According to the text in section 7.2.2.1 of H.323v4 (11/2000), as a response to a RRQ, a GK may respond with an RCF containing a timeToLive that is equal to or less than the timeToLive from the RRQ. This raises two questions: 1) Why isn't it allowed for a GK to respond with a greater value, or any value for that matter? 2) If the EP does not specify a timeToLive, should this be treated as a value of "0" and does this make it impossible for a GK to impose any timeToLive value? Frank ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ For help on this mail list, send "HELP ITU-SG16" in a message to listserv@mailbag.intel.com
Frank, For (1), It was felt that this was the appropriate action. I believe that Chris Purvis had some good comments on this topic some time back-- perhaps he can address this. As for (2), absence of the TTL means nothing more than that the endpoint does not support TTL. However, I have implemented a GK such that, if a value is not provided, I will provide one in the RCF. However, to ensure interoperability, I do not expect to get LW RRQs-- I use the old V1 procedure of sending IRQs :-) Nothing has changed in this area since V2. Essentially, the whole procedure is optional, though I will also say highly important and should be implemented by the endpoints. Paul ----- Original Message ----- From: "Frank Derks" <frank.derks@PHILIPS.COM> To: <ITU-SG16@mailbag.cps.intel.com> Sent: Monday, January 29, 2001 8:05 AM Subject: H.323's "keepAlive" mechanism
According to the text in section 7.2.2.1 of H.323v4 (11/2000), as a response to a RRQ, a GK may respond with an RCF containing a timeToLive that is equal to or less than the timeToLive from the RRQ. This raises two questions:
1) Why isn't it allowed for a GK to respond with a greater value, or any value for that matter? 2) If the EP does not specify a timeToLive, should this be treated as a value of "0" and does this make it impossible for a GK to impose any timeToLive value?
Frank
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ For help on this mail list, send "HELP ITU-SG16" in a message to listserv@mailbag.intel.com
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ For help on this mail list, send "HELP ITU-SG16" in a message to listserv@mailbag.intel.com
According to the text in section 7.2.2.1 of H.323v4 (11/2000), as a response to a RRQ, a GK may respond with an RCF containing a timeToLive that is equal to or less than the timeToLive from the RRQ. This raises two questions:
1) Why isn't it allowed for a GK to respond with a greater value, or any value for that matter?
There's no point. After all, the endpoint is permitted to renew its registration earlier than necessary, so if the gatekeeper put in a longer timeToLive an endpoint would not lose by (and probably would) use its preferred value anyway.
2) If the EP does not specify a timeToLive, should this be treated as a value of "0" and does this make it impossible for a GK to impose any timeToLive value?
I would suggest that this would imply that the endpoint does not support this feature. I would, on the other hand, have thought (unless one of the standards says otherwise) that a value of "0" would mean that the registration expires immediately, which may not be terribly useful! Not quite the same thing. Regards, Chris -- Dr Chris Purvis -- Development Manager ISDN Communications Ltd, The Stable Block, Ronans, Chavey Down Road Winkfield Row, Berkshire. RG42 6LY ENGLAND Phone: +44 1344 899 007 Fax: +44 1344 899 001 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ For help on this mail list, send "HELP ITU-SG16" in a message to listserv@mailbag.intel.com
I agree with Chris. Some other sig protocols use the value of 0 in registration msgs to ask for unregistration or indicate registration has expired. This is consistent with the definition of ttl and h323 implementors should agree on that. Even though h225 does provide urq msg to properly unregister a terminal, thus permitting to use the 0 value to indicate that the ras capability of the endpoint for keepalives, allowing to use the value of ttl==0 to indicate that keepalives are not supported is *not* recommended. Jean-Francois Mule. Chris Wayman Purvis wrote:
2) If the EP does not specify a timeToLive, should this be treated as a value of "0" and does this make it impossible for a GK to impose any timeToLive value?
I would suggest that this would imply that the endpoint does not support this feature. I would, on the other hand, have thought (unless one of the standards says otherwise) that a value of "0" would mean that the registration expires immediately, which may not be terribly useful! Not quite the same thing.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ For help on this mail list, send "HELP ITU-SG16" in a message to listserv@mailbag.intel.com
participants (4)
-
Chris Wayman Purvis -
Frank Derks -
Jean-Francois Mule -
Paul E. Jones