Hyperic, known for its tools that let users monitor and manage their applications running on the Web, has launched a new hosted service for users of Amazon Web Services. In this case, however, it might be a good idea to first say what the free service does not do. It does not provide users with specific information on how their particular application is doing. It does not tell them what’s going on in their pieces of the cloud. It does not dulcetly intone, “Excuse me, Dave, but I have detected a throughput problem in storage sector 6.”
Not yet, anyway.
What Hyperic’s new CloudStatus.com does right now is give users an independent, graphical dashboard view into the “general health” of AWS’s Elastic Compute Cloud (EC2), Simple Storage Service (S3), and other elements of the big utility. The idea, the company says, is to provide a comprehensive, overall measure of availability, throughput and latency.
“What CloudStatus.com does is basically break down each service in AWS and help to answer some of the questions cloud users and developers would have,” says Javier Soltero, co-founder and CEO of Hyperic. “For example, measuring how long it takes to launch a machine image across the various availability zones. We track the time it actually takes to provision new instances and track that historically, then present that data graphically. For the S3 storage system, we’re able to look inside and outside the cloud and measure the storage and retrieval throughput in various zones. We can show how effective the service is, and developers or potential customers will know if it meets their application’s requirements.” CloudStatus.com also provides graphs of measurements from Amazon’s Simple Queue Service, Simple DB database service and Flexible Payment system.
Hyperic deploys agents at multiple global points in the cloud that provide real-time and historic data. “You can actually watch the trends, the different oscillations, if you will,” Soltero says. “We think we get as close as you can to approximating what the individual user’s experience would be.”
“Over time we’ll add other clouds and services, including Google AppEngine, force.com, and so on. We want to be able to provide whatever depth of measurement is needed to determine the health of your cloud operation,” Soltero says.
The company plans regular updates that over time should evolve CloudStatus.com into a global monitoring-and-managing tool like operators of old-school networks have. “We will roll out more personal monitoring features in the next few months, like the ability to get an e-mail alert if something you have running in the cloud is not meeting service levels, or in response to some other performance requirement,” Soltero says.
CloudStatus.com is based on top of Hyperic HQ, the company’s system for monitoring and managing Web infrastructure. (Hyperic’s technology is used by clients running some very large Web sites, such as CNET, and by datacenter operator Rackspace’s cloud division, Mosso.) HQ’s extensible design allows addition of new agents that can work in new environments, like the cloud, Soltero says. The HQ server can communicate with these agents and collect data whether they’re in the local datacenter or in the cloud. Hyperic HQ’s ability to reach out into more traditional networks as well as into cloud-based services will make it a good tool for companies that want to operate with a hybrid network, or want to closely watch performance as they migrate to the cloud.
CloudStatus.com can’t do this yet, but Hyperic’s goal is to provide a single scope that shows what’s going on with all your cloud-based services, and in greater detail than currently available. Hyperic HQ will be able to provide personal data feeds, sort of a “MyCloudStatus,” Soltero says.
Hyperic expects CloudStatus.com will serve a bigger purpose: resolving some concerns some companies might have about running in the cloud. “Without proper visibility into the cloud, many organizations will be reluctant to adopt the cloud for business,” Soltero says. “Applications that depend even partially on cloud-based services need the same monitoring and management they get in the more traditional network. That’s what we think we’re going to give people with CloudStatus.com.”
Steve Brasen, analyst with Enterprise Management Associates, says CloudStatus.com “is a step in the right direction,” adding that much of the concern limiting cloud computing deployment revolves around supportability and provisioning, and if you consider the extensive number of systems needed to support a cloud infrastructure, the need for an automated monitoring solution becomes critical.
“Since cloud implementations are often intended for critical business drivers, they must attain a high level of reliability to convince business executives of their value. That level of trust only can only be achieved through an independent performance evaluation,” says Brasen. “Hyperic’s solution certainly provides the verification, but it will be up to the cloud vendors to prove long-term high-availability and reliability. The next step is the development of reliable standards against which cloud infrastructures can be audited.”
First and foremost, Hyperic addresses the issue of creating a third-party source to do performance monitoring of AWS, RedMonk analyst Michael Cote tells GRIDtoday. “Admittedly, there are many more other metrics and tests one would want to see. But this, after all, is a first release.” As Cote wrote on his blog at redmonk.com, CloudStatus “starts the ball rolling” toward genuine cloud management by providing two important fundamentals: visibility into EC2 and historical tracking of performance. While it doesn’t do this latter function “extremely well (you’re restricted to the date range selected in each graph), Hyperic is building up a historical record on AWS’s uptime,” Cote says. “Getting a handle on the performance trends for stuff in the cloud is key for people who want to build enough trust in the cloud to actually use it.”
Now, whether the measurements this initial release of CloudStatus.com currently provides are of value to people making business decisions can be answered only by people who make business decisions. It will be a different story if CloudStatus develops into a tool that provides deep performance details on all of a user’s operations and across different clouds. “We’re tying to provide the equivalent of shining a very bright light in an otherwise darkened room,” Soltero says.
CloudStatus.com is not yet the eyes in the cloud for which users, developers, and operations managers might yearn, but it is a big step beyond Amazon’s current red light/green light.