Try #2, this time, in text mode.
I have a weird/flacky problem. I'm looking for thoughts on directions I can
take on this.
We have a MS/Access application that access DB2/VSE over a DRDA connection.
The VSE that hosts the DB2 engine is connected to the network via VSWITCH and
we go out to the "world" via an OSA card on the z/890.
There are two types of functions this application does, transactional and a
monthly "batch" update.
We upgraded from z/VM 5.1 to z/VM 5.2 Service Level 0602 on January 13, 2007
and apparently the "batch" update function stopped working at that time.
The user that does the "batch" processing was out on a medical leave so we
didn't know there was a problem until she got back in late February. And me,
being very busy, made a lot of changes to VSE in those couple months. So it
has been very difficult to backtrack the VSE side to find where things broke.
Well, the broken part is not in VSE. The broken production system STLESA2 also
has a test system STLESA2F which is a flash copy of the original system. Only
the IP address in the IPINIT00 CONFIG is different. That system was flashed on
February 23. And the user has always been able to run the "batch" updates on
this test system.
Well, over the weekend, I backed up to take the STLESA2F system, and then
flashed STLESA2 to STLESA2F and brought up STLESA2F. Other than the IP
address, they are both identical. She can't do the "batch" updates on STLESA2
but still can on STLESA2F.
So, the problem resides outside of VSE.
The users PC and software are obviously the same, so the problem must lie in
the network, or rather some configuration item in the network that is different
for STLESA2 vs STLESA2F. And that is what I need help on.
1. Is there any VM commands, other then Q VSWITCH that will show me some/all
configuration items that may be set for a particular connection?
q vswitch detail
VSWITCH SYSTEM VSW1 Type: VSWITCH Connected: 8 Maxconn: INFINITE
PERSISTENT RESTRICTED PRIROUTER Accounting: OFF
VLAN Unaware
State: Ready
IPTimeout: 5 QueueStorage: 8
Portname: UNASSIGNED RDEV: 2000 Controller: DTCVSW2 VDEV: 2000
Portname: UNASSIGNED RDEV: 2003 Controller: DTCVSW2 VDEV: 2003 BACKUP
VSWITCH Connection:
RX Packets: 2949792 Discarded: 120 Errors: 0
TX Packets: 4208954 Discarded: 0 Errors: 0
RX Bytes: 405358583 TX Bytes: 2002478169
Device: 2002 Unit: 002 Role: DATA
Adapter Owner: STLESA2 NIC: 0800 Name: THD390
RX Packets: 105 Discarded: 0 Errors: 0
TX Packets: 112 Discarded: 0 Errors: 0
RX Bytes: 7338 TX Bytes: 8080
Device: 0802 Unit: 002 Role: DATA
Options: IPv4 VLAN
Unicast IP Addresses:
192.168.99.5 MAC: 02-00-00-00-00-04
Adapter Owner: STLESA2F NIC: 0800 Name: THD390
RX Packets: 3329 Discarded: 0 Errors: 0
TX Packets: 9765 Discarded: 0 Errors: 0
RX Bytes: 786073 TX Bytes: 1160926
Device: 0802 Unit: 002 Role: DATA
Options: IPv4 VLAN
Unicast IP Addresses:
192.168.99.6 MAC: 02-00-00-00-00-07
>From SYSTEM CONFIG:
DEFINE VSWITCH VSW1 RDEV 2000 2003 PRIROUTER
MODIFY VSWITCH VSW1 GRANT TSTESA5
MODIFY VSWITCH VSW1 GRANT TSTESA5F
VMLAN MACPREFIX 030001
>From USER DIRECT:
USER TSTESA5F ~~~~~~
NICDEF 800 TYPE QDIO LAN SYSTEM VSW1
The controllers don't specify anything for a particular machine:
q controller
Controller DTCVSW2 Available: YES VDEV Range: * Level 520
Capability: IP ETHERNET VLAN_ARP
SYSTEM VSW1 Primary Controller: * VDEV: 2000
SYSTEM VSW1 Backup Controller: * VDEV: 2003
Controller DTCVSW1 Available: YES VDEV Range: * Level 520
Capability: IP ETHERNET VLAN_ARP
BTW, what we see on the console when the "batch" update fails is the IP
disconnect message:
But buffer length can be different values but seem to always be between 2000
and 3500.
Right now, I've pretty well cleared DB2/VSE and VSE/ESA 2.7 as being the
problem. The only other change in the "communication" path was the z/VM
update. That is my current target.
Thanks for any suggestions.
Tom Duerbusch
THD Consulting
