RE: [ws-rx] [NEW ISSUE] Rollover underspecified,state of sequence is unclear

I think we did, but I can’t find it
in WD 15.

I think that we decided that it might
close the sequence but that is not in the text unless you know where it is.

Thanks

-bob

 

From: Doug Davis
[mailto:] 

Sent: Monday, June 26, 2006 12:25
PM

To: 

Subject: Re: [ws-rx] [NEW ISSUE]
Rollover underspecified, state of sequence is unclear

 

Bob, 

 
if a MsgNumRollover fault is generated I don't think the RMD should necessarily
terminate the sequence.There may be other messages, with lower numbers, that
could still be received and processed successfully.  Didn't we discuss
this already - it sounds familiar. 

-Doug

 

  
  
"Bob Freund-Hitachi" <> 

  
06/26/2006 11:33 AM 

  
  
  
   

    
    
To

    
    
    
"[WS-RX]"
    <> 

    
   

   

    
    
cc

    
    
    
 

    
   

   

    
    
Subject

    
    
    
[ws-rx] [NEW ISSUE] Rollover underspecified,
    state of sequence is unclear

    
   

  
  
 

  
   

    
    
 

    
    
    
 

    
   

  
  

  
 

Description 

 

WD
15 lines 564-570 state that if the internal limitation for MessageNumber is
exceeded by either RMS or RMS, then a MessageNumberRollover fault is generated.

Section
4.5 which describes the fault provides no further information 

 

 

Justification 

 

Possibility
for protocol lockup or non-interoperable behavior 

 

 

Target:
core 

 

Proposal: 

 

Insert
after line 570: 

Generation
or receipt of a MessageNumberRollover fault MUST immediately terminate the
sequence 

 

Insert
in section 4.5 

 

Sent
by: RMS or RMD 

Condition:
message numbers for a Sequence have been exhausted 

Action
upon Generation or receipt: The Sequence is terminated immediately.