Tuxedo 7.1 - /WS Client Does Not Fail Over To Second Network Address Set In WSNADDR (Doc ID 773655.1)

Last updated on SEPTEMBER 09, 2024

Applies to:

Goal

DESCRIPTION:

Two /WS clients attempt to connect to an application.

Both clients have the same WSNADDR, which consists of a list of two network addresses.

For example:

  WSNADDR=//host1:port1,//host2:port2

The WSL specification in the ubbconfig looks like this:

  WSL     SRVGRP=WSLGRP SRVID=1000 RESTART=Y GRACE=0 MAXGEN=2 SEQUENCE=1
          CLOPT="-A -t -- -n //host1:port1 -m 1 -M 1 -x 1 -T 60"

  WSL     SRVGRP=WSLGRP SRVID=1100 RESTART=Y GRACE=0 MAXGEN=2 SEQUENCE=1
          CLOPT="-A -t -- -n //host2:port2 -m 1 -M 1 -x 1 -T 60"

The first /WS client successfully connects to Tuxedo on //host1:port1.

The second /WS appears to hang until the first client's request is complete instead of failing over immediately to the second WSL on //host2:port2.

The ULOG shows the following:

  124519.host1!?proc.xxx.1.-2: 08-13-2003: Tuxedo Version 7.1, 32-bit
  124519.host1?proc.xxx.1.-2: Calling tpinit(), PID=xxx
  124520.host1?proc.xxx.1.0: Successfully called tpinit(), PID=xxx
  124522.host1?proc.xxx.1.-2: 08-13-2003: Tuxedo Version 7.1, 32-bit
  124522.host1?proc.xxx.1.-2: Calling tpinit(), PID=xxx
  124624.host1!WSL.27373.1.0: WSNAT_CAT:1175: WARN: Disconnecting listener connection to workstation, request timed out
  124624.host1!?proc.xxx.1.0: LIBWSC_CAT:1037: ERROR: Network message receive failure
  124624.host1!?proc.xxx.1.0: LIBWSC_CAT:1059: ERROR: WSL returns error on connection request
  124625.host1!?proc.xxx.1.0: Successfully called tpinit(), PID=xxx

Why does the second /WS client not detect that the WSL on //host1:port1 is not available and immediately attempt to connect to the next network address listed in WSNADDR?

Solution

My Oracle Support provides customers with access to over a million knowledge articles and a vibrant support community of peers and Oracle experts.