Issue 2939: IANA ports for IIOP need to be documented in Ch 13 and 15 (interop) Source: Micro Focus (Dr. Jishnu Mukerji, jishnu(at)microfocus.com) Nature: Uncategorized Issue Severity: Summary: Chapter 15 (formal/99-07-19), page 15-51 last para of section 15.7.2, and Chapter 13 (formal/99-07-17), page 13-36 last para, assert that "no 'well known' ports have been allocated", and then proceeds to state that individual implementations must use some unallocated port and document it. The statement about "no well known port has been allocated" needs to be revised to reflect that a well known port (683 for IIOP) has been allocated. Then we can change the statement about what individual implementations are expected to do by merely replacing the "must use some unallocated port" by a "may use the well known port or may use some unallocated port and document it" or some such. Resolution: see below Revised Text: The spec mentions port 683, but doesn’t say what it’s for (which is pointless). No-one is actually using 683 (or at least not for anything that’s well-defined). In addition, 683 is rather useless anyway because (at least under UNIX), you can use that port only if you are root. Change the first para on page 15-54 to read: Agents may freely choose TCP port numbers for communication; IIOP supports multiple agents per host. Actions taken: September 21, 1999: received issue May 13, 2002: closed issue Discussion: End of Annotations:===== Sender: jis@fpk.hp.com Message-ID: <37E7ABD3.2758546C@fpk.hp.com> Date: Tue, 21 Sep 1999 12:01:23 -0400 From: Jishnu Mukerji Organization: Hewlett-Packard EIAL X-Mailer: Mozilla 4.08 [en] (X11; U; HP-UX B.11.00 9000/889) MIME-Version: 1.0 To: issues@omg.org, interop@omg.org Subject: IANA ports for IIOP need to be documented in Ch 13 and 15 Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=us-ascii X-UIDL: a48b9932a0f1ff7400394707c8ad1a44 Chapter 15 (formal/99-07-19), page 15-51 last para of section 15.7.2, and Chapter 13 (formal/99-07-17), page 13-36 last para, assert that "no 'well known' ports have been allocated", and then proceeds to state that individual implementations must use some unallocated port and document it. The statement about "no well known port has been allocated" needs to be revised to reflect that a well known port (683 for IIOP) has been allocated. Then we can change the statement about what individual implementations are expected to do by merely replacing the "must use some unallocated port" by a "may use the well known port or may use some unallocated port and document it" or some such. Jishnu. -- Jishnu Mukerji Systems Architect Email: jis@fpk.hp.com Hewlett-Packard EIAL, Tel: +1 973 443 7528 300 Campus Drive, 2E-62, Fax: +1 973 443 7422 Florham Park, NJ 07932, USA. Date: Thu, 21 Oct 1999 18:34:48 -0400 From: Paul H Kyzivat Organization: NobleNet X-Mailer: Mozilla 4.0 [en] (WinNT; I) MIME-Version: 1.0 To: Juergen Boldt CC: interop@emerald.omg.org Subject: Re: issue2939 -- Interop RTF issue X-Priority: 3 (Normal) References: <4.1.19991021161936.00a407c0@emerald.omg.org> Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=us-ascii X-UIDL: 4oB!!"d!"!d > This is issue # 2939 [snip] > The statement about "no well known port has been allocated" needs to > be revised to reflect that a well known port (683 for IIOP) has been > allocated. Then we can change the statement about what individual > implementations are expected to do by merely replacing the "must use > some unallocated port" by a "may use the well known port or may use > some unallocated port and document it" or some such. If we are going to do this, then we need to be more specific about the intended use of the well known port. Merely putting some corba orb/server on that port doesn't guarantee anything to the client. It is my understanding that port 683 was assigned to the OMG specifically as a port on which a nameservice should be available. Namely, the url corbaloc::foo.bar:683/NameService should be valid if anything is. There is no particular reason to believe that this port is valid for anything else, though that is not excluded. Also, the possibility of multiple orbs coexisting on a node presents problems. It should be up to the administrator of a node to decide whether to deploy something on this port or not. Orbs should always be capable of using another port. It would however be reasonable for an orb implementation to deploy a name service on this port by default. Date: Thu, 21 Oct 1999 15:41:34 PDT Sender: Bill Janssen From: Bill Janssen To: Paul H Kyzivat Subject: Re: issue2939 -- Interop RTF issue CC: interop@emerald.omg.org In-Reply-To: <380F9508.6421A657@roguewave.com> References: <4.1.19991021161936.00a407c0@emerald.omg.org> <380F9508.6421A657@roguewave.com> Content-Type: text X-UIDL: @> Juergen Boldt wrote: > > > > This is issue # 2939 > [snip] > > The statement about "no well known port has been allocated" needs > to > > be revised to reflect that a well known port (683 for IIOP) has > been > > allocated. Then we can change the statement about what individual > > implementations are expected to do by merely replacing the "must > use > > some unallocated port" by a "may use the well known port or may > use > > some unallocated port and document it" or some such. > If we are going to do this, then we need to be more specific about the > intended use of the well known port. Merely putting some corba > orb/server on that port doesn't guarantee anything to the client. Exactly right. The Interop chapters should continue to say, ``No well-known port has been allocated for general CORBA IIOP usage.'' Any mention of a specific port allocated by the OMG for a specific service should be in the document describing that service. Two services that come to mind which might profit from standard ports are Naming and Interface Repository. Of the two, I'd suspect that Naming would profit more from a `protected' port (< 1024), whereas the IR would probably profit more from a non-protected port. Bill Date: Thu, 21 Oct 1999 18:34:48 -0400 From: Paul H Kyzivat Organization: NobleNet X-Mailer: Mozilla 4.0 [en] (WinNT; I) MIME-Version: 1.0 To: Juergen Boldt CC: interop@emerald.omg.org Subject: Re: issue2939 -- Interop RTF issue X-Priority: 3 (Normal) References: <4.1.19991021161936.00a407c0@emerald.omg.org> Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=us-ascii X-UIDL: 6e;e9nSYd9-'>e9[2,!! Juergen Boldt wrote: > > This is issue # 2939 [snip] > The statement about "no well known port has been allocated" needs to > be revised to reflect that a well known port (683 for IIOP) has been > allocated. Then we can change the statement about what individual > implementations are expected to do by merely replacing the "must use > some unallocated port" by a "may use the well known port or may use > some unallocated port and document it" or some such. If we are going to do this, then we need to be more specific about the intended use of the well known port. Merely putting some corba orb/server on that port doesn't guarantee anything to the client. It is my understanding that port 683 was assigned to the OMG specifically as a port on which a nameservice should be available. Namely, the url corbaloc::foo.bar:683/NameService should be valid if anything is. There is no particular reason to believe that this port is valid for anything else, though that is not excluded. Also, the possibility of multiple orbs coexisting on a node presents problems. It should be up to the administrator of a node to decide whether to deploy something on this port or not. Orbs should always be capable of using another port. It would however be reasonable for an orb implementation to deploy a name service on this port by default.