E1 fails to start on NT App Server

[EHOss]

When we attempt to start the E1 service on an NT app Server the services 'starts,' but onle jdesnet.exe process actually runs. There are no logs created (not just nothing useful; I mean no logs). This is a multi-foundation configuration that has been functioning for several months. This weekend the one of the E1 instances was upgraded from 8.95_I2 to 8.98.1.4, and the upgraded instance is running fine. The enterprise server is an AS400 and is running fine in both instances. The event log on the NT machine states, "the description for Event ID ( 108 ) in Source ( PSFT E810 B9 Network ) cannot be found. The local computer may not have the necessary registry information or message DLL files to display messages from a remote computer. The following information is part of the event: The event log file is corrupt..." Essentially the same message shows up for both instances on the machine.

We have restarted both instances on the enterprise server and app server with no change. Have restored the non-working instance to prior to the aforementioned upgrade to the working instance with no affect. Changed profile E1 service starts under from Domain user to LocalSystem and back again with no affect.

I am open to suggestions on how to approach the problem.

 

[Sebastian Sajaroff]

Hi

Windows NT???
What are you running? A museum?

 

[Sebastian Sajaroff]

Hi

Being serious now...
I'm quite sure that Windows NT is no longer supported
under newer Tools like 8.98
You should definitely upgrade your NT boxes to W2003
or W2008.

 

[EHOss]

Sebastian, thanks for the reply - both of them. Actually we're not running NT, it's become for me like kleenex, although in this case it means an operating system from Microsoft rather than something to blow your nose in. I should have been more specific.

I waited all mornig for my question to get posted so I could send an update. I mentioned that there had been a tools release update on another instance running on the server, and that turned out (not surprisingly) to be the source of the problem. When performing the upgrade the guys who did the work found it necessary to add a system variable JDE_B9_ICU_DATA in order to get the upgraded instance to start (see Oracle Doc ID 883602.1). The value for the system variable was the path where the icudt32l.dat file. Once I learned about this little tidbit I deleted the system variable and the PRD instance was restored to a working state. Oddly enough the other instance started up successfully as well.

So, long story short it's working again.