ActiveState announced the general availability of Stackato 1.0, an application for creating private Platform as a Service environments in any cloud environment, today. Stackato claims a unique status in the PaaS space because of its ability to accommodate “any language, any stack” and “any cloud.” Whereas few other PaaS vendors support all languages and technology stacks, Stackato joins PaaS vendor CloudBees in enabling its PaaS platform to run on any cloud environment such as Amazon Web Services or HP Cloud Services. Private cloud Stackato users can configure and deploy the PaaS software within a matter of minutes. Enterprises that are wary of public clouds can take advantage of Stackato’s ease of deployment to create a private cloud that boasts management utilities such as a web-based management console and performance monitoring technology from New Relic.
Stackato’s support for “any language” and “any stack” goes back to its origins as a participant in the Cloud Foundry open source project. As the Cloud Foundry Community Lead for Python, Stackato represents an “enterprise-ready, commercially supported private PaaS version of Cloud Foundry” that claims even more expansive support for languages and technology stacks than Cloud Foundry. Moreover, Stackato supports a streamlined process for data migration of legacy applications to its platform, including applications coded in UNIX. Stackato’s ability to support any technology platform, its ease of deployment and support for private cloud PaaS environments gives it a clear leadership position in the increasingly crowded PaaS space. Competitors include Red Hat’s OpenShift, CloudBees and Cloud Foundry, though few vendors match its combination of flexibility and ability to deploy private cloud PaaS environments. Today’s announcement about the general availability of Stackato 1.0 comes roughly six weeks after a release that revealed Stackato’s support for OpenStack, Linux KVM and the Citrix XenServer.
Last Wednesdsay, Amazon Web Services announced the availability of the Amazon Simple Workflow service that allows developers to automate and orchestrate discrete programming tasks in order to more effectively manage complex development ecosystems. AWS CTO Werner Vogels summarized the new workflow service in a blog post as follows:
Amazon Simple Workflow enables applications to be built by orchestrating tasks coordinated by a decider process. Tasks represent logical units of work and are performed by application components that can take any form, including executable code, scripts, web service calls, and human actions. Developers have full control over implementing and orchestrating tasks, without worrying about underlying complexities such as tracking their progress and keeping their state.
Developers implement workers to perform tasks. They run their workers either on cloud infrastructure, such as Amazon EC2, or on-premise. Tasks can be long-running, may fail, may timeout and may complete with varying throughputs and latencies. Amazon SWF stores tasks for workers, assigns them when workers are ready, tracks their progress, and keeps their latest state, including details on their completion. To orchestrate tasks, developers write programs that get the latest state of tasks from Amazon SWF and use it to initiate subsequent tasks in an ongoing manner. Amazon SWF maintains an application’s execution state durably so that the application can be resilient to failures in individual application components.
The Simple Workflow Service features the following components:
• Domain, defined as a set of related workflows
• Workflow, understood as the automation of a business process
• Deciders, that orchestrate qua manage and sequence tasks
• Tasks, that constitute “logical units of work”
At one level, Amazon Simple Workflow represents a brilliant solution to the problem of orchestration. Deciders manage dependencies between tasks and manage task scheduling and the concurrency of simultaneous, inter-related processes. As Chirag Mehtra notes, however, the Simple Workflow solution may be viewed as another service that increases vendor lock-in with respect to AWS given that the orchestration specific to their application ecosystem runs within a managed service offering uniquely configured by Amazon Web Services. Developers seeking to migrate an application from AWS to another cloud environment will need to rebuild the orchestration apparatus architected using Simple Workfow in order to enjoy the benefits of automated workflow coordination and management.
This week, Infochimps revealed details of its Infochimps Platform for deriving business value from Big Data sets. The Infochimps Platform complements its data marketplace for over 200 data sets from companies such as Twitter, FourSquare and OkCupid for enterprises seeking to embed these data sets into their own software applications. Hosted on the Amazon Web Services cloud or within a customer’s private cloud, the heart of the Infochimps Platform consists of Ironfan, a technology that allows customers to configure and operationalize a Big Data stack quickly. Users can leverage Ironfan to determine which databases and analytic applications need to be selected to optimize their Big Data analyses.
An open-source technology developed by Infochimps, Ironfan allows business users to seamlessly scale and re-architect their Big Data processing infrastructure as their Big Data stack evolves. Ironfan constitutes the technology that enabled Infochimps to develop a “special social influence score” called Trstrank based on 20 million tweets per day. Manually configuring the server infrastructure for Trstrank on Amazon Web Services would have taken Infochimps four weeks. Configuring the technology stack for Trstrank using Ironfan, however, took only two hours.
Key features of the Infochimps Platform include:
• Apache Flume to manage delivery of data sets from point A to point B
• Support for databases such as HBase, Cassandra, Elastic Search, MongoDB and MySQL
• Elastic Hadoop that allows users to access only as much Hadoop resources as required
• Analytics that leverage Pig, Wukong (Infochimps) and other Hadoop-compatible analytics software frameworks
Infochimps will continue to offer its data marketplace alongside its turnkey Big Data platform to promote its mission of democratizing access to Big Data. The platform aptly illustrates the growing trend of the convergence of Big Data and cloud platforms. IBM, Microsoft, Oracle and Karmasphere join Infochimps in promoting cloud-based Big Data platforms. The Infochimps Platform for Big Data also represents the most recent example of a Platform as a Service infrastructure running on an Infrastructure as a Service public cloud.
A schematic of the Infochimps Platform architecture can be found below:
At the 2012 Parallels Summit, Jelastic announced its Platform as a Service offering of the same name will exit beta and become generally available in March 2012. Geared toward Java applications, Jelastic is offered through the U.S. and German hosting partners ServInt and dogado respectively. Jelastic’s availability through hosting partners means that customers need not manage complex installations of the Jelastic PaaS platform. In a manner similar to AWS, customers can deploy a new cloud environment in seconds with a few clicks of the mouse. Jelastic’s configuration technology handles automated connections between a customer’s server instances and the software stack selected within the hosting environment.
Key features of Jelastic include the following:
• Capability to “run and scale any Java application with no code changes required.”
• The flexibility to select customized software stack groupings such as Tomcat, Glassfish, Jetty and JBoss Java application servers and SQL (MariaDB, PostgreSQL, MySQL) or NoSQL (MongoDB, CouchDB) database platforms.
• A utility pricing model that charges on the basis of RAM and CPU consumption instead of machine size.
• Pricing in units of 128 MB, called cloudlets, of $0.02 per cloudlet per hour from ServInt, and 0.016 euro per cloudlet per hour from dogado.
• Ease of deployment including no installation and autoscaling.
Jelastic’s utility pricing model is revolutionary within the PaaS landscape and leverages its technology for automatically allocating resources in relation to consumption, allowing partners Servint and dogado to charge for usage alone. ServInt COO Christian Dawson commented on Jelastic’s utility pricing model by noting: “We think the near-utility approach we’re taking with Jelastic on pricing and provisioning is going to break the platform-as-a-service industry wide open. The product is miles ahead of its competitors on technological features, and now we’re adding game-changing, customer-friendly pricing to the mix.” Jelastic Beta customers will enjoy two weeks of free service following the March launch date for the paid service.
Jelastic’s announcement comes days after CloudBees revealed details of AnyCloud, which allows customers to deploy the AnyCloud PaaS across multiple hosting environments. Meanwhile, ActiveState’s Stackato recently announced a new release featuring the ability to deploy applications to private and public clouds as the battle for market share in the PaaS space intensifies.
Rackspace has partnered with Redapt to deliver Rackspace Cloud: Private Edition, featuring deployment of an OpenStack-based private cloud offering in addition to support. Under the partnership, Redapt will deliver hardware installed with OpenStack that conforms to Rackspace’s infrastructure specifications. Redapt will perform all hardware-related integration testing off customer premises such that the “hardware will arrive at a customer’s data center of choice in as few as 14 days after ordering, needing only power and network connectivity.” While the Rackspace Cloud: Private Edition is based on Rackspace’s customized version of OpenStack, the San Antonio based company is looking forward to offering support for other deployments of OpenStack from vendors that similarly deliver commercial OpenStack deployments. Partnerships with other commercial OpenStack distributors would allow other vendors to take advantage of Rackspace’s legendary support services while increasing their distribution channel. Meanwhile, Rackspace’s partnership with Redapt illustrates the commercial interest in enterprise grade private clouds based on the OpenStack cloud operating system.
This week, PaaS vendor CloudBees announced the availability of AnyCloud, a platform that empowers customers to enjoy applications across a range of IT infrastructures including on-premise data centers and hosted data center environments. Whereas most PaaS platforms require the installation of software within a private cloud environment, AnyCloud enables customers to deploy applications in diverse environments such as Amazon Web Services, other IaaS hosting environments and enterprise data centers. AnyCloud manages the deployment of applications across different environments and thereby allows customers to differentially deploy applications in different hosting environments in order to:
• Comply with regulatory policies specific to a geographic region
• Comply with company policies about which data centers can be used for application deployment
• Minimize data latency and security issues
François Déchery, CloudBees Vice President of International Business Development, explained the significance of AnyCloud by noting: “As we worked with customers around the world, CloudBees was being asked more and more for deployment options across various IT environments and hosted providers – particularly in Europe. We decided to satisfy the demand for flexible deployment options in a very different way than our competition.” AnyCloud supports any JVM-language including Java, Spring, JRuby, Grails, Scala, Groovy and others.
AnyCloud illustrates the emerging co-implication of PaaS platforms within IaaS infrastructures. This week, for example, Nimbula’s IaaS Nimbula Director 2.0 platform revealed support for third party PaaS platforms including the VMware Cloud Foundry PaaS. Current CloudBees customers include Digg, The Gerson Lehman Group, Lawrence Livermore National Laboratory, Netflix, Symantec and Yale University. CloudBees was founded in 2010 by Sacha Labourey, former CTO of JBoss.
On Monday, Nimbula announced Nimbula Director 2.0 with a markedly enhanced version of its cloud operating system software for public and private clouds. The platform features enhanced capabilities to integrate with third party software in the form of storage repositories, PaaS platforms, cloud security and management applications and other software. Notably, Nimbula Director 2.0 supports the VMware ESXi hypervisor and VMware Cloud Foundry PaaS. The platform also boasts improved application lifecycle management capabilities involving the ability to automate the provisioning of applications and monitor application performance over the duration of their operation. Nimbula Director 2.0 is currently in Beta and debuts in general availability mode in March 2012. With this release, Nimbula dramatically expands the flexibility of its operating system by way of its support for Cloud Foundry and VMware’s ESXi hypervisor. Customers can now integrate a pre-existing PaaS platform with Nimbula’s IaaS cloud operating system. Nimbula Director 2.0 appears to have hit upon an increasingly common theme in contemporary IaaS branding, namely, the delivery of the performance and scalability of a public cloud for private clouds of varying sizes and development needs.