Network Management Strategies for the CIO
We have a web server located in a data center, and our customers access our webpage over the Internet. A few customers report slow web page access (application performance degradation), but within the data center there doesn't seem to be a problem. How can I determine whether there is an ISP connection problem causing slow web page access for some customers? What network troubleshooting tools can I use to track down the problem?
Requires Free Membership to View
Ah. The ever elusive 'it's the network problem.' I am so fond of this problem – and it's a good thing, since that's what I do every day!
There are several major groups where a web-based application hosted over the Internet may experience performance degradation and other problems causing slow web page access. These major categories fall into end-user workstation, network, server, or application silos. While there are several products in the market place that will provide comprehensive analytical data once you've identified the major category, it's essential that you first identify whether it's the client PC, network, server, or application itself.
To best quantify if the application performance degradation occurs in the network, server or application arena, the least intrusive way to really understand the nature of the application is to use troubleshooting tools to do a packet capture or analyze the traffic on the switch the servers are directly connected to. If you use a port mirror to point the traffic to a packet analyzer, you would have the ability to get a view of the traffic going to and from the server without having to instrument anything on the client machines.
There are several network troubleshooting tools on the market that will help you with this step, including but not limited to Wireshark (free), Network Instruments Observer (download a demo), and Wild Packets OmniPeak. If you want to capture response time statistics that provide a comprehensive report about the response times experienced without having to analyze the packets yourself, I would recommend you investigate options like NetQoS SuperAgent or OpNet's Ace Live. These products specialize in helping you analyze whether the delays experienced across the enterprise are related to the network, server, or application.
These types of network troubleshooting tools are used to help determine if the Internet-facing application is experiencing network delays due to Internet brown-outs or congestion constraints imposed by the connected network. While it's easy to troubleshoot when an application is just inaccessible, slow performance tends to be more difficult to understand and quantify. First identify the major area in which the problems are occurring, and then use more advanced troubleshooting tools to gain better insight into the root cause of the hosted application problems.
There are several major groups where a web-based application hosted over the Internet may experience performance degradation and other problems causing slow web page access. These major categories fall into end-user workstation, network, server, or application silos. While there are several products in the market place that will provide comprehensive analytical data once you've identified the major category, it's essential that you first identify whether it's the client PC, network, server, or application itself.
To best quantify if the application performance degradation occurs in the network, server or application arena, the least intrusive way to really understand the nature of the application is to use troubleshooting tools to do a packet capture or analyze the traffic on the switch the servers are directly connected to. If you use a port mirror to point the traffic to a packet analyzer, you would have the ability to get a view of the traffic going to and from the server without having to instrument anything on the client machines.
There are several network troubleshooting tools on the market that will help you with this step, including but not limited to Wireshark (free), Network Instruments Observer (download a demo), and Wild Packets OmniPeak. If you want to capture response time statistics that provide a comprehensive report about the response times experienced without having to analyze the packets yourself, I would recommend you investigate options like NetQoS SuperAgent or OpNet's Ace Live. These products specialize in helping you analyze whether the delays experienced across the enterprise are related to the network, server, or application.
These types of network troubleshooting tools are used to help determine if the Internet-facing application is experiencing network delays due to Internet brown-outs or congestion constraints imposed by the connected network. While it's easy to troubleshoot when an application is just inaccessible, slow performance tends to be more difficult to understand and quantify. First identify the major area in which the problems are occurring, and then use more advanced troubleshooting tools to gain better insight into the root cause of the hosted application problems.
This was first published in February 2010
Join the conversationComment
Share
Comments
Results
Contribute to the conversation