Assessing OpenStack, CloudStack And Amazon Web Services: An Interview With Floyd Strimling Of Zenoss

In recent weeks, speculation about the commercial viability of OpenStack as a non-proprietary alternative to Amazon Web Services has mounted given the launch of the OpenStack Foundation, the release of the Essex version of its code and the deployment of Beta versions of OpenStack-based public clouds by Rackspace and HP . The debate about OpenStack’s readiness for the enterprise has been rendered all the more intense because of Citrix’s surprise decision to open-source CloudStack to the Apache Software Foundation, thereby creating a competitor to OpenStack overnight. Meanwhile, Amazon Web Services continues its relentless unfurling of feature after feature, and product after product onto its cloud computing platform in an effort to become the web’s one stop shopping ground for cloud software deployment, whether that involves IaaS, PaaS, SaaS, big data, cloud automation, an online marketplace or some combination thereof.

In an effort to obtain industry insight into the three horse battle between CloudStack, OpenStack and Amazon Web Services, I spoke to Floyd Strimling, Technical Evangelist and the Senior Director of Marketing & Community at Zenoss, a leading provider of management software for physical, virtual, and cloud-based IT infrastructures. As elaborated in his responses to the questions below, Strimling noted that OpenStack is tremendously promising but faces multiple challenges in the form of competition from CloudStack, the “staggering” rate of innovation of Amazon Web Services, the danger of Android-like fragmentation within the OpenStack community, and the technical challenge of delivering a “stable, flexible, and useable release that is ready for production deployments.” Strimling also identified “Red Hat as the most likely vendor to succeed in commercializing OpenStack” but refused to count out Piston Cloud Computing and Nebula from the list of candidates that will deliver successful commercialized versions of OpenStack.

Cloud Computing Today: How would you describe the place of OpenStack and CloudStack in relation to Amazon Web Services? Technically, for example, how would you compare CloudStack to OpenStack?

Floyd Strimling: While it is very tempting to compare OpenStack and CloudStack to Amazon Web Services, it is really not a productive exercise. OpenStack/CloudStack are enabling technologies that are used to build private and public clouds while AWS is itself a public cloud. Instead, you’ll need to compare the implementations of OpenStack or CloudStack to AWS and see how they fare. Thankfully, we’ll have that opportunity as two high profile OpenStack public clouds, Rackspace and HP, are released. While most of the technical elite ponders the future of OpenStack, the fact is most AWS customers are less worried about the enabling technologies underlying AWS. Instead, customers are focused on the services AWS offers, the features they provide, and the usability and stability of the solution. This puts Amazon in an extremely powerful position as they are focused on their own path and vision while essential obfuscating the underlying technology.

With regards to the place of OpenStack and CloudStack within the cloud market, they are extremely important as the represent a disruptive force that is open, disruptive, innovative, yet unproven.

I’m not about to step into the emotionally charged arguments of CloudStack vs. OpenStack. It’s really the difference between a commercially polished solution and one that has great promise. In the end, the needs of the users themselves will determine which is the superior solution as well as risks with each product.

Cloud Computing Today: As everyone knows, OpenStack has the potential to significantly impact the balance of cloud computing market share, particularly as it relates to IaaS. What dangers do you foresee for OpenStack in the coming year or two?

Floyd Strimling: While OpenStack has the potential to impact the balance of cloud computing market share, the questions are all about execution. Additionally, OpenStack has plenty of competition within CloudStack, vCloud, Eucalyptus, and more. Will OpenStack clouds have the chance to compete against established public clouds such as Amazon and Microsoft? If so, how?

The reality is OpenStack is full of large and powerful companies with competing agendas. Does Rackspace want to “share” the cloud business with HP or dominate? Does Red Hat want to share the commercialization of OpenStack with the likes of upstart Piston Cloud? This leads these competitors to differentiate their solutions and has the potential to create fragmentation of OpenStack a la Google’s Android. While each provider starts with the same core solution, they innovate around it via features, integrations, or other capabilities.

With that said, the biggest danger to OpenStack may be its ability to create a stable and mature release for its “customers.” After all, OpenStack’s customers are gunning for Amazon and they certainly aren’t standing still. OpenStack must resist the temptation to boil the ocean and instead focus on providing a stable, flexible, and useable release that is ready for production deployments. Time will tell if the OpenStack community is up for the task.

Cloud Computing Today: How do you foresee OpenStack’s most significant technical challenge? What must OpenStack achieve in order to become a credible alternative to Amazon Web Services?

Floyd Strimling: Perhaps the most significant technical challenge for OpenStack is building a commercially viable solution that satisfies the needs of its members while keeping pace with its competitors. It’s a daunting task to know that you have competition from other open source projects as well as public cloud providers.

Again, as OpenStack is an enabling technology, it needs a commercially viable offering that is in production to compete against Amazon Web Services. However, this is only the first step. These OpenStack clouds must have the features, availability, ease of use, scale, etc. to stand up to AWS. The challenge is as OpenStack struggles for technical parity with AWS, AWS is moving forward at a staggering pace offering innovative solutions with seemingly endless price reductions.

Cloud Computing Today: Which vendor or vendors do you see as most likely to succeed in commercializing OpenStack?

Floyd Strimling: I see Red Hat as the most likely vendor to succeed in commercializing OpenStack. After all, Red Hat has done this before and they finally are focused on the cloud. However, Red Hat has plenty of work to do to make this a reality. Additionally, an intriguing partnership of sorts would be if Red Hat and Rackspace created a joint offering. Red Hat would provide the enterprise or private cloud solution while Rackspace would handle the public cloud. In essence, this would give customers the ability to create hybrid environments backed by proven commercially viable companies.

Additionally, I wouldn’t count out Piston Cloud or Nebula as they offer unique solutions that are based on OpenStack. Piston’s ability to utilize a customer’s existing network hardware, such as Arista Networks, to build a secure, open, and easy to deploy cloud is a very compelling solution. In contrast, Nebula’s appliance-based solution offers a simple way to build an OpenStack cloud that appeals to those that prefer hardware-based solutions.

Bio

Floyd Strimling is a Technical Evangelist and the Senior Director of Marketing & Community at Zenoss. Floyd has been following the Cloud computing/autonomic computing (and predecessors), datacenter automation, virtualization, networking and security areas now for over a decade. Floyd also writes on technology trends at his personal blog, The Platen Report.

Zenoss is a leading provider of management software for physical, virtual, and cloud-based IT infrastructures. Over 35,000 organizations worldwide have deployed Zenoss to manage their networks, servers, virtual devices, storage, and cloud infrastructure, gaining complete visibility and predictability into their IT operations. Customers include Rackspace, VMware, Hosting.com, LinkedIn, Motorola and SunGard.

NTT and Verio Select CloudStack For Commercial IaaS Public Cloud Offering

NTT Communications Corporation and U.S.-based Verio, wholly owned subsidiaries of Nippon Telegraph and Telephone (NTT) Communications, today announced their selection of Citrix’s CloudStack IaaS platform as a key component of its public cloud service, CloudN. CloudN integrates the robust network infrastructure capabilities of Verio and NTT with Citrix’s CloudStack platform to provide customers with cloud solutions that offer “efficient, limitless scale and fast deployment – all on an open source software platform,” according to Sameer Dholakia, Citrix’s Group Vice President and General Manager. In the case of Verio, CloudN is currently available exclusively to its channel partners and targets cloud solutions focused on the needs of small and medium-sized businesses. Customers will enjoy unlimited inbound and outbound bandwidth for a provisional period after signup. Verio’s strategy of leveraging its channel partnerships in the hosting provider space to additionally offer cloud solutions for small and medium sized businesses represents a key departure from the stampede for enterprise cloud market share that has characterized the overwhelming majority of cloud solutions. NTT’s endorsement of Citrix’s CloudStack, meanwhile, is highly notable given how NTT has been a participating member of OpenStack and recently collaborated with Dell, Rackspace and HP to launch TryStack, a sandbox for OpenStack development.

Citrix Turns CloudStack Away From OpenStack To Apache Software Foundation

Today, Citrix revealed plans to open-source its CloudStack IaaS solution to the Apache Software Foundation under an Apache 2.0 license. Previously, Citrix had intended to integrate OpenStack with CloudStack in the form of an offering known as Project Olympus, which was unveiled last year as one of the first commercialized distributions of OpenStack. Because OpenStack is one of the world’s largest open source collaborations on cloud computing, a CloudStack product offering whose IaaS code was based on OpenStack would have represented an inter-operable, open source-derived product that avoids vendor lock-in by interfacing with other OpenStack-based products. Today’s announcement, however, represents an abrupt change of course by Citrix with significant implications for OpenStack, Amazon Web Services and cloud inter-operability more generally.

CloudStack will be released for public deployment immediately under the stewardship of the Apache Software Foundation (ASF), which takes credit for successfully incubating the likes of Hadoop and Cassandra. Citrix’s press release commented on the decision to open-source CloudStack with ASF as follows:

Elevating CloudStack into a full open source Apache project will further accelerate its mission of delivering a powerful, proven, hypervisor-agnostic platform that helps customers of all sizes build true Amazon-style clouds…. This proven method for incubating and advancing leading open source cloud projects is ideal for bringing a vibrant community of vendors and developers together to accelerate innovation, interoperability and standardization.

Rather than waiting for OpenStack to iron out issues related to its governance structure and choose whether to officially endorse the Amazon Web Services API, Citrix decided to open source CloudStack immediately. In a conference call the day before the publication of the press release, Citrix identified four attributes required of the ecosystem supporting their platform: (1) the platform must be fundamentally cloud-based; (2) production-level ability to scale; (3) Amazon Web Services compatibility; and (4) open source. Citrix’s Sameer Dholakia, Vice President and General Manager of the Cloud Platforms Group, spoke to GigaOM about the difficulties of integrating with OpenStack by noting: “Our very explicit public statement had been that we were going to try and build atop the OpenStack platform…[But] we can’t afford to wait a year or two for the technical maturation process that needs to happen [in order to integrate CloudStack and OpenStack].” Gartner’s Lydia Leong elaborated on Dholakia’s concerns about OpenStack’s immaturity by comparing CloudStack with OpenStack:

What makes this big news is the fact that OpenStack is a highly immature platform (it’s unstable and buggy and still far from feature-complete, and people who work with it politely characterize it as “challenging”), but CloudStack is, at this point in its evolution, a solid product — it’s production-stable and relatively turnkey, comparable to VMware’s vCloud Director (some providers who have lab-tested both even claim stability and ease of implementation are better than vCD). Taking a stable, featureful base, and adding onto it, is far easier for an open-source community to do than trying to build complex software from scratch.

The other reason for Citrix to turn away from OpenStack concerns its equivocation about support for the Amazon Web Services API. Lydia Leong continues: “OpenStack’s community has been waffling about whether or not they want to continue to support an Amazon-compatible API; at the moment, OpenStack has its own API but also secondarily supports Amazon compatibility.” Citrix continues its support for the Amazon Web Services API for its Apache-based CloudStack offering as well as its commercial counterpart although Amazon Web Services has yet to officially endorse it.

Turning over CloudStack to the Apache Software Foundation makes good business sense, according to many analysts. James Staten of Forrester Research noted that Citrix’s decision to hand over CloudStack to the Apache Software Foundation constitutes an astute strategic maneuver to recoup some of its investment in the $200 million Cloud.com acquisition that formed the foundation for its CloudStack offering. However, Staten observes that the question now is whether other vendors will follow suit and break away from OpenStack as well:

For a company that needs revenue now and has a more mature solution, a break away from OpenStack, while politically unpopular, is clearly the right business decision. The key question this raises is whether the move will be followed by others who need revenue now. At the fall 2011 OpenStack Design Summit there was a low rumble of discontent that the OpenStack code was not maturing fast enough.

Staten’s remarks point to concerns that OpenStack code was failing to develop at the pace required to stay commercially competitive in the IaaS space. Given the recent Amazon Web Services-Eucalyptus deal, which consolidates the positioning of the AWS API, and today’s move by Citrix to turn over CloudStack to an Apache 2.0 license, the obvious question is whether OpenStack can innovate at a pace that maintains its contributing base of supporters and roster of commercial deployments. Once seen as a possible counterweight to Amazon Web Services that offers the lure of cloud-interoperability, OpenStack is running up against the clock ticking on the maturity of its code as well as the robustness of its API as an alternative to the Amazon Web Services API.