Understanding Cloud Computing

Open any popular IT publication and you’re bound to see countless articles on how cloud computing is going to change everything from application management to climate change. With all these claims, it can be hard to understand what cloud computing really is and how it will impact network management.

In this article, we’ll cut through the hype and provide practical advice for monitoring the cloud.

What’s Old is New

The first thing to understand is that we’ve been here before. Terms like “private clouds” refer to existing practices of running applications over the Internet and across WAN connections. Going back even further, you might recall when time-sharing was the prominent model of computing. By allowing a large number of users to interact concurrently with a single computer, time-sharing dramatically lowered the cost of providing computing capability, and made it possible for individuals and organizations to use a computer without owning one.

The Internet has brought the concept of time-sharing back into popularity. Expensive corporate server farms and private data centers can host thousands of customers all sharing the same common resources. “New” business models, such as Software as a Service have become popular due to the cost/benefit ratio they provide customers.

Cloud Computing Platforms

In addition to the earlier mentioned term of private clouds, there are three types of public cloud computing:

Software as a Service (SaaS): Rather than buying and internally hosting software, organizations can lease the application as a service through a provider. The application is hosted, managed, and upgraded by the provider, and customers access it via an Internet connection. Examples of SaaS include Salesforce.com and Google Apps.

Platform as a Service (PaaS): Using PaaS, customers lease a platform where they develop their code, and the provider uploads and presents it on the web. Essentially, the organization is leasing raw computer power and the environment to develop applications in a quick and cost-effective manner. Examples of PaaS include Microsoft Azure and Salesforce.com’s Force.com.

Infrastructure as a Platform (IaaS): IaaS is a pay-as-you-go model for leasing the computing infrastructure for operating intense applications. IaaS provides extremely flexible scalability for handling sudden application demand shifts. For applications that require a lot of bandwidth and resources but are only used for a short time, this can be an excellent option. Examples of IaaS include Amazon EC2 and Rackspace.

Loss of Control
When implementing any type of cloud service, an organization is taking an application that was internally hosted and introducing new players–the ISP and cloud provider–who now control certain aspects of the service delivery and performance. When the application suffers any type of performance delay or users complain about slow application response, how does the network team deal with these new partners? Be aware of the types of performance reporting the cloud provider offers, and the ability of your company’s monitoring tools to integrate with these reports. Also, when dealing with an IaaS or PaaS provider, service level agreements (SLAs) become key to having any recourse if you’re plagued by performance issues. In the case of many SaaS vendors, effective SLAs may not be an option.

Loss of Visibility
In addition to giving up control over application performance, organizations need to evaluate whether visibility and analysis of performance will suffer by moving to the cloud. Are current tools able to provide similar monitoring to connections moving across the WAN and Internet? In monitoring cloud-based activities using the Observer platform, it provides key metrics for managing performance across the WAN/Internet, SLA monitoring and enforcement, ensuring application precedence, and pinpointing the cause of delay to the internal network, ISP, or cloud provider.

Although cloud computing might seem daunting, the technologies are largely based upon existing network concepts. Migrating to the cloud requires a shift in mindset, but uses many of the same solutions you have for managing your internal network and infrastructure. As the next step in readying your network and engineering team for the migration, read the Network Instruments’ Tech Advisory on Preparing for the Cloud.

Linking Cloud Reports into ORS

In shifting to cloud services, organizations need to be sure their monitoring solutions provide visibility into traffic as it flows from their internal network over the ISP to the cloud service provider. SaaS providers like SalesForce.com provide web-based reports for tracking service and server performance.

Observer Reporting Server allows you to logically integrate these web-based performance reports into your troubleshooting workflow.

First, we’ll create the report that displays the URL.
1) From the ORS main screen, click on the Reports tab and Network Trending Reports tab then select the New button to create the web report.

2) Within the Report Setup window, name the report, check the Redirect to external website box, and input the entire URL address including http://. Click the Ok button.

Secondly, we’ll incorporate the URL report as a part of the drill-down process from the current status element.
1) Within the report layout, double-click on the specific current status element tracking the cloud service.

2) In the Current Status Element Setup window, click the Select button next to the title: ORS drill down report. This allows you to assign the report you just created as a drill-down action.

3) In the Select Drill Down Reports window, select the report you created and drop it into the Selected Reports for Drill Down window. Select Ok button to complete the action.

After tying the reports together, upon discovering a problem with cloud services you can drill from the current status element to the cloud provider reports.

This example shows Saleforce response time is slower than desired. Next, click on the white arrow icon to choose from a list of reports for investigation, including the new SalesForce Trust Link report.

The providers’ reports are now a part of the troubleshooting workflow allowing you pinpoint or rule them out as the source of problems.

New NI University Courses

Looking to sharpen your troubleshooting skills? Check out the new NI University courses. We’ve recently launched classes on Application Analysis, VoIP Troubleshooting, and GigaStor Data Mining and Forensics. Learn more.

© 2010 Network Instruments, LLC. All rights reserved.
Network Instruments, Observer, GigaStor, and all associated logos are trademarks
or registered trademarks of Network Instruments, LLC.
All other trademarks, registered or unregistered, are sole property of their respective owners.