[Info-vax] TCPIP Services SMTP config problems

[Rich Jordan]

This seems to be the week for wierd problems.  I think we may have a
corrupt TCPIP$CONFIGURATION.DAT file

Standalone DS20e, OpenVMS Alpha V8.4, TCPIP V5.7 ECO 3, all class 1
and 2 ECOs up to April 2012.

The system had a bare SMTP config set up  for sending mail.  The
customer needed it changed to use a 'smarthost' for all email bound
offsite (a common request and often required by the ISP).

When we checked the system today it had 4 SMTP queues  (NODE is the
system nodename):

TCPIP$SMTP  (generic pointing to TCPIP$SMTP_NODE_1
TCPIP$SMTP_NODE_1
TCPIP$SMTP_NODE_00   (generic pointing to TCPIP$SMTP_NODE_01
TCPIP$SMTP_NODE_01

The first two queues are the active ones; all email process through
generic TCPIP$SMTP to TCPIP$SMTP_NODE_1; the other two queues, which
appear to have the correct and expected names, are not active.  The
GENERIC queue defined in SMTP config matches the second generic queue,
which is not used.

However when we ran @SYS$STARTUP:TCPIP$SMTP_SHUTDOWN, the TCPIP
$SMTP_NODE_1 queue does not stop or go into stopping state; it remains
idle.

TCPIP SMTP config is:

SMTP Configuration
 
Options
Initial interval:   0 00:30:00.00       Address_max:    16
NOEIGHT_BIT
Retry interval:     0 01:00:00.00       Hop_count_max:  16
NORELAY
Maximum interval:   3 00:00:00.00
TOP_HEADERS

Timeout             Initial       Mail    Receipt       Data
Terminate
  Send:                   5          5          5          3
10
  Receive:                5

Alternate gateway:  not defined
General gateway:    not defined

Substitute domain:  not defined
Zone:               not defined

Postmaster:         TCPIP$SMTP
Log file:           SYS$SPECIFIC:[TCPIP$SMTP]TCPIP$SMTP_LOGFILE.LOG

Generic queue       Queues   Participating nodes

TCPIP$SMTP_NODE_00      1     NODE


We made the following changes:

TCPIP> SET CONFIG SMTP /GATEWAY=ALTERNATE=newgateway.com
TCPIP> SET CONFIG SMTP /ZONE=HAL.APEXOIL.COM
TCPIP> SET CONFIG SMTP /OPTION=RELAY

SMTP.CONFIG and the local relay file were properly set up also.  These
changes ARE shown if we do another

$ TCPIP SHOW CONFIG SMTP

However on restart, the TCPIP$SMTP and TCPIP$SMTP_NODE_1 queues came
back into operation but completely ignored the changes.  Reviewing the
log shows the changes are not reflected in the queue startup output.
Reviewing headers of sent mail show they are still being sent direct;
the 'smarthost' is not in the route.

At the customer's request we did another TCPIP$SMTP_SHUTDOWN and had
to manually shut down the execution queue again (it remained idle).
We then deleted all four queues, defined TCPIP
$SMTP_NO_SUBS_DOMAIN_INBOUND logical, and restarted.  The two "wrong"
queue names were created again (TCPIP$SMTP and TCPIP$SMTP_NODE_1), the
queue logs still do not reflect the config changes, and the entry for
NO_SUBS_DOMAIN_INBOUND is shown as FALSE, so the logical did not get
implemented either

So given the queue name, several config params, and a logical are all
being ignored, while the management util (TCPIP SHOW CONFIG SMTP) does
display the correct info, is this likely to be a coruption in the TCPIP
$CONFIGURATION.DAT file, or could there be some other issue?

Thanks for any thoughts.