Very easy to use and gather value from the start. Easy to deploy on servers with little impact. Easier to gather information from the Server than from the Network Switch when there are multiple servers to combine for an application study.
Room for Improvement:
Marketing is complex and not well understood, so often this product is not considered a good fit just for a surface look. Applications are getting more complex as there are islands of services that are combined to provide value to the customer. The response time delay is hidden within service calls that are being made. This family of tools provides the insight needed to know where that response delay is happening. It also provides insight as to the number of service calls that are being made. While each service call may be short, doing 1,000 of them, brings major delays to the overall response time. Finding where that delay is happening is key to improving the application. These tools provide that insight and provide the ability to show the total response time that is taking place. As a side note:
Often I used the Capture side of the product more than the Analysis side of the product. The ability to do captures without having to use tcpdump on each system is very helpful. This ability takes the grunt work out of gathering the basic information that is required for deeper analysis.They (Opnet) also use Wireshark in their analysis phase and then add even more analysis for some deeper dive information. This is helpful for general analysis. Also some of their decodes have been helpful as well.
Use Cases: SOA brings a whole new meaning to the word distributed processing. Finding where work is really happening when an application is using SOA concepts is a challenge. In some cases, multiple traces are required as each trace reveals other servers and services that are engaged that were unknown to the original calling program of the SOA service. These are buried within the process of calling the first SOA service.