5 replies to this topic

[Harold Donivan]

Quote

<< 2009-12-28 17:02:03 4403418/uxord /GAS39P41 /134452338 - %UNI_-E-U_EGAH39P3202, Procedure unknown in this environment
<< 2009-12-28 22:02:13 4403418/uxord /GAS39P41 /134452338 - %UNI_-E-U_EGAH39P3202, Procedure unknown in this environment
<< 2009-12-29 01:01:29 4403418/uxord /GAS39P41 /134452338 - %UNI_-E-U_EGAH39P3202, Procedure unknown in this environment
<< 2009-12-29 03:00:10 3723516/uxext /U_EXTR_C /134449010 - %UNI_-E-U_ENOEXT, No articles were extracted
<< 2009-12-29 03:00:10 3723516/uxext /UXEXT /134449010 - %UNI_-E-U_ENOEXT, No articles were extracted

[men]

Hi Harold,

What is your question exactly?

What issue are you facing?

And as usual what's the DU release and OS?

Michel

[Lindy_So]

I have this error message in the log file as well.

%UNI_-E-U_EGAH39P3202, Procedure unknown in this environment

Dollar Universe 5.12

MS Windows Server 2003 Standard

[men]

Hi,

512 is not maintained by ORSYP. You should upgrade ASAP to DU 513 or to DU 533 in both cases with the latest mandatory patches.

Michel

[Monika]

I'm afraid upgrading doesn't help.
We use DUS533 and have the same messages.
I can offer you some explanations:
There is a session with two or more nodes involved. You did not distribute all uprocs of the session to all concerned nodes but only those which should be executed on the specific node.
1. You want to execute a uproc on node A which does not exist (forgot to distribute). It aborts. Message is written in logfile of node B
2. A uproc executing on node A writes an event. You want to delete this event as part of a completion instruction of a uproc on node B and the uproc of node A is not present. The the message is written in the logfile of node B.
3. One of the nodes is offline, the session stops. The last node before switching to the offline node has the message written.

But I also see these messages on nodes which are stand-alone. They do not use external scripts or undefined variables. One of these jobs for example is just "exit 0".

Regards

Monika

[men]

Hello Monika,

Thank you for your feedback.

Some consultants recommend to distribute all Session objects to all Nodes where the Session is to run. We're also aware that some customers prefer not to do so. What could be done is at least distribute:

- the Sesion header to all Nodes (that was mandatory at a time)

- distribute to the involved Nodes the last Uproc to run on one Node and the first to run on the other node.

- distribute to the involved Nodes the conditioning Uprocs.

Hope I'm clear enough and that this reduces or stops the messages.

While reviewing PEs, we've selected several "Maintenance" PEs (they don't bring new functions, but enhance supportability, performance or an existing function) regarding messages so that we always have the object name inthe message.

Michel