This is a guest post written and contributed by Tanya Bragin, Lead Product Manager at ExtraHop Networks, a Rackspace Cloud Tools Partner. ExtraHop provides application performance management (APM) solutions to help ensure business-critical transactions do not fail.
Companies have a number of reasons to deploy their applications in the cloud: scalability, simpler management, and cost-efficiency are among the top benefits. It’s true that there are situations where traditional on-premises deployments are preferable, but technological advances such as more plentiful bandwidth and server virtualization have made the cloud an enticing new alternative. This is the great thing about the IT industry—nothing stays the same for long when processing power, network bandwidth, and storage capacity increase exponentially.
In the same way that IT trends changed the way people deploy their applications, they are also changing the way that people monitor those applications. Traditionally, IT departments harvested performance metrics from systems all along the application delivery chain. With relatively simple client-server architectures that changed infrequently, this approach worked fairly well.
However, today’s application environments are increasingly heterogeneous, distributed, and dynamic. There are more systems in the mix and more device types. Applications change more rapidly while the underlying infrastructure is constantly in flux. Datacenters must deal with a flood of Big Data, proliferating virtual machines, and geographic distribution. Additionally, it seems like everyone and their brother is developing a mobile version of their application.
This wild-and-woolly application environment often breaks or obscures traditional monitoring that uses host-based instrumentation. Agents that monitor application performance are sensitive to changes in that application, only work with tested software frameworks, and can behave in unexpected ways if not configured correctly. Since they run on host machines, agents encounter difficulties in virtualized environments.
If reading this blog post is stressing you out, relax. Necessity is the mother of invention, and smart technologists have harnessed IT trends for a different approach to application monitoring. It’s called network-based application performance management (APM) and it uses advances in processing power and storage capacity to analyze application transactions as they traverse the network instead of host-based instrumentation.
Poof! That’s the sound of your application monitoring headaches disappearing. No more worrying about software compatibility, no more overhead on your applications (or in cloud terms, no extra CPU cycles), no more tweaking configurations, and no need to reboot your virtual instances. Sure, host-based monitoring tools have their place just as traditional on-premises deployments do. But network-based APM, just like cloud computing, is a new disruptive technology that deserves examination.
The ExtraHop system, the pioneering network-based application performance monitoring solution, analyzes transactions in real-time from L2 to L7, including application-level details such as the URI included in an HTTP 500 error or a slow stored procedure in a database. The ExtraHop system decodes web, database, and storage protocols, as well as protocols for supporting services such as LDAP and DNS. One IT manager at LexisNexis calls the ExtraHop system an EKG that lets him monitor the health of his environment in real time. Leading IT organizations such as those at Adobe, Alaska Airlines, Concur, and Microsoft are already seeing great success with network-based APM from ExtraHop Networks.
We’re excited to announce a partnership with Rackspace that makes an offline version of ExtraHop technology available to Rackspace customers at an extremely affordable price: FREE. Network Timeout tool is a cloud-based transaction analysis tool that can help when tuning performance and troubleshooting network and application problems.
To use Network Timeout, just complete a free registration and upload a packet capture. You can view the analysis in an interactive web interface. Start with summaries of network, web, database, and storage performance, and then drill into specific transaction metrics for each device. Video tutorials, forum discussions, and FAQs are also available at www.networktimeout.com.