Re: LookupDiscovery MulticastTimeToLive
Dan Creswell <[email protected]>
| Newsgroups | gmane.comp.java.sun.jini |
|---|---|
| Message-ID | <[email protected]> |
Rikard Elofsson wrote: > Mark Brouwer skrev: >> Rikard, can you explain why you have to set the TTL to 0 for production? > > Ah, sorry no for production I want to set multicast TTL to 1 (or so). > Since I've been experiencing problems with a TTL of 0 I just want to be > sure that a TTL of 1 will actually work > From experience thus far, TTL of 1 will have the desired/expected effects although there _may_ be some bridges/routers/firewalls out there that don't do multicast right. I'd recommend you do a trivial multicast test to validate your production assumptions. There are various bits of test code lying around which can be used to ping around trivial multicast traffic and test routing behaviours. Let me know if you want help with this (I've done this test enough times to do it blindfold!) >> Although you probably can find reasons to want to have the dynamics of >> multicast discovery on the localhost, for me this is nearing the >> qualification of "uninteresting" (borrowing the words of somebody else >> :-). > > Hehe, the only reason for me to set ttl to 0 during development is to > use the same code in development and deployed while islolating my local > services from everybody else during development. I wasnt expecting to > run into trouble with it, but it seems like there is some confusion > around what ttl=0 actually should mean regarding leaving localhost or not. > Just to be clear - ttl=0 should mean packets don't leave the host but certain OS'en have network stacks that don't honour this properly. >> With regard to net.jini.discovery.ttl I couldn't find the system >> property in the docs or src of the 2.1 release so I doubt that one is >> still there, it seems you have to specify constraints. > > OK I'll go with constraints and then see what actually happens. > If nothing else, Constraints are probably the more future-proof/proper way to doo it. > //R > Dan. -------------------------------------------------------------------------- Getting Started: http://www.jini.org/wiki/Category:Getting_Started Community Web Site: http://jini.org jini-users Archive: http://archives.java.sun.com/archives/jini-users.html Unsubscribing: email "signoff JINI-USERS" to [email protected]