NonStop Server Guide for BEA WebLogic Server 8.1

Using the WS Plug-in
HP NonStop Server Guide for BEA WebLogic Server 8.1—529891-001
8-27
Troubleshooting
Cause. The WS Plug-in was unable to communicate with the TCP/IP process while
trying to contact a target application server process.
Effect. The WS Plug-in stops trying to send the request to the currently selected
application server process and looks for an alternate target application server process.
The RLS periodically retries.
Recovery. Investigate why the TCP/IP process is not responding and take corrective
action.
Troubleshooting
•
The TS/MP numstatic setting for the rmt.pway Plugin should be at least 2 to 1
with regard to the TS/MP numstatic setting of the httpd process. For example,
if numstatic is set to 5 for the httpd serverclass, the rmt.pway serverclass
should have a numstatic setting of at least 10. (The TS/MP maxservers setting
for the rmt.pway serverclass would also have to be set to at least 10). If the
rmt.pway setting is not correct, the rmt.pway serverclass gets TS/MP error
1034.
•
If access.log shows that it is using HTTP/1.0 even if the rmt.pway serverclass
has been configured with HTTP11_SUPPORT ON, could signify that the current
configuration settings for the httpd rmt.pway serverlass coupled with the BEA
WebLogic Server, are not well tuned or adequate for the current environment. This
can be accompanied by error messages in the error log that state, “Zero length
reply received from server, Aborting”. In general, this means that configured
resources are overloaded.
Check /usr/tandem/webserver/logs/httpd.log for iTP Secure WebServer
messages.
For more troubleshooting information, see the HP iTP Secure WebServer
documentation.