SIGTRAN Plugtest Day 4
"Barry Nagelberg" <[email protected]>
| Newsgroups | gmane.ietf.sigtran |
|---|---|
| Message-ID | <[email protected]> |
Greetings from Moscow! Here is the wrap-up report for Day 4 of the SIGTRAN Plugtest. Protocols tested: M2PA - 4 implementations M3UA - 7 implementations M2UA - 5 implementations SUA - 5 implementations There was one issue raised at the Day 4 wrap-up. The issue is a case of non-interoperability that has been observed between 2 fielded implementations, and has already been discussed on the mailing list (see "Question related with DAUD Message" from Jan 31 - Feb 2 2007). The specific issue that was raised on the mailing list concerns one implementation sending 1024 instances of the same parameter in a DAUD msg, and another implementation not prepared to accept so many instances of the same parameter. The broader issue is that there is no specified limit on the size of an xUA message, or a limit on the number of times a single parameter may be repeated in a msg. In order to protect itself from running out of memory resources, each implementation is setting ad-hoc limits on what it is willing and able to receive. By their nature, these ad-hoc limits are all different, and we have wound up with non-interoperability issues. The lack of response from the xUA editors on this issue has been disapointing, and there is concern among the group that this issue isn't being addressed. Several options to address this issue were suggested during the wrap-up meeting: 1. Allow unlimited repetition of individual parameters, but set an upper limit on the total size of a non-data msg. The suggested limit is 8096 bytes. 2. Add functionality to the specs to allow the xUA endpoints to negotiate the max non-data msg size during startup. 3. Use the maximum transmission unit (MTU) size obtained via SCTP "Path MTU Discovery" to set the limit on the total size of a non-data msg. Barry Nagelberg