Whether it’s to check emails with our morning coffee, book a hotel room or run a global organization, everyone relies on the Internet, and its effective operation is now a crucial aspect of businesses around the world. With over three…

The post Keeping Data Safe: How Comprehensive Data Center Protection Solutions Can Maximize System Reliability appeared first on The Data Center Journal.

 

The post ServerHub Goes Sleepless in Seattle. Expanding into the Emerald City. appeared first on Web Hosting Talk.

 

Selamat Datang!

We are so excited to announce the arrival of one of the first enterprise public clouds to Jakarta, Indonesia!
A thriving, exciting city and region, home to more than 250 million people! Expand your business to Indonesia today, on the production ready cloud (and get 50% off, this week only!)

Based in the heart of Jakarta, our platform offers fantastic local peering and …

The post dediserve expands into Jakarta, Indonesia appeared first on Web Hosting Talk.

 

The post Server Mania Commits to Excellence with the Latest Intel Xeon Processor appeared first on Web Hosting Talk.

 

Application performance is critical to accomplishing business objectives, and data center managers are tasked with optimizing application performance to the fullest. But application performance depends on infrastructure support, and application- and infrastructure-management teams typically reside in different organizations and use separate, nonintegrated tool sets—there is no system that monitors application performance and analyzes from the application down into the supporting infrastructure such as network, storage, compute resources and, especially, virtualization. Owing to the recent momentum behind software-defined and hyperconverged data centers, physical-infrastructure resources are highly virtualized and abstracted with the most dynamic allocation for virtualized applications. As a result, infrastructure-management teams simply lack the real-time application visibility required to ensure that data center infrastructure is optimized for application performance.

This article looks at the challenges of visualizing, analyzing and optimizing infrastructure for virtualized-application performance and discusses the requirements for a solution that brings application-performance visibility to infrastructure management.

Application Performance: It’s About the Infrastructure

Assuming an application’s code is working properly, the performance of the application ultimately depends on physical infrastructure. In virtualized environments, shared physical resources (CPU, memory, storage and networking) support virtual machines (VMs), and those resources may be overcommitted—especially during peak usage times. For example, a VM may be underprovisioned because it is configured with inadequate memory, a situation similar to that of an application running on a physical server with inadequate memory. The impact is constant memory swapping, which slows application performance.

Even if a VM is properly provisioned with sufficient memory, the server’s physical memory may at times be insufficient to support multiple VMs. In this case, the hypervisor must perform memory swaps and suspend VM execution, also reducing application performance. As is true for memory availability, so too do CPU cycles, storage IOPS and network bandwidth affect VM performance.

Data Center Management Silos

The problem with determining how infrastructure configurations affect application performance is that data center management teams operate in silos of responsibility, and nobody has the overall picture. Application teams use application-performance-management (APM) tools, and operations teams use infrastructure-performance-management (IPM), and network-performance-management (NPM) tools.

  • APM solutions focus on identifying application-performance issues caused primarily by the application code, such as Java, .Net and PHP scripting. APM solutions are typically limited to applications developed in house instead of off-the-shelf alternatives. Issues resulting from infrastructure-resource overcommitment, such as congested storage devices or inadequate memory configurations, are not identifiable by APM solutions because these tools look only at the application code for custom (developed in house) application servers. A virtualized data center, however, could have hundreds or thousands of application servers, many of them running off-the-shelf applications.
  • IPM and NPM solutions both suffer from limited visibility into the entire infrastructure. The former focus on server CPU, memory and storage allocations, whereas the latter manage network activity and bandwidth. Neither has application visibility and neither offers correlation across the entire infrastructure.

Without application visibility, the infrastructure team can only react to application-performance issues. They lack prevention capability as well as planning guidance for improving application performance and infrastructure reliability. Furthermore, when application-performance problems do happen, identifying the root infrastructure cause means getting the application team and the infrastructure teams (server admin, network engineers and storage experts) together with their own favorite but nonintegrated tools to compare notes on paper. This process is painful, time consuming and often fruitless.

application performanceFigure 1: Virtualization increases the application-visibility gap for infrastructure teams.

For example, a company’s e-commerce website may occasionally suffer performance degradation in the form of slow end-user response time. The infrastructure-management team investigates web-server VMs, physical-resource usage, storage performance and network utilization, but fails to discover the root cause. In this case, the real root cause resides in a virtualized storage device heavily shared by many virtual applications, including the back-end database server used by the e-commerce web server. APM tools wouldn’t identify the virtual-storage IOPS usage collision by various virtualized applications, and IPM and NPM tools won’t be able to correlate infrastructure (storage IOPS) metrics with application performance because they entirely lack application visibility.

Solution Benefits

Eliminating the application blind spot for infrastructure-management teams requires a solution that monitors application performance and correlates it with physical and virtual infrastructure. Such application-aware IPM (AA-IPM) tools allow infrastructure teams to visualize, analyze and optimize application performance.

infrastructureFigure 2: AA-IPM tools bring application-performance awareness to infrastructure-management teams.

  • Visualization: An AA-IPM tool provides complete real-time monitoring of data center performance, from the application’s performance down to the physical and virtual infrastructure. Users can see at a glance where the physical and virtual infrastructure is failing to provide adequate support for each application, whether it’s a custom or off-the-shelf application.
  • Analysis: An AA-IPM solution delivers immediate root-cause identification with correlated real-time and historical performance evidence of applications, networks and compute/storage infrastructure.
  • Optimization: IT operations lack sufficient information to identify infrastructure-upgrade opportunities in order to improve or prevent degradation of application performance. An AA-IPM tool enables detailed and correlated trend analysis and system health status on applications, CPU utilization, memory usage, storage usage and network resources. By checking a real-time performance dashboard, users can identify trouble spots and correct them before they cause serious problems.

Requirements for a Solution

First and foremost, an AA-IPM solution should provide complete visibility into applications as well as virtual and physical infrastructure performance. It should identify application-performance problems down to the root cause. This process requires some form of deep packet inspection and virtualization integration.

Second, the solution should be easy and nondisruptive to deploy. Some solutions require the use of agents, which reside inside an application server. Agents are tedious to deploy and maintain from version to version, especially in a virtualized data center, which may contain hundreds or thousands of application servers. More importantly, the infrastructure team lacks the access authority (and neither do they want it) to reach inside application servers to deploy agents.

Finally, the AA-IPM solution should employ big-data analysis to identify trends and anomalies, extract underlying behavior patterns, and deliver hidden operational insights.

Supporting Business Goals

AA-IPM solutions enable such business goals as improving IT efficiency and application performance. They actively identify the cause of poor application-server response time down to a specific server or storage device, making it relatively easy to upgrade infrastructure to better support business goals.

In addition, AA-IPM solutions improve uptime and customer satisfaction by enabling IT managers to optimize application performance. Issues can be spotted and resolved in minutes rather than the hours or days it would have taken before.

Businesses rely on application performance, and lost productivity or unhappy customers result when they fail to optimize this performance. AA-IPM solutions allow IT managers to spot, analyze and correct application-performance issues so the company is always running with optimum resources.

Leading article image courtesy of Dennis Van Zuijlekom under a Creative Commons license

About the Author

Chia-Chee Kuan is CEO of Uila, Inc. He has a distinguished 20-year-plus career developing unique intellectual property and building game-changing companies focused on computer-networking technologies and industries. From early pre-Internet TCP/IP stack and standards development to IP multicast video-streaming applications to Wi-Fi security, Chia-Chee is a pioneering technologist who has been granted 14 U.S. patents. After cofounding AirMagnet in 2001 and merging with Fluke Networks in 2009, he was named General Manager of the Wireless Business Unit, doubling revenue during his four-year tenure at Fluke Networks. Before AirMagnet, he was a founding member at Precept Software, which was later acquired by Cisco Systems, where he continued to lead the Cisco Video Internet Service Unit product-development team. Chia-Chee holds a master’s degree in computer science from Stanford University and a bachelor’s degree in information engineering from National Taiwan University.

The post Bridging the Gap Between Application Performance and Infrastructure Management appeared first on The Data Center Journal.

 

1 Feb 2016,Kochi-

Ezeelogin, a ssh gateway software for securing access to Linux servers and a centralized Linux server management software is happy to announce the release of the new version 6.6.0

The post Secure SSH Gateway & centralized server management software Ezeelogin launches 6.6.0 appeared first on Web Hosting Talk.

 

1 Feb 2016,Kochi-

Ezeelogin, a ssh gateway software for securing access to Linux servers and a centralized Linux server management software is happy to announce the release of the new version 6.6.0

The post Secure SSH Gateway & centralized server management software Ezeelogin launches 6.6.0 appeared first on Web Hosting Talk.

© 2012 Webhosting news Suffusion theme by Sayontan Sinha