Re[2]: I'm working

mfm@arinc.com
Tue, 13 May 1997 11:46:43 -0400

Cameron,

Now that I've had a little more caffine.

#2 - Set SockOpt to Reuse address

Mike Marek

______________________________ Reply Separator _________________________________
Subject: Re: I'm working
Author: mfm@ccmail.arinc.com at SMTPGATE
Date: 5/13/97 7:46 AM

Cameron

I would have to think about the first problem a bit.

Question: Is the straps to Scotty connection TCP or UCP?

#2: This is most likely the lingering socket problem. I would agree
with your observation. Simple cure is Connect loop with pause.

Mike Marek
Sr. Engineering Manager
Arinc, Inc.


______________________________ Reply Separator _________________________________
Subject: I'm working
Author: Cameron Laird <claird@Starbase.NeoSoft.COM> at SMTPGATE
Date: 5/12/97 8:53 PM


on what we'll call a "mission-critical" application built
atop Scotty technology. I have a couple of problems with
the straps mechanism. I'm still not certain how to arti-
culate them, but I'm close enough now that I think it's
worth starting to ask others:
1. Suppose I have a scotty session, with a trap-re-
ception binding. straps is happily re-directing
TRAPs, my session is receiving and acting on them,
everything is fine. For some reason utterly ex-
ternal to my application--maybe someone simply
"kill $WRONG_PID"--straps goes away. My scotty
session no longer receives TRAPs. Worse, it
doesn't realize it's never going to.

Is there some way within Scotty that I can validate
the health of the straps-scotty connection?

I could do heartbeats from within the Scotty session,
but that feels very ugly.

My application in fact does not expose the Scotty
interpreter, but if I understand a solution for this
situation, I'm confident I can generalize it.
2. Imagine, again, that we have a host where straps
instances are coming and going. Sometimes when I
try to bind a trap-handler, Scotty reports "can not
connect straps socket: broken pipe". Immediately
afterward, binding is successful.

I have a customer who reports some of the same, with
a "broken plate" message (!?).

What's the story? Is this one of those LINGERing
sockets?

I confess, I haven't looked through much of the source code. I
know sockets well enough that I'm confident I could unravel
whatever is going on. I'm hoping to save myself some of the
work, and also learn what others are doing to cope with these
exceptions.

Cameron Laird http://starbase.neosoft.com/~bodi/nesi.html
Network Engineered Solutions +1 713 763 8366
claird@NeoSoft.com +1 281 996 8546 FAX
Houston WWW Business Guide: http://starbase.neosoft.com/~bodi/HouGuide.html

--
!! This message is brought to you via the `tkined & scotty' mailing list. 
!! Please do not reply to this message to unsubscribe. To subscribe or
!! unsubscribe, send a mail message to <tkined-request@ibr.cs.tu-bs.de>. 
!! See http://wwwsnmp.cs.utwente.nl/~schoenw/scotty/ for more information. 
--
!! This message is brought to you via the `tkined & scotty' mailing list. 
!! Please do not reply to this message to unsubscribe. To subscribe or
!! unsubscribe, send a mail message to <tkined-request@ibr.cs.tu-bs.de>. 
!! See http://wwwsnmp.cs.utwente.nl/~schoenw/scotty/ for more information.
--
!! This message is brought to you via the `tkined & scotty' mailing list.
!! Please do not reply to this message to unsubscribe. To subscribe or
!! unsubscribe, send a mail message to <tkined-request@ibr.cs.tu-bs.de>.
!! See http://wwwsnmp.cs.utwente.nl/~schoenw/scotty/ for more information.