RE: DRN Label Issues -- Issue #1
"Haresign Lincoln" <[email protected]>
| Newsgroups | gmane.ietf.sigtran |
|---|---|
| Message-ID | <849535E338E99741B7F7413F73253EDB04E8BC4F@us-nj-mail1.comverse.com> |
Brian, I would assume that anyone who is implementing the DRN routing mechanism will use a mechanism that assures a one to one routing between DRN and DLR. Otherwise the mechanism would be useless. This seemed obvious to me. But if you feel that we need to clarify this in the IG, I'd be happy to specify this. Regards, Lincoln -----Original Message----- From: Brian F. G. Bidulock [mailto:[email protected]] Sent: Wednesday, October 11, 2006 3:27 PM To: Haresign Lincoln Cc: Barry Nagelberg; [email protected] Subject: DRN Label Issues -- Issue #1 Lincoln, Here is the DLR/DRN problem restated: RFC 3868 says: 4.7.3.2. SCCP Connection Oriented traffic Messages not containing a destination reference number (DRN), i.e., a Connection Request, MAY be loadshared among the available ASPs. The load distribution mechanism is an implementation issue. When a DRN is present, the SG extracts the label and selects the ASP that corresponds with it. If an ASP is not available, the SG discards the message. But messages arriving from the SS7 network do no have DRNs. They have DLRs. There is no 1:1 mapping requirement between a DLR and a DRN. Therefore, because no CO SCCP SS7 message has a DRN in it, I assume that all messages can be loadshared and that DRN Label is completely ineffective. --brian -- Brian F. G. Bidulock [email protected] http://www.openss7.org/