DCOM Explained
by Rosemary Rock-Evans Digital Press ISBN: 1555582168 Pub Date: 09/01/98 |
Previous | Table of Contents | Next |
|
We have looked at the platform support for DCOM; the next big question most people are likely to ask is does it work over the Internet? This chapter thus looks at Microsofts approach for supporting the Internet, the products it has, the strategy it recommends its users take, and the links with DCOM.
We will be starting this chapter with a description of the approach taken by some of the other middleware vendors so that you can contrast the approach Microsoft has taken with that used by them. Microsofts approach is quite different, and we will be looking at the reasons why and the implications of this in this chapter.
Several middleware vendors currently support the InternetNCR with TOP END using a module called Java Remote Client services, Transarc with Encina using DE-Light Web Client, and BEA with Tuxedo using a module called Jolt. All of them use an identical approachan approach which uses Java, Web browsers, and Web serversand all of the above products are aimed at organizations wanting to support commercial transactions on the Internet using the Web.
Figure 12.1 DCOM and the Internet
The Internet and its protocols are normally unable to recognize the state of a transaction, which makes the multistep, often complex interactions which take place within a transaction largely impossible to support. DE-Light Web Client, JRC, and Jolt solve this problem by combining distributed transaction processing with software that can sustain transaction interactions over the Web.
Let me start by explaining the way the Web normally works. If you are a Web expert you can ignore the next few subsections, otherwise read on for an explanation of Web and Internet technology.
The Internet and the World Wide Web normally work using TCP/IP as the network protocol, Web browsers communicating with Web servers, URLs as the addressing mechanism, CGI to invoke programs, HTML to describe documents, and HTTP as the communication protocol. The Web is accessed using a Web browser on a client, which connects with a Web server. The Web browser (for example, Netscapes Navigator or Microsofts Internet Explorer) running on a PC or similar machine is used as the users point of access to the Internet. The Web server is located elsewhere on the network and handles the storage of information and retrieval on behalf of the client.
URLs
When a user wants to connect to a Web server, he uses a logical address name called the URLUniform Resource Locator. The URL is used for all Internet communication and has a format like that shown below:
aa://aa.aa.aa.aa/aa/aa/aa/aa.aa#aa
The first set of aas describe the data source/protocol. The Internet actually supports a number of protocols:
The Internet can thus act as an e-mail application, file transfer application, and data access application, among other things. The particular protocol used for Web access is the HTTP protocol.
The rest of the URL is then the addressin the case of Web access, the Web server name, the directory in which to find the Web page youre looking for, and even down to a particular spot in the particular page. The address, in other words, is not just the address of the Web server, it is the address of the particular bit of information youre wanting to look at on the Web server. This means that a company can set up a number of Web sites telling you about different thingssupport and help, general information, and so on.
HTTP
Communication from the PC to other nodes is via the protocol HTTP-HyperText Transfer Protocol. HTTP is just the particular protocol used for communication in the World Wide Web. It is not a network protocol, but an application protocolit defines how Web browsers and Web servers talk to one another. The Internet is, of course, based on the network protocol TCP/IP.
Information exchanges on the Web happen in four parts all classed as message types for HTTPconnection, request, response, and close. These message types bear a strong similarity to the RPC mechanisms we have discussed in this book.
Once HTTP has completely transferred the information requested, it has done its part. The browser then takes over to interpret and display the information.
Previous | Table of Contents | Next |