And THAT's a winner!

I couldn't tie any of the numbers I've seen together to make any sense.
I forgot that a couple months ago, we had to put some 30 users directly on 
TUBES (a session manager).  We did this by having them "dial tubes".

So when I was counting VTAM sessions, we seemed to fail on a weird, odd, number.
But when I add in a rough number of tubes sessions, I do, indeed, come very 
close to the LDEVRANGE parm in my config file.

That is much easier to fix, then trying to figure out how we were generating a 
few thousand extra LDEV sessions in order for us to hit the MAXLDEV parm.  

Time to revise my Reader Comment Form, I already sent in...

Tom Duerbusch
THD Consulting



>>> <[EMAIL PROTECTED]> 8/24/2007 12:46 PM >>>
My memory may be "fuzzy", but I seem to recall a TCPIP parameter that 
defined the LDEV range that TCPIP will use.
Although I've slept a few times (and drank a lot of beer) since then.
 




Tom Duerbusch <[EMAIL PROTECTED]> 
Sent by: The IBM z/VM Operating System <[email protected]>
08/24/2007 01:41 PM
Please respond to
The IBM z/VM Operating System <[email protected]>


To
[email protected] 
cc

Subject
Re: zVM 5.2 TCPIP problem






Hi Alan

I take it that you believe that we hit the default 4096 mark.
That's very interesting.....

Anyway, I did submit a reader comment form via email.

Tom Duerbusch
THD Consulting

>>> Alan Altmark <[EMAIL PROTECTED]> 8/24/2007 12:00 PM >>>
On Friday, 08/24/2007 at 12:03 EDT, Tom Duerbusch 
<[EMAIL PROTECTED]> wrote:
> Yesterday, we started getting the following error messages from the 
TCPIP stack:
> 
> 11:12:25 DTCSTM259E CONN 237: LDSFINITIATE RETURNED TRYING TO USE LDSF 
FOR TOO 
> MANY DEVICES.
> 11:12:25 DTCSTM163I TELNET SERVER: CONN 238:CONNECTION OPENED 08/23/07 
AT 
> 11:12:25
> 11:12:25 DTCPRC150I    FOREIGN INTERNET ADDRESS AND PORT:  NET ADDRESS = 


> 205.235.239.50, PORT= 1163
> 
> The first error message "DTCSTM258E", when looked up, doesn't give much 
of a 
> reason.

Tom, please submit a Reader's Comment Form (details in the book) so that 
we can improve the message to reference SET MAXLDEV.  (That command didn't 

exist when TCP/IP was born!)

Alan Altmark
z/VM Development
IBM Endicott

Reply via email to