DSXchange

"Connection reset by peer" indicates that something, somewhere in the network has dropped out. It is most usually a timeout - perhaps an inactivity timeout, as no traffic passes while rb_tmu is running on the remote machine. You may need to increase the inactivity timeout for the rsh connection.

You may need to increase the inactivity timeout for the rsh connection.

Ray:
How I do that??? Can you tell me???

Thanks in advance !!

Jman.

It's something that the UNIX administrator on the remote machine does. Outside the scope of my expertise to provide details, I'm afraid. Try man rsh (or man remsh as appropriate) - it may help.

But Ray, I launch de rsh command from the Windows 2000 Server.

I know that in the UNIX version of rsh you can configure timeout, but in the Windows version???


Thank you Ray

jman.

If the issue is time out, you can modify the rb_tmu code as follows:


rsh 10.7.145.2 -l redbrick rb_tmu -i nrows %2


where you set nrows to the number of rows after which the loader will write an information message back to the client.

Red Brick Manual says: "Directs the TMU to print a progress message
to the system message file for every n rows of data that
are loaded. Because the TMU processes rows in small
batches, it reports the number of rows processed at the
end of the batch in which an interval occurs. Hence the
number reported might not be exactly nrows. Frequent
intervals slow down the load process."


This should prevent your session from timing out, and give you abetter understanding of the progress of your load.

Thanks to all !!!!

jman