Hi Martin,
I didn't look up my notes now but trying to remember what we discussed:
Both agents are allowed to end a transaction at any time.
Messages can be sent at any time from any peer. There is no way to control
that client and server
do not send their TE messages simulatinously.
When an agent sends a TE message it will free the data it has associated
with this transaction.
Still in many cases it will receive messages for this transaction from the
other peer.
But those messages do then belong to an unknown transaction and therefore
must be ignored.
So, whether the other peer sends a TE or not is irrelevant.
I think, it would be at least good practice if an agent sends a TE when
it is done with the transaction, no matter how this decision has been done,
even it is convinced
that the peer has already sent his TE. Sending its own TE does not harm but
if no side is sending
a TE we have a problem. That's why all examples I did have a TE from both
ends.
Regards
Martin Stecher
-----Original Message-----
From: owner-ietf-openproxy(_at_)mail(_dot_)imc(_dot_)org
[mailto:owner-ietf-openproxy(_at_)mail(_dot_)imc(_dot_)org] On Behalf Of
Martin Thelian
Sent: Dienstag, 6. Februar 2007 16:17
To: Openproxy
Subject: OCP - Transaction End (TE)
Hi!
I have a question regarding the Transaction-End (TE) message
of the OPES callout-protocol.
It is unclear to me whether both OCP-agents have to send a TE
message, or if the transaction is terminated with the first
TE-message send or received. Is this an XOR or an OR disjunction?
I've found the following statements in rfc-4037 (OCP-core):
2.5. Termination
[...]
Usually, the callout server terminates adapted
application message
delivery and the transaction.
[...]
4. Transactions
[...]
The OPES processor or the callout server can terminate the
transaction by sending a corresponding message to the other side.
[...]
A transaction ends with the first Transaction End (TE)
message sent
or received, explicit or implied. A TE message can be sent by
either side.
[...]
The figure below illustrates a possible message sequence [...]
P: TS 10;
P: AMS 10 1;
... processor sending application data to the callout server
S: AMS 10 2;
... callout server sending application data to the processor
... processor sending application data to the callout server
P: AME 10 1 result;
S: AME 10 2 result;
P: TE 10 result;
[...]
11.5. Transaction Start (TS)
[...]
A callout server MUST maintain the state until it
receives a message
indicating the end of the transaction or until it terminates the
transaction itself.
[...]
11.6. Transaction End (TE)
[...]
An OCP agent MUST send a Transaction End (TE) message immediately
after it makes a decision to send no more messages related to the
corresponding transaction.
So, most parts of the RFC - especially the example shown in
chapter 4, indicate that this is an XOR, but chapter 11.6
does not clarify this.
But in the message flow examples shown in RFC4236 (OCP HTTP
profile), both agents are sending TE messages to the other endpoint.
Can someone explain me what's the correct usage of the TE
message? Thanks.
Regards,
Martin Thelian