Healthcare.gov, the Department of Health and Human Services site for the federal health insurance marketplaces offered under the Affordable Care Act (a.k.a. Obamacare), will be finding a new home next year.
The agency has awarded Hewlett-Packard a $38 million contract to host Healthcare.gov starting in March, The Wall Street Journal reported last week. Verizon's Terremark unit now hosts the site but the decision to move to a new provider was made before the Oct. 1 public launch of Healthcare.gov, which has suffered ongoing and widely criticized outages. The Verizon datacenters have experienced numerous outages since the launch and, according to the report, agency members were aware of prior problems.
It bears noting that the outages are just one of many reasons Healthcare.gov, one of the hallmarks of Obama's presidency, has failed. The heart of the reason many people were initially unable to register was an application designed to authenticate individuals and application integration issues. The blame largely falls on how the project was managed, not any one component, as I reported back in October when I described it as "the biggest IT failure ever."
Certainly, losing Healthcare.gov gives Verizon a black eye. At the same time, Verizon is in the midst of revamping its cloud service portfolio and will have an opportunity to regroup. The company said in October that its new IaaS offering is scheduled for release early next year.
For HP, the move will give the company a chance to showcase its ability to run a business-critical site. Of course, if it's not working well by then, things could get pretty ugly.
Posted by Jeffrey Schwartz on 12/05/2013 at 12:29 PM0 comments
Salesforce.com and Hewlett-Packard have inked a deal to let customers build virtual instances of the Salesforce.com CRM platform.
The companies announced the pact at Salesforce.com's annual Dreamforce conference, taking place this week in San Francisco. Using HP's "Converged Infrastructure" of servers, storage and network gear, the companies will collectively build the Salesforce.com Superpod.
"The Salesforce Superpod will allow individual customers to have a dedicated instance in the Salesforce multi-tenant cloud," said Marc Benioff, the company's chairman and CEO, in a statement announcing the deal.
However, Salesforce.com will host the Superpods in its own datacenters, not HP's. In fact, the Superpods will be identical to the existing 15 pods in Salesforce.com datacenters used to host the company's CRM platform, InformationWeek reported. The key difference is that Salesforce.com will equip the Superpods with HP infrastructure.
Furthermore, Salesforce.com is only offering the Superpods to the largest of enterprises, the InformationWeek report pointed out, adding that they're intended for those who have governance and security requirements. "For the vast majority of customers, this is not appropriate," Benioff reportedly said. "But there are customers who want to go to another level."
That may address governance and security questions for some, but for those organizations who must keep data on their premises, the virtual instances may not be enough. On the other hand, Salesforce.com is not seeing a slowdown in its growth.
The company on Monday reported third-quarter revenues of $1.08 billion, up 36 percent over the same period a year ago. Benioff noted on the company's earnings call that it's the first time Salesforce.com had a quarter that broke the $1 billion in revenues barrier. And it comes just four years after Salesforce.com first exceeded $1 billion in a single year. Now the company is in a $5 billion run rate for fiscal year 2015, Benioff said.
Posted by Jeffrey Schwartz on 11/21/2013 at 1:13 PM0 comments
While there's little hard data to show market share for the use of public cloud computing services, most would agree that Amazon Web Services (AWS) is by far the most widely used Infrastructure as a Service (IaaS) to date.
Analyst Ben Schachter of Macquarie Capital said AWS' revenues this year could hit $4 billion. The Wall Street Journal reported this week, though, that Schachter's figure sounds much higher than prior estimates, which pegged revenues in the $1 billion to $2 billion range. As rivals such as Google, Hewlett-Packard, IBM, Rackspace and Microsoft are gunning for that business by expanding their own services, Amazon is clearly getting under Big Blue's skin, in particular.
After giving up on its contested losing bid for a $600 million cloud contract for the CIA to Amazon late last month, IBM took to the media last week, arguing it has a larger cloud business than Amazon. IBM ran an ad that claims its cloud powers 270,000 more Web sites than Amazon. Having acquired SoftLayer this year, IBM also argued that its cloud service lets customers run apps on dedicated servers without interference from "virtual neighbors."
The timing of the ads -- right after losing the CIA bid and just days before Amazon's re:Invent customer and partner conference -- is also telling. I spoke with Mark Dietz, director of IBM SmartCloud solutions, last week and asked if the ad campaign was timed in concert with AWS re:Invent, and he denied any connection.
But IBM is apparently running ads on buses in Las Vegas, where the conference is taking place, noted AWS Senior VP Andy Jassy in his keynote address at AWS re:Invent. Jassy didn't mention IBM by name, but left little doubt that he was referring to Big Blue.
"An old-guard technology company based in New York seems to be pretty worked up about AWS these days," Jassy said as he held up the ad.
"The advertising is creative, I'll say that," he continued. "I don't think anybody that knows anything about cloud computing would argue this company has a larger cloud computing business than AWS. But it's a way to jump up and down and to hand-wave to try to distract and confuse customers instead of trying to build comparable functionality to what AWS has. We believe that customers are smart. And we believe that customers are not going to allow them to have the wool pulled over their eyes easily. And we believe that in this day and age of the Internet, and the transparency that that provides, and given how inexpensive it is to try these technology infrastructure platforms, that customers will figure out what's what."
Dietz doesn't see it that way. "We're making bold moves in the cloud and this ad campaign really underscores that," he said, referring to IBM's acquisition of SoftLayer, which brings 21,000 IaaS customers into the fold.
Nevertheless, IBM's campaign targeting Amazon was surprising. I asked Dietz if there were sour grapes over the CIA deal.
"That's definitely not the case," he insisted. "I wouldn't say anything to diminish the importance of the CIA contract but I'd emphasize that's one deal. If you look at just the federal sector alone, we have bigger deals and more customers in the federal sector, let alone businesses large and small using different aspects of our overall cloud portfolio."
IBM has made significant investments this year that further its efforts to push into cloud computing. Embarking on such an ambitious campaign challenging Amazon is the latest sign IBM means business when it comes to cloud computing. But the two companies have very different philosophies on the end game.
In his re:Invent keynote, Jassy spelled out Amazon's philosophy: "We have a pretty different view on how hybrid is evolving than a number of the old-guard technology companies. I think the old-guard technology companies believe that what enterprises want is to run most of their workloads on-premise behind the firewall, and then they just want to have a little bit of an ability to surge to the cloud. We have a very different view of that. We believe in the fullness of time that very few enterprises are going to own their own datacenters, and those that do will have much smaller footprints than they have today."
IBM, Microsoft and VMware are emphasizing a hybrid world for many years to come, and in the end the wares they're building are preparing them for a world that Amazon envisions. The only subject of debate that remains is whether datacenters will become the exception rather than the rule, and how long that will take to materialize.
Posted by Jeffrey Schwartz on 11/14/2013 at 11:31 AM0 comments
The Cloud Security Alliance (CSA) yesterday launched an initiative that aims to pave the way for organizations to use cloud computing services to protect their own infrastructures.
A new working group called the Software Defined Perimeter (SDP) project represents a departure for the organization that was formed with the mission of improving cloud security. The SDP Working Group seeks to provide a standard way to use cloud services to protect their infrastructures in the age of bring-your-own-device (BYOD) and employees' own use of cloud services.
"What we're proposing is actually very new in the sense that the cloud actually does become your perimeter versus thinking of the cloud to get low-cost CPU cycles," said Junaid Islam, co-chair of the SDP Working Group and president and CTO of Vidder, a provider of security solutions.
The committee is developing a framework that identifies devices, implements standard authentication, and creates one-time use of a VPN on an application server that ensures the user can see only data he or she is permitted to access, Islam explained.
"We want to take all of this and run it in a cloud service," Islam said. "What the CSA is doing, instead of everyone in the industry coming up with their own version of this, is creating a standard, public domain, free-to-use-without-restrictions-framework that pulls all these concepts together in a framework that is well thought-out and vetted by a team of experts."
The CSA is already collaborating with major cloud providers, Islam added, though he declined to name any.
Islam pointed to a large corporate customer with a well-known brand that is working with a "gigantic" cloud provider, both of which he would not name, that will represent the first deployment of SDP. The deployment will be announced at the RSA Conference in late February, presuming it doesn't leak out sooner.
Bob Flores, former CTO of the CIA and now CEO of Applicology, a consulting firm, is the other co-chair of the SDP Working Group. Flores will outline the new framework at the CSA Congress conference in Orlando next month. The committee will also publish a whitepaper outlining the framework.
The workgroup is also creating APIs that customers, systems integrators and developers can use to implement these security protocols. The group will release previews next quarter and hopes to have a working API by mid-year. One such API will use SAML to enable the use of device certificates, which Islam explained are rarely used today.
"What we're doing is pulling it together, saying, 'Here's a way to get a device cert. Now that you have it use it to create this mutual TLS connection, then use it to set up your credentials,'" Islam said.
To be sure, SDP is in its early stages. "It's important to understand, what starts rolling out in the first quarter of next year is not the be all, end all. This will build on itself as time rolls on and best practices are identified," Flores said. "It may start off with something as simple as authenticating end users to a cloud, which today is not done universally. From there, we will go on to other issues related to security."
Posted by Jeffrey Schwartz on 11/14/2013 at 11:40 AM0 comments
Amazon Web Services said it wants to shake up the struggling VDI market with a cloud-based alternative that requires no hardware, software or datacenter infrastructure. Amazon's solution is Amazon WorkSpaces, which it claims it can offer services at half the cost with better performance than traditional virtual desktop infrastructure platforms today.
Amazon Web Services senior VP Andy Jassy revealed the new cloud-based VDI offering in his opening keynote address at the company's second annual re:Invent customer and partner conference taking place in Las Vegas.
Saying VDI hasn't taken off because it's complex to setup and manage, Jassy told the thousands of attendees and online viewers in his keynote that Amazon WorkSpaces promises to reduce those barriers. It will allow organizations to move their desktop licenses to Amazon and provides integration with Active Directory.
"You can access your Amazon WorkSpace from any of your devices whether it's a desktop, laptop or an iOS device," Jassy said. "And you get persistent sessions, so if you're using a WorkSpace on your laptop, and you switch to your Android [or any other] device, the session picks up just where you left off. What's also nice, because it's a cloud service, all of the data lives in the cloud -- it doesn't live local to those devices, which of course is a concern for an IT administrator."
The company described in a blog post a use case with 1,000 employees that would cost just $43,333 using Amazon WorkSpaces. This would be 59 percent less expensive than an on-premise VDI deployment that would cost $106,356 (which includes datacenter investments).
Amazon will initially offer a Standard service that costs $35 per month for one virtual CPU, 3.75 GB of memory and 50 GBytes of capacity; and a Performance plan that costs $60 for two virtual CPUs, 3.75 GB of memory and 100 GB storage per user. A Performance Plus package will come with 7.5 GB of memory. Customers that don't have licenses to move over can purchase licenses for Microsoft Office and antivirus software firm Trend Micro for $15 per month per user.
Jassy said the company intends to first offer invitation-only trials. He did not disclose general availability. Customers can register for the preview now.
Posted by Jeffrey Schwartz on 11/13/2013 at 12:00 PM0 comments
Red Hat Software released an upgraded version of its CloudForms tools used to deploy and manage private and hybrid Infrastructure as a Service (IaaS) clouds with support for OpenStack. The new CloudForms 3.0 debuted at the OpenStack Foundation Summit, taking place in Hong Kong this week.
Also at the large industry event, Red Hat said its OpenShift Platform as a Service (PaaS) software can now be deployed on OpenStack clouds. The move lets developers and IT pros use OpenStack infrastructure to create PaaS environments.
"OpenStack is one of the foundational pillars on which we are building out our open hybrid cloud strategy and portfolio," said Bryan Che, Red Hat's general manager of cloud, speaking on a Web cast from Hong Kong. Since committing to OpenStack two years ago, Che said Red Hat has become the largest contributor to the open source project, including the most recent Havana release. The company said it has 87 engineers working on 69 various OpenStack projects.
Red Hat was able to integrate the OpenStack management functionality into CloudForms after it acquired ManageIQ in the beginning of the year for $104 million. CloudForms 3.0 also gained improved management of Amazon public and hybrid cloud services and VMware infrastructure from ManageIQ. "Enterprises are now going to be able to deploy OpenStack, and they'll be able to bring enterprise-class management capability on top of it, as well," Che said.
The new release also includes tools for authoring and administrating service catalogs. The company offers it as a virtual appliance. Red Hat offers it as a standalone management platform, as well as part of the Red Hat Enterprise Linux Platform, which includes the Linux distribution bundled with OpenStack.
In addition to the new wares, Red Hat has extended its OpenStack certification programs. The company's partner network, launched in June, now has 140 members that have introduced over 900 certified solutions available in the Red Hat Marketplace. "As a result of the partners who are engaged with us in this program, we're seeing hundreds of customers with POCs," said Mike Werner, Red Hat's senior director for global ecosystems.
While the current certifications focused on core compute, storage and networking, Red Hat said it is adding OpenStack SWIFT object storage and extensions to advanced networking features in the community called Neutron.
On top of certification for OEMs, independent system builders and ISVs, Red Hat is extending the certification program to systems integrators, managed service and cloud service providers, and its channel partners.
Posted by Jeffrey Schwartz on 11/07/2013 at 11:11 AM0 comments
Systems integrator CSC has filled a key hole in its cloud service portfolio by agreeing to acquire ServiceMesh, a key cloud service automation provider that facilitates running various apps on multiple public cloud services.
The deal gives CSC the ServiceMesh Agility Platform, software that automates the deployment and management of enterprise apps running on public, private and hybrid clouds. The ServiceMesh policy engine provides governance, compliance and security of enterprise information running in various cloud environments, the company said.
"With ServiceMesh, we will empower an ecosystem of enterprise software providers by lowering the friction for companies to execute a multi-vendor hybrid cloud strategy while maintaining central governance, policy and administration," said CSC Chief Technology Officer Dan Hushon in a statement announcing the deal.
Forrester Research analyst Dave Bartoletti noted in a blog post that adding ServiceMesh gives CSC a much-needed unified service catalog, orchestration and governance platform; cloud neutrality; and an orchestration engine that rivals like Dell, HP and IBM already have.
"CSC has picked up one of the few independent hybrid/multi-cloud management vendors," Bartoletti noted. "CSC can combine its strong managed services capabilities and IT management tools expertise with the application lifecycle (DevOps) focus of ServiceMesh to reach a powerful cloud buyer: the app owner and developer. Apps are where the cloud action is."
CSC said the deal is slated to close by the end of the quarter.
Posted by Jeffrey Schwartz on 10/31/2013 at 1:24 PM0 comments
In a move aimed at broadening the enterprise reach of its application marketplace, Salesforce.com on Thursday launched its Salesforce Private AppExchange. The new exchange lets IT organizations roll out their own apps, which can be accessed on mobile devices and Web-based clients.
Built on the foundation of its existing Salesforce AppExchange, the new private iteration lets enterprise customers build and distribute internally developed software that can be accessed via the single sign-on service the company announced earlier this month called Salesforce Identity. The private enterprise marketplace is designed to let IT organizations consolidate access to applications and give employees broader choices of tools that are on the common platform.
"You can distribute any application, so it could be an app that's delivered by one of our ISVs that's listed on the AppExchange, or it could be an application that runs on another cloud service that doesn't have an integration point to the platform," explained Sara Varni, senior director of Salesforce.com's AppExchange.
Asked if the Private AppExchange requires pre-integration with other cloud services or application stores, Varni said, "It's really a mechanism to distribute applications to end users more so than where the application exists."
One key benefit of the Salesforce Private AppExchange is it will allow organizations to integrate Salesforce.com's Chatter enterprise social networking service into their apps, explained Constellation Research vice president and principal analyst Alan Lepofsky.
"Employees can now engage with their colleagues to discuss the applications, provide self-support, weed out the best apps, etc.," Lepofsky said. "This is more effective than accessing a generic public app store and having to fend for yourself to discover the best tools. Several social business vendors provide integrated application stores, but many are just catalogs and are lacking the social elements that Salesforce is looking to provide."
Users can move from the Private AppExchange to the Salesforce identity app launcher, enabling them to see what apps they have access to, Varni added. Any Salesforce customer that has access to the AppExchange via an enterprise license has rights to the private exchange, as well, she said. Those without enterprise licenses can provide access to the new private exchange and the new Salesforce Identity service for $5 per user, per month.
Posted by Jeffrey Schwartz on 10/31/2013 at 1:21 PM0 comments
Remember when EarthLink was the Internet service provider many people used if they didn't want to rely on AOL for their dial-up access?
EarthLink, which has a formidable communications backbone these days, is also looking to make its mark as a cloud infrastructure service provider.
The company this week launched a private cloud hosting service, targeted at those who require dedicated clouds rather than shared multitenant services. The company says its new Private Cloud Hosting is intended for workloads that require high availability. It's available with four service levels that can deliver N+1 redundancy with its self-healing platform, which offers automated server failover and SSAE 16 SOC2 certified datacenters and a domestic MPLS network and a SAN.
EarthLink's cloud infrastructure is based on VMware's vCenter cloud management platform, which supports customizable dashboards and real-time performance monitoring and analytics. Users can specify most builds of Windows Server from Windows 2000 to Windows Server 2012, as well as popular Linux server distributions.
The service is targeted at implementations that require 10 virtual machines or more. The company doesn't publish its pricing; rather, it issues quotes based on configuration requirements. While building up its communications backbone over the years -- most recently with its OC-192 10 Gbps fiber backbone, voice over IP portfolio, IP-based MPLS and EVDO wireless service -- EarthLink has also assembled cloud hosting assets.
Over two years ago, the company acquired LogicalSolutions.net, an enterprise cloud service provider. Also that year, EarthLink acquired Business Vitals, a managed IT security and professional services company; xDefenders, which also provides managed IT security services; and Synergy Global Solutions' cloud-based applications business. EarthLink this year also picked up CenterBeam, which provides remote IT services and has added four datacenters for its cloud hosting platform.
The only question that remains is will IT decision-makers be able to shake off their image of EarthLink as that provider of dial-up services (and, later, ISDN and DSL communications) when considering it for cloud hosting services?
Posted by Jeffrey Schwartz on 10/24/2013 at 10:13 AM0 comments
Cloud Technology Partners is best known for its consulting services. But the company has a new business aimed at making it easier for developers to migrate their .NET and Java applications to different cloud service providers.
The company's new PaaSLane tool aims to remove the time-consuming and labor-intensive process of evaluating software source code to ensure an application can work properly in a cloud environment. That's especially important as providers roll out new services and update existing ones.
Cloud Technology Partners this week released a free beta version of PaaSLane, available for download here.
PaaSLane is designed to continuously track changes in cloud environments and provide recommendations to developers to remediate their code, said Ben Grubin, director of product management at Cloud Technology Partners. "We provide overall cloud readiness metrics and reports for each app run through the software," Grubin said, adding that PaaSLane's engine can process 250,000 lines of code in 10 minutes. "It doesn't just find issues -- it estimates the development effort to address those issues."
This is important, Grubin said, to keep up with different cloud platforms, including offerings from Amazon Web Services, Microsoft's Windows Azure and the various OpenStack-based services.
"This is a complex problem because there are thousands of changes from different cloud providers, and developers don't spend time keeping up with all of these platforms," he said. "You can spend a lot of time researching how to develop for each platform, and that can have a huge productivity impact."
Posted by Jeffrey Schwartz on 10/24/2013 at 10:51 AM0 comments
The OpenStack Foundation on Thursday released the eighth version of its open source cloud infrastructure operating platform.
The new OpenStack Havana release has 400 new features, but the most noteworthy are support for monitoring, orchestration and global cluster support with the platform's new object storage system.
Havana also offers support for improved quality of service settings and end-to-end encryption across all block storage drivers, SSL built into all of the OpenStack service API, an upgraded VPN, firewall support and the ability to boot from a volume, enabling live migrations.
OpenStack Foundation Executive Director Jonathan Bryce told me these new features are critical for enterprise and business-critical applications, including ERP and virtual desktop infrastructure.
"Enterprises can deploy more and more applications into their OpenStack environments," Bryce said in a briefing Wednesday. "That's critical if you're going to run your business-critical systems in the cloud."
Bryce described the metering and monitoring capability, which can be exposed through a Web-based dashboard, as critical for those who require visibility into usage across all of the OpenStack services.
"Before, if you wanted to pull out who was using more resources or determine where there were bottlenecks and hotspots, you needed to use different tools," Bryce explained. "The OpenStack metering and monitoring services pre-integrate with the other OpenStack services and aggregate all that data into a central location."
That's useful from an operational standpoint but also for those who need to bill users and/or do chargeback, as well as for issuing alerts when certain conditions arise, he added. The monitoring and metering capability can be added to dashboards and other tools via a new central API.
"It's a powerful way to see what's going on in your cloud app at an overall level or to see which users or apps are driving any usage," Bryce said. Administrators can track network and storage users and drill down on a per-app or per-user basis.
The new orchestration capability is based on a template-driven engine, which lets developers describe the infrastructure resources needed to deploy an application. In the template, a developer can define compute nodes such as the number of Web servers, database servers and networks, and storage resources feed all of that into the orchestration engine that provisions those resources and issues the necessary calls to set everything up, Bryce said.
"What's cool is because it ties into the metering and monitoring service, you can set thresholds, and if they receive alerts, they can take actions like auto-scaling," he said. "These two features in combination provide a high level of functionality on top of the core OpenStack components."
The template language is JSON and Bryce said there are already quite a few available in the open source community. He anticipates more will surface now that the Havana release has gone live.
The global cluster support on the object storage system enables a single storage system to span multiple datacenters, which is important for disaster recovery to enable replication across datacenters and for allowing users to connect to the nearest cloud facility, said Joe Arnold, CEO of SwiftStack, a private cloud storage provider which led the development of the global cluster support in OpenStack Havana's object storage system.
Concur, which provides the widely used, cloud-based expense reporting Software as a Service (SaaS), is among SwiftStack's customers that have deployed the new global clustering capability in its OpenStack private cloud, which the company will talk about at next month's OpenStack Summit in Hong Kong.
"They have two datacenters set up and they're sending their users to the nearest datacenter," Arnold said. "When a customer takes a picture of a receipt with their phone, it's uploaded to the datacenter nearest to them. That just gives the user a better experience. And if there's a failure in either datacenter, they can send all of the users to the other datacenter and not miss anything."
Red Hat is among those in the OpenStack community that have released new distributions based on Havana. Red Hat's RDO Havana is aimed at deployment across Red Hat's various Linux distributions, including Red Hat Enterprise Linux, Fedora, CentOS and Scientific Linux. Canonical's new Ubuntu Server 12.04 LTS and 13.10 also include the new OpenStack Havana release.
Posted by Jeffrey Schwartz on 10/17/2013 at 12:12 PM0 comments
At this week's VMworld 2013 Europe conference in Barcelona, VMware announced key updates to several of its products. Most important is improved performance monitoring of Hyper-V, SQL Server and Exchange within VMware vCenter Operations Management Suite 5.8. VMware also updated its vCenter Log Insight analytics tool.
The vCenter Operations Management Suite was able to monitor and manage Hyper-V workloads as a guest OS by placing its Hyperic agent in the box. But now, through Microsoft's System Center Operations Manager or VMware's Hyperic management packs for vCenter Operations Manager, it can manage Hyper-V hosts and virtual machines. It also adds support for Amazon Web Services (AWS) EC2 infrastructure as a service (IaaS).
The new version collects data on the CPU, network, memory and other components, and feeds that into its analytics engine to separate normal performance behavior from unhealthy activity and then provides alerts.
Expanding support for Hyper-V is a smart move that will be welcomed by VMware and Microsoft customers alike, says Pund-IT principal analyst Charles King. "By expanding its support for and visibility into Hyper-V and public cloud services like AWS, VMware is highlighting its continuing technical leadership," King said. "Since these new features are also coming with no additional premium, adding them also enhances the value proposition of VMware's solutions and services."
In tandem with the vCenter Operations Management Suite upgrade, VMware also updated its recently launched vCenter Log Insight analytics tool. The new Log Insight 1.5 release provides real-time analytics to provide searchable information and dashboards. Released in June, the Log Insight upgrade supports "content packs" for specific systems such as Exchange Server and SQL Server, as well as products from Hytrust, EMC, NetFlow and VMware's own Nicira software-defined networking platform, said Mark Leake, director of product marketing in VMware's cloud management business unit.
"These new out-of-the-box capabilities enhance the discovery and topology," Leake said. "So you get deeper discovery of app instances and components and you can apply the analytics that we have in vCenter Operations Manager to them."
Posted by Jeffrey Schwartz on 10/16/2013 at 4:49 PM0 comments