Advantages & Disadvantages of Webservices RRS feed

  • Question

  • Here are some general pros and cons..




    Web Services offer many benefits over other types of distributed computing architectures.

    • Interoperability - This is the most important benefit of Web Services. Web Services typically work outside of private networks, offering developers a non-proprietary route to their solutions. Services developed are likely, therefore, to have a longer life-span, offering better return on investment of the developed service. Web Services also let developers use their preferred programming languages. In addition, thanks to the use of standards-based communications methods, Web Services are virtually platform-independent.

    • Usability - Web Services allow the business logic of many different systems to be exposed over the Web. This gives your applications the freedom to chose the Web Services that they need. Instead of re-inventing the wheel for each client, you need only include additional application-specific business logic on the client-side. This allows you to develop services and/or client-side code using the languages and tools that you want.

    • Reusability - Web Services provide not a component-based model of application development, but the closest thing possible to zero-coding deployment of such services. This makes it easy to reuse Web Service components as appropriate in other services. It also makes it easy to deploy legacy code as a Web Service.

    • Deployability - Web Services are deployed over standard Internet technologies. This makes it possible to deploy Web Services even over the fire wall to servers running on the Internet on the other side of the globe. Also thanks to the use of proven community standards, underlying security (such as SSL) is already built-in.




    • Although the simplicity of Web services is an advantage in some respects, it can also be a hindrance. Web services use plain text protocols that use a fairly verbose method to identify data. This means that Web service requests are larger than requests encoded with a binary protocol. The extra size is really only an issue over low-speed connections, or over extremely busy connections.
    • Although HTTP and HTTPS (the core Web protocols) are simple, they weren't really meant for long-term sessions. Typically, a browser makes an HTTP connection, requests a Web page and maybe some images, and then disconnects. In a typical CORBA or RMI environment, a client connects to the server and might stay connected for an extended period of time. The server may periodically send data back to the client. This kind of interaction is difficult with Web services, and you need to do a little extra work to make up for what HTTP doesn't do for you.

    • The problem with HTTP and HTTPS when it comes to Web services is that these protocols are "stateless"—the interaction between the server and client is typically brief and when there is no data being exchanged, the server and client have no knowledge of each other. More specifically, if a client makes a request to the server, receives some information, and then immediately crashes due to a power outage, the server never knows that the client is no longer active. The server needs a way to keep track of what a client is doing and also to determine when a client is no longer active.

    • Typically, a server sends some kind of session identification to the client when the client first accesses the server. The client then uses this identification when it makes further requests to the server. This enables the server to recall any information it has about the client. A server must usually rely on a timeout mechanism to determine that a client is no longer active. If a server doesn't receive a request from a client after a predetermined amount of time, it assumes that the client is inactive and removes any client information it was keeping. This extra overhead means more work for Web service developers.

    Monday, November 19, 2007 4:02 PM