portion of this potential comes from the demand generated by the legacy applications that are sitting in the myriad of corporate data centers. The footprint from these legacy workloads alone is staggering. Start adding in the workloads that sit in secondary data centers that often do not get included in many metrics and one can quickly see the opportunity.

ORACLE STARTS FROM THE GROUND UP

At Tech Field Day’s Cloud Field Day 3, I had the opportunity to meet with the team from Oracle Cloud Infrastructureto discuss their Infrastructure as a Service (IaaS) cloud portfolio. Oracle is trying to attract the current Oracle customer to their cloud-based offerings. Their offerings range from IaaS up through Software as a Service (SaaS) for their core back-office business applications.

The conversation with the Oracle team was pretty rough as it was hard to determine what, exactly, that they did in the IaaS space. There were a number of buzzwords and concepts thrown around without covering what the Oracle IaaS portfolio actually offered. Eventually, it became clear during a demo, in a configuration page what the true offerings were: Virtual Machines and Bare Metal. That’s a good start for Oracle, but unfortunate in how it was presented. Oracle’s offering is hosted infrastructure that is more similar to IBM’s SoftLayer(now called IBM Cloud) than Microsoft Azure, Amazon AWSor Google Cloud.

ORACLE DATABASE AS A SERVICE

Beyond just the hardware, applications are one of the strengths of Oracle’s enterprise offerings. And a core piece of the puzzle has always been their database. One of the highlights from the conversation was their Database as a Service (DBaaS)offering. For enterprises that use Oracle DB, this is a core sticking point that keeps their applications firmly planted in the corporate data center. With the Oracle DBaaS offering, enterprises have the ability to move workloads to a cloud-based infrastructure without losing fidelity in the Oracle DB offering.

Digging deeper into the details, there were a couple interesting functions supported by Oracle’s DBaaS. A very cool feature was the ability to dynamically change the number of CPUs allocated to a database without taking an outage. This provides the ability to scale DB capacity up and down, as needed, without impact to application performance.

Now, it should be noted that while the thought of a hosted Oracle DB sounds good on paper, the actual migration will be complicated for any enterprise. That is less a statement about Oracle and more to the point that enterprise application workloads are a complicated web of interconnects and integrations. Not surprisingly, Oracle mentioned that the most common use-case that is driving legacy footprints to Oracle Cloud is the DB. This shows how much pent-up demand there is to move even the most complicated workloads to cloud. Today, Oracle’s DB offering runs on Oracle Cloud Infrastructure (OCI). It was mentioned that the other Oracle Cloud offerings are moving to run on OCI as well.

Another use-case mentioned was that of High-Performance Computing (HPC). HPC environments need large scale and low latency. Both are positive factors for Oracle’s hardware designs.

While these are two good use-cases, Oracle will need to do things that attract a broader base of use-cases moving forward.

THE CIO PERSPECTIVE

Overall, there seems to be some glimmers of light coming from the Oracle Cloud offering. However, it is hard to get into the true differentiators. Granted that Oracle is playing a bit of catch-up compared with other, more mature cloud-based offerings.

The true value appears to be focused on existing Oracle customers that are looking to make a quick move to cloud. If true and the two fundamental use-cases are DBaaS and HPC, that is a fairly limited pool of customers when there is significant potential still sitting in the corporate data center.

It will be interesting to see how Oracle evolves their IaaS messaging and portfolio to broaden the use-cases and provide fundamental services that other cloud solutions have offered for years. Oracle does have the resources to put a lot of effort toward making a bigger impact. Right now, however, it appears that the Oracle Cloud offering is mainly geared for existing Oracle customers with specific use-cases.

Business · CIO · Cloud · IoT

Three things to look for at Amazon’s upcoming AWS re:Invent conference

screen-shot-2016-11-23-at-8-10-21-am

As folks in the US prepare for the Thanksgiving holiday, those of us in technology are looking to Amazon’s annual cloud confab AWS re:Invent the week after Thanksgiving in Las Vegas. In preparation for the sold-out conference, there are a few things to look for at the conference.

ENTERPRISE ENGAGEMENT

Amazon has done a good job of attracting the Webscale and Startup markets. One could go so far as to say that Amazon has cornered these markets. For these folks, the options are wide open to address their specific and scaling requirements. The requirements for enterprises, however, are vastly different from their webscale and startup counterparts.

Look for indications that Amazon is starting to learn how to engage enterprises and is moving in that direction. The method, language and solutions vary greatly when considering a prospective customer that has an existing footprint over those that are looking to build their first.

INTELLIGENCE: AI & MACHINE LEARNING

Amazon already has a Machine Learning solution in their portfolio today. Look for further expansion beyond just tools and into the realm of intelligence. Amazon has done a great job of create a bevy of infrastructure tools. However, moving into the intelligence space is both necessary and the logical next step in maturing Amazon’s position in this market. Also, look for Amazon’s response to the growing interest in AI solutions. In addition, the combination of AI & Machine Learning is paramount to more mature IoT solutions.

MOVING UP THE STACK

To date, most of Amazon’s portfolio targets the infrastructure end of the stack. There are a few solutions that move up the stack, but not many. Even so, Amazon has done a stellar job of their existing efforts. If, however, Amazon is intending to capture more of the enterprise market and move beyond being simply a tool provider, it needs to move up the stack into the application realm. To date, it is not clear if Amazon has both the capability and strong interest to do so.

Across the board, Amazon’s competition may not have the depth in IaaS cloud that Amazon has today. Nor do they have the ecosystem that Amazon has worked hard to build over the past 10 years. However, what they lack in IaaS depth is countered by their breadth up and down the stack. And while they may lack the breadth of features in the IaaS space compared with AWS, each are quickly catching up and are posting impressive growth rates.

Next week should provide another exciting event for Amazon and those working in the Cloud space. Whether coming from the startup, webscale or enterprise space, all eyes are on Amazon and their next move.

CIO · Cloud

Microsoft Azure Stack fills a major gap for enterprise hybrid cloud

Azure Stack is billed as a version of Azure that runs in your corporate data center. Originally announced on May 4, 2015, Microsoft Azure Stack presented a significant change to the enterprise cloud spectrum of options. Prior to Azure Stack, enterprises looking for a private cloud option were left to build their own. While possible, not a trivial feat for most enterprises.

Today, Microsoft announced availability of their Technical Preview 1 (TP1); a series of planned technical previews leading up to Azure Stack’s General Availability later in 2016.

 

Azure Map

A PRIVATE CLOUD FOR ENTERPRISES

Azure Stack represents an on premise version of Microsoft’s Azure public cloud that runs in your corporate data center. If you are familiar with Azure, you are already familiar with Azure Stack.

Unlike many solutions that start small and scale up, Microsoft was challenged with the opposite problem; to scale down Azure. Azure Stack is essentially a scaled down version of Azure and the code between the two versions of Azure is remarkably similar. The further up the stack, the more similar the code base gets. For developers, this means a more consistent experience between Azure and Azure Stack.

Many enterprise customers are hesitant or incapable of making the leap from workloads on premise to full-on public cloud. Those reasons range from cultural resistance to regulatory considerations. Azure Stack provides a solution that fills the gap between full-on public cloud (Azure) and the prospect of creating a private cloud from scratch. Moreover, because of the consistent experience, customers are able to develop applications on Azure Stack and then move them fairly seamlessly to Azure. Many of the services are similar between the solutions, however, there are some obvious differences inherent to public vs. private cloud.

For now, Microsoft has drawn the line with Azure to focus on the Platform as a Service (PaaS) and Infrastructure as a Service (IaaS) cloud tiers. That may be true for Azure, however, Microsoft continues to grow their SaaS based solutions such as Microsoft Office 365. Microsoft stated that they are in the process of moving Microsoft 365 to Azure. It is anticipated that traditional enterprise core services such as Microsoft SQL server in addition to newer solutions like Internet of Things (IoT) will move to Azure in the form of a deployable ‘Template’.

It should not be minimized that the effort to move existing enterprise applications from their legacy footings is not a trivial effort. This is true for applications moving to Azure Stack, Azure, along with cloud solutions including Amazon AWS, Blue Box, and others.

 

Enterprise Cloud Map

THE KEY TO AZURE STACK LOCAL TARGET

The beauty of a local version of public cloud solutions is in its ability to sidestep many of the challenges that public cloud presents. In the case of regulatory or data privacy issues, Azure Stack provides the ability to leverage the benefits of cloud, while adhering to local regulatory issues surrounding location of data.

In the most simplistic form, one could consider Azure Stack another ‘Region’ in which to deploy applications. Microsoft’s management application, Azure Resource Manager (ARM) is able to deploy directly to Azure Stack as another target Region just as one would deploy to West US or East US. In the case of Azure Stack, the Region is Local. Customers do have the option to deploy internal (Local) Regions in a single zone or in separate zones.

DEVELOPING ON AZURE

One of the core benefits to Azure Stack is in the ability to build applications for Azure Stack (or Azure) and deploy them to either solution. Microsoft Visual Studio already has the ability to update locations in real-time from Azure and Azure Stack. The core of Azure deployments come in the form of a Template. There are already a number of Templates on GitHub for immediate download:

Quick Start ARM templates that deploy on Azure Stack: http://aka.ms/AzureStackGitHub

The Software Development Kit (SDK) for Azure Stack supports both PowerShell and Command Line Interface (CLI) just like Azure. In addition, deployment tools such as Chef and Puppet are supported via the ARM API to Azure Stack.

GETTING STARTED WITH AZURE STACK

While the download for Azure Stack TP1 will not be available until January 29th, there are a number of minimum requirements to get started. Keep in mind that this is the first Technical Preview of Azure Stack. As such, there is quite a bit of code to optimize for local use vs. the full Azure cloud. With Azure, the minimum configuration covered a full 20 racks! With Azure Stack, the minimum footprint has shrunk to a cluster of four systems with a maximum of 63 systems per cluster. Jeffrey Snover (Chief Architect, Microsoft Azure and Technical Fellow at Microsoft) outlined the minimum and recommended requirements in his blog post last month.

One may notice the Windows Server certification requirement. That is due to Azure Stack running on a base of Microsoft Windows Server. However, the Microsoft team believes that this will evolve over time. The memory requirements may also evolve. When running Azure Stack, the components take up approximately 24GB of RAM per system. While this may get optimized over time, additional components (such as clustering) may increase the memory consumption.

One may express concern at the very mention of a local cloud based on Windows Server. If anything, for the purposes of patching processes. Azure Stack is built to evacuate workloads off resources prior to patching. But Microsoft is looking at a wholly different approach to patching. Instead of applying the traditional Windows Server patches, Microsoft is looking to complete redeploy a new copy of Windows Server for the Azure Stack underpinnings. It will be interesting to see how this plays out.

There are two ways to get started with Azure Stack:

  1. Do It Yourself: Leverage reference architectures from Dell, HP and others that list the parts needed to support Azure Stack.
  2. Integrated Systems: Purchase a fully assembled, standardized solution.

 

IN SUMMARY

Azure Stack presents a significant game changer for Microsoft and the Enterprise cloud spectrum by filling a gap long-since needed. There are a number of other benefits to both enterprises and Managed Service Providers (MSPs) that Azure Stack brings. We will leave those for a later post.

 

UPDATE: The download for Azure Stack TP1 is live. You can get it here.

Business · CIO · Cloud · Data

Are the big 5 enterprise IT providers making a comeback?

Not long ago, many would have written off the likes of the big five large enterprise IT firms as slow, lethargic, expensive and out of touch. Who are the big five? IBM (NYSE: IBM), HP (NYSE: HPQ), Microsoft (NASDAQ: MSFT), Oracle (NYSE: ORCL) and Cisco (NASDAQ: CSCO). Specifically, they are companies that provide traditional enterprise IT software, hardware and services.

Today, most of the technology innovation is coming from startups, not the large enterprise providers. Over the course of 2015, we have seen two trends pick up momentum: 1) Consolidation in the major categories (software, hardware, and services) and 2) Acquisitions by the big five. Each of them are making huge strides in different ways.

Here’s a quick rundown of the big five.

IBM guns for the developer

Knowing that the developer is the start of the development process, IBM is shifting gears toward solutions that address the new developer. Just look at the past 18 months alone.

  • February 2014: Dev@Pulse conference showed a mix of Cobol developers alongside promotion of Bluemix. The attendees didn’t resemble your typical developer conference. More details here.
  • April 2014: Impact conference celebrated 50 years of the mainframe. Impact also highlighted the SoftLayer acquisition and brought the integration of mobile and cloud.
  • October 2014: Insight conference goes further to bring cloud, data and Bluemix into the fold.
  • February 2015: InterConnect combines a couple of previous conferences into one. IBM continues the drive with cloud, SoftLayer and Bluemix while adding their Open Source contributions specifically around OpenStack.

SoftLayer (cloud), Watson (analytics) and Bluemix are strengths in the IBM portfolio. And now with IBM’s recent acquisition of BlueBox and partnership with Box, it doesn’t appear they are letting up on the gas. Add their work with Open Source software and it creates an interesting mix.

There are still significant gaps for IBM to fill. However, the message from IBM supports their strengths in cloud, analytics and the developer. This is key for the enterprise both today and tomorrow.

HP’s cloudy outlook

HP has long had a diverse portfolio that addresses the needs of the enterprise today and into the future. Of all big five providers, HP has one of the best matched to the enterprise needs today and in the future.

  • Infrastructure: HP’s portfolio of converged infrastructure and components is solid. Really solid. Much of it is geared for the traditional enterprise. One curious point is that their server components span the enterprise and service provider market. However, their storage products are squarely targeting the enterprise to the omission of the service providers. You can read more here.
  • Software: I have long since felt that HP’s software group has a good bead on the industry trends. They have a strong portfolio of data analytics tools with Vertica, Autonomy and HAVEn (being rebranded). HP’s march to support the Idea Economy is backed up by the solutions they’re putting in place. You can read more here.
  • Cloud: I have said that HP’s cloud strategy is an enigma. Unfortunately, discussions with the HP Cloud team at Discover this month further cemented that perspective. There is quite a bit of hard work being done by the Helion team, but the results are less clear. HP’s cloud strategy is directly tied to OpenStack and their contributions to the projects support this move.

HP will need to move beyond operating in silos and support a more integrated approach that mirrors the needs of their customers. While HP Infrastructure and Software are humming along, Helion cloud will need a renewed focus to gain relevance and mass adoption.

Microsoft’s race to lose

Above all other players, Microsoft still has the broadest and deepest relationships across the enterprise market today. Granted, much of those relationships are built upon their productivity apps, desktop and server operating systems, and core applications (Exchange, SQL, etc). There is no denying that Microsoft probably has relationships with more organizations than any of the others.

Since Microsoft Office 365 hit its stride, enterprises are starting to take a second look at Azure and Microsoft’s cloud-based offerings. This still leaves a number of gaps for Microsoft; specifically around data analytics and open standards. Moving to open standards will require a significant cultural shift for Microsoft. Data analytics could come through the acquisition of a strong player in the space.

Oracle’s comprehensive cloud

Oracle has long been seen as a strong player in the enterprise space. Unlike many other players that provide the building blocks to support enterprise applications, Oracle provides the blocks and the business applications.

One of Oracle’s key challenges is that the solutions are heavy and costly. As enterprises move to a consumption-based model by leveraging cloud, Oracle found itself flat-footed. Over the past year or so, Oracle has worked to change that position with their cloud-based offerings.

On Monday, Executive Chairman, CTO and Founder Larry Ellison presented Oracle’s latest update in their race for the enterprise cloud business. Oracle is now providing the cloud building blocks from top to bottom (SaaS PaaS IaaS). The message is strong: Oracle is out to support both the developer and business user through their transformation.

Oracle’s strong message to go after the entire cloud stack should not go unnoticed. In Q4 alone, Oracle cloud cleared $426M. That is a massive number. Even if they did a poor job of delivering solutions, one cannot deny the sheer girth of opportunity that overshadows others.

Cisco’s shift to software

Cisco has long since been the darling of the IT infrastructure and operations world. Their challenge has been to create a separation between hardware and software while advancing their position beyond the infrastructure realms.

In general, networking technology is one of the least advanced areas when compared with advances in compute and storage infrastructure. As cloud and speed become the new mantra, the emphasis on networking becomes more important than ever.

As the industry moves to integrate both infrastructure and developers, Cisco will need to make a similar shift. Their work in SDN with ACI and around thought-leadership pieces is making significant inroads with enterprises.

Summing it all up

Each is approaching the problem in their own ways with varying degrees of success. The bottom line is that each of them is making significant strides to remain relevant and support tomorrow’s enterprise. Equally important is how quickly they’re making the shift.

If you’re a startup, you will want to take note. No longer are these folks in your dust. But they are your potential exit strategy.

It will be interesting to watch how each evolves over the next 6-12 months. Yes, that is a very short timeframe, but echoes the speed in which the industry is evolving.