>The three SFS filepools that I share (2 on CP system, 1 on IFL system)
are
>all GLOBAL resources.
>I do find that once in a while IPGATE gets "confused" and
non-responsive and
>I have to recycle the SVM on both sides of the link.
>Tom Kern
I have the same problem with AVS. Once in awhile, a link just stops,
and has to be restarted. I was thinking of switching to IPGATE to avoid
that problem, so I'm disappointed to hear that IPGATE is no better. I'm
also concerned that there's no support for IPGATE, because it isn't an
official part of VM. Buying support from SNA or another third party is
probably not an option for us. Someone from IBM asked us a few years
ago if making IPGATE official would solve our problems with AVS, and we
said yes. Alan Altmark, if our name isn't on the requirement, would you
please add it?
In the meantime, the internal version of IPGATE would be an improvement
over what we have. I have an IPGATE VMARC file that I uploaded to VM on
5 Nov 2007, but I don't recall where it came from. It includes a
version of the fix that Kris Buelens posted to the list on 29 Jun 2006,
although it seems that some code was added along with the code that
moved:
>IPGATE did not pass the alternate userid of CMS clients to the SFS
server
>(in most cases that is).
>The fix is simple:
> - take the IPGATE1Y (S)MTREXX file,
> - move the 4 lines staring with "address 'CPICOMM' 'XCECSU ... "
> to before "if word(partner_LU_name,1) <> '*USERID'"
>Kris Buelens
Mine has:
address 'CPICOMM' 'XCECSU cm_convid security_user_ID ',
'security_user_ID_length rc'
if cm_return_code.rc = 'CM_OK' then do /* */
Secu_user = left(security_user_ID,security_user_ID_length)
If Secu_user <> Work_user then /* User has Alternate userid */
parse value Secu_User Work_user'('Secu_user')',
with Work_User Disp_user
end
if word(partner_LU_name,1) <> '*USERID' then do
/* remote access -----------------------------************************/
Dave Jones offered a version of IPGATE to the list on 18 Jun 2007, that
included encryption:
>I've added an encryption capability to IPGATE so that the data flowing
>across the TCP/IP network is shielded from nosy folks....drop me a note
>off-list if you're interested. And many thanks to Perry and Holger for
>helping me understanding the code well enough to add this feature in.
I did send Dave a note, but I don't think what I have was from him.
Searches for the strings "Jones" and "crypt" do not return any hits.
Most of the files are dated 19 Mar 2006. The Mar 2006 files are either
identical to the files in SG245164 VMARC (e.g. RES* EXEC, or aren't in
SG245164 VMARC (e.g. SAMPMAIN MTREXX, TIME MTREXX, VMTIMED MTREXX).
IPGATE1Y (S)MTREXX includes Kris' fix, and is dated 29 Jun 2006.
IPGATE1 (S)MTREXX is dated 3 Jul 2006. The only note file in the
package is MTREXX NOTES, which contains Perry Ruiter's original notes on
MTREXX.
The difference in IPGATE1 is one line:
NEW 62 parse value socket('Bind',socketid,'AF_INET' listen_port
'INADDR_ANY') with rc .
OLD 62 parse value socket('Bind',socketid,'AF_INET' listen_port
tcp_host_id) with rc .
Dennis
"A pistol! Are you expecting trouble Sir?" "No Miss, were I expecting
trouble I'd have a rifle."
