Cloud Cruiser: Bean counting in the heavens
Attack of the accountants
There seem to be a zillion companies that want to help you build cloudy infrastructure, but Cloud Cruiser, a startup that has just come out of stealth mode, wants to help you do something else: keep track of the money you are allocating for clouds.
Cloud Cruiser was founded in January 2010 by Dave Zabrowski, who among other things was in charge of Hewlett-Packard's US PC operations a while back and eventually rose to be general manager of HP's Enterprise Computer Organization in the pre-Compaq days. In 2002, Zabrowski was hired to be chief executive officer at 10 Gigabit Ethernet network interface card maker Neterion, which was just shut down last week by its owner, Exar, which bought Neterion for $11m in February 2010; the company had raised more than $42m in venture funding over the years, pushing the envelope for 10 Gigabit Ethernet technology on servers.
Only seven months after its founding, Cloud Cruiser lined up an undisclosed bag of cash from Wavepoint Ventures, and also has a bunch of Silicon Valley angel partners that have helped get the software and services company on its feet. (Zabrowski used to work at Wavepoint, according to a report in the Sacramento Business Journal. Zabrowski is CEO at Cloud Cruiser, but co-founder Greg Howard is the expert on chargeback methodology for IT infrastructure. Howard was director of development at CIMS Lab, creating a middleware program that kept track of IT assets and did chargeback into accounting systems. IBM bought CIMS Lab in 2006 and has merged it into its Tivoli systems management stack.
Zabrowski knows that people are going to use a mix of physical and virtual assets internally as well as hybrid and public clouds, and his goal with Cloud Optimizer, the company's first product, is to provide a dashboard that shows what resources are being consumed across the company, no matter where they physically or logically reside. Having figured out where everything is and what it costs - this is often called showback - then the next step is to allow IT managers to do chargeback. Once different lines of business and divisions see what their IT habits cost, then comes the final step, which is to look at the entire cloudy infrastructure stack and make decisions about costs, performance, and security to let lines of business actually make their own decisions about where they want to plunk their workloads to reduce their own expenses.
If that doesn't scare the hell out of CIOs, it is hard to imagine what will.
But there is a plus side to doing cloud bean counting: CIOs have tight control over internal resources, but less and less over external resources as more business units and individuals "go rogue" and buy their own IT on clouds such as Amazon's EC2. By allowing their clients - the managers of company lines of business - to make a choice between internal and external cloudy infrastructure, CIOs have to compete and justify their existence. And by having a tool like Cloud Optimizer, CIOs can have a single pane of glass to count all the bits and beans to help the business units make proper decisions. And aside from that, if the CIOs don't put such chargeback tools into place that span data centers and external clouds, CFOs will. And the bean counters might start wondering why they pay so much for their CIOs. . . .
Zabrowski says that there are plenty of companies that are "cloudwashing" their system management tools. "But they are not suitable for managing clouds," he says, and that is because in many cases they don't look at the money side of the IT equation or they only do chargeback for one particular stack, as is the case with the add-ons for VMware's vSphere virtualization tools.
Cloud Optimizer has a number of different modules. The Cloud Collect module can accept input data from just about any device: virtual and physical servers, network devices, storage, and so forth. "We can take feeds from anything," says Zabrowski. "Even your telecom usage and costs."
The Cloud Map module aggregates the data feeds from all the devices into a single logical view. Cloud Charge is what keeps track of usage and does the chargeback for all of the resources in the internal and external clouds. Cloud Report is the data analytics engine that sites on top of this usage data, allowing you to drill down into it to generate reports. And Cloud Share is a set of APIs that let Cloud Optimizer link into other management and monitoring tools. The tool is created using a mix of technologies, including the Google Web Toolkit for its graphical user interface, a mix of Java and the Spring framework for its code, Jasper reports for reporting, a slew of REST APIs, and the MySQL database.
Right now, Cloud Optimizer has some limits of its own, in that it only supports x64-based systems running Hyper-V, KVM, or VMware hypervisors. Cloud Cruiser is in the middle of certifying its Cloud Optimizer to work with the popular public clouds, but Zabrowski is not willing to name names just yet. It would be nice if it supported Unix, mainframe, and other proprietary systems used in the data center, and XenServer would seem to be a necessary hypervisor as well.
Cloud Optimizer is being offered as a test drive during its beta phase for $3,000 a month for a 90-day alpha period. Zabrowski says that on on-site license will be priced in the five to six figures per year, with the cost scaling with the number of input feeds, number of devices, and amount of data that Cloud Optimizer has to chew on. Cloud Optimizer will go into beta testing on June 1, and is expected to be generally available at the end of August. ®