Bugs item #964508, was opened at 2004-06-01 15:53 Message generated for change (Comment added) made by davetron5000 You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=376685&aid=964508&group_id=22866
Category: JBossServer Group: v3.2 Status: Closed Resolution: Invalid Priority: 5 Submitted By: Dave Copeland (davetron5000) Assigned to: Scott M Stark (starksm) Summary: RMI Codebase URL value is determined incorrectly Initial Comment: In server/src/main/org/jboss/web/WebService.java on line 293, the following call is made: address = ServerConfigUtil.fixRemoteAddress(address); This converts the configured jboss.bind.address (or whatever was used in jboss-service.xml) and turns it into a name-based InetAddress. This seems wrong, since if you specified a jboss.bind.address to be an ip address, you want the RMI Codebase URL to use the ip address. This URL gets sent to clients and the clients may not have DNS (or worse, may not have correct DNS). Yes, it is overridable by a system property, but I can think of no reason to default to sending the hostname. It may make sense if the jboss.bind.address is the default of 0.0.0.0, but if it is specified, that exact value should be used. ---------------------------------------------------------------------- >Comment By: Dave Copeland (davetron5000) Date: 2004-06-15 15:06 Message: Logged In: YES user_id=751883 I am not sure what you mean by "address". The way the code is, fixRemoteAddress is called if the java.rmi.server.codebase property is null. That seems OK to me. The next step is to create the codebase to use. It does this by taking the value of the variable address (which is set to the value of the bind address), looking it up in DNS to get a name and using that as the codebase. Since 0.0.0.0 is a special case, I can see the need to get *some* sort of host name or ip address to send over as the rmi server code base. What I have a problem wiht is that when I explicitly set the jboss.bind.address in my config file, I think that the name lookup should be skipped. For one thing, it isn't needed. For another, it runs the risk of producing an incorrect result (as it does for me). The "fix" was for me to explictly set java.rmi.server.codebase, and now I have to keep both that and the jboss.bind.address in sync, when JBoss could easily figure out the codebase from the bind.address for me. As for you -b, I'm not sure what that is, but I don't see how the existence of some command line option has any bearing on the behavior of the code in question. ---------------------------------------------------------------------- Comment By: Scott M Stark (starksm) Date: 2004-06-15 14:11 Message: Logged In: YES user_id=175228 This analysis is incorrect. The only time the hostname is returned from ServerConfigUtil.fixRemoteAddress(String address) is when address is null or equal to '0.0.0.0'. If I run with -b localhost, l see a codebase of http://localhost:8083/. If I run with -b 172.17.66.52, I see a codebase of http://172.17.66.52:8083/ ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=376685&aid=964508&group_id=22866 ------------------------------------------------------- This SF.Net email is sponsored by The 2004 JavaOne(SM) Conference Learn from the experts at JavaOne(SM), Sun's Worldwide Java Developer Conference, June 28 - July 1 at the Moscone Center in San Francisco, CA REGISTER AND SAVE! http://java.sun.com/javaone/sf Priority Code NWMGYKND _______________________________________________ JBoss-Development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
