Ravello Systems Extends Its Platform For Migrating Workloads To Public Clouds To Google Compute Engine

Ravello Systems today announced the general availability of its cloud hypervisor platform for Google Compute Engine. As a result of the announcement, Ravello customers can use its SaaS nested virtualization technology to migrate workloads to the Google Compute Engine (GCE) public cloud with a few clicks of the mouse and thereby take advantage of the agility of public cloud environments to accelerate their development and testing efforts. Ravello’s process for the migration of workloads to public cloud environments simplifies migrations by preserving network topologies as well as storage and computing configurations. Customers use Ravello’s user interface to upload their virtual machines into Ravello’s private library, drag and drop them onto an “application canvas,” specify the requisite network configuration, publish the configuration of VMs to the cloud and then create an “application blueprint” that serves as a snapshot of the application at the time of upload. Developers can use the blueprint to spin up the application for parallel testing and QA purposes or connect Ravello to a continuous integration server that synchronizes the blueprint with changes to the on-premise version of the application to ensure that development and testing efforts are performed on the latest version of the application.

Today’s announcement means that Ravello now supports four public clouds in the form of Amazon Web Services, HP Cloud, Rackspace and Google Compute Engine. Integration with Microsoft Azure remains on the product roadmap as told to Cloud Computing Today in a phone interview with Shruti Bhat, Ravello’s Director of Product Marketing. The general availability of Ravello’s nested virtualization technology on GCE represents a particularly special moment for Ravello insofar as its leadership team built the KVM hypervisor that constitutes the basis of virtualization within the Google Compute Engine platform. Ravello’s partnership with Google affirms the threat Google Compute Engine poses to the IaaS market supremacy enjoyed by Amazon Web Services and additionally promises to provide data about cloud deployments within the public clouds it supports by means of the Ravello Cloud Dashboard, which features data about VM provisioning time and VM provisioning-related error rates per cloud provider. From an industry perspective, Ravello’s integration with GCE continues to underscore the importance of public cloud environments for accelerating development and testing. Other use cases for the platform include disaster recovery, backup and replication, but the core use case will revolve around an embellishment of the platform’s capabilities for dev and test in ways that leverage the flexibility of the public cloud to disrupt contemporary protocols for software testing in favor of the massive, algorithmic parallel testing enabled by Ravello’s “blueprint” concept and cloud-based, nested virtualization platform.

Joyent Fires Salvo At Amazon Web Services With Enhanced Joyent Cloud

On Thursday, Joyent announced the launch of a new public cloud computing platform that takes direct aim at Amazon Web Services in the increasingly competitive Infrastructure as a Service space. The newly improved Joyent Cloud offering from the San Francisco based company incorporates the company’s August 15 reconfiguration of its SmartOS operating system that allows users to deploy applications on Windows and Linux operating systems in addition to SmartOS. The new Joyent Cloud boasts four principal innovations to an upgraded infrastructure management system called SmartDataCenter:

Enhanced analytics
Customers will have increased visibility to the performance of their cloud infrastructure thanks to the deployment of DTrace, an open source analytics tool that Joyent had previously leveraged exclusively for corporate, troubleshooting purposes.

Safe storage, Increased speed, Data security
Joyent claims its Windows, Linux and SmartOS machines have superior processing speeds, data security standards and secure storage. SmartOS machines deliver even more granular analytics than their Windows and Linux counterparts.

Lower computational costs
Because Joyent’s servers are reportedly up to 14 times faster than comparable Amazon EC2 machines, computing costs for customers are amongst the lowest in the industry.

More predictable pricing
Joyent has shifted to a pay per use pricing model at a rate starting at $.085/hour, in contrast to their previous subscription model.

The enhanced analytics give Joyent a competitive advantage over Amazon Web Services, which has often been characterized as a black box when it comes to providing customers with visibility about the performance of their deployments. “Customers are going to get, in their user interface, the ability to measure real-time latency from the infrastructure all the way up through the application stack,” said Steve Tuck, General Manager for Joyent Cloud. Amazon Web Services customers frequently use third party applications such as RightScale in order to gain more insight into latency within a cloud stack.

Joyent customers can now deploy applications on Windows and Linux operating systems because of the company’s “porting” of the KVM hypervisor onto its SmartOS operating system. The integration of the KVM hypervisor onto SmartOS allows for hardware virtualization in addition to operating level system virtualization. Joyent founder and chief scientist Jason Hoffman hailed the new Joyent SmartOS as the “first hypervisor platform to emerge in five years” and the only cloud solution in the industry “that can manage both KVM hardware virtualization and operating system-level virtualization on a single OS.” After porting the KVM onto its hypervisor, Joyent open sourced its revamped cloud SmartOS cloud operating system. The outspoken Hoffman claimed that “this combination of virtualization options, data consistency through ZFS and access to DTrace for rapid troubleshooting, is the most powerful and efficient collection of technologies in cloud application development. I invite developers who use VMWare, Citrix, Red Hat or Microsoft hypervisor tools to try this open source package.”

Hardware virtualization refers to a scenario whereby a hypervisor enables one server to function as several servers operating independently of each other. For example, one server might thereby be virtualized into three servers, each of which runs a different operating system such as SmartOS, Windows and Linux simultaneously. Operating system-level virtualization, on the other hand, refers to a case where the operating system itself is virtualized. In this case, a virtualized operating system features discrete cases of the same operating system running independently. Speaking of the impetus for the company’s recognition of hardware virtualization in its August 15 KVM integration, Joyent’s Steve Tuck noted that “there are a lot of developers out there that say, ‘I just want Linux or I just want Windows. I don’t want to worry about a couple of small differences, even if they are minor, before I code.”

Joyent Cloud claims 13,000 customers including high profile names such as LinkedIn, Kabam, StackMob, and Gilt Groupe. The company’s aggressive push of its Infrastructure as a Service offering comes in the wake of competition from increasing OpenStack deployments, the enterprise oriented Direct Connect offering from Amazon Web Services, Dell’s announcement of a VMWare based cloud offering and HP’s forthcoming OpenStack based cloud. Joyent is a member of the Open Virtualization Alliance, an association dedicated to the adoption of the KVM hypervisor as a robust alternative to proprietary cloud solutions.