No. We know what the error was so we had no need to look at EREP.

I'm not sure if VM even will get an error to be processed. I guess a 
programm (in this case DDR) will initiate an IO, the subsystem will process 
the IO. If the subsystem decides to process that IO on channel 13 it will 
do just that. And since the hardware is the same, both 3390-3 DASD, there 
is no need for the hardware or the subsystem to issue an error. It just 
does what it is told to do. Or am I missing something?

Based on POP, if I understand correctly, for every IO-start the path is 
selected again. Would that be a logical selection? I.e. first path 10, next 
path 11 etc. In that case the error would only be noticed if channel 10-12 
are occupied and 13 is free to start an IO. That would also be the case 
when I query the label. There is just a little chance to get the IO on 
channel 13. But with the 1000's of IO's when running DDR the chance that at 
least 1 IO is running on channel 13 increases.

That still leaves DDR. When I copy data from tape to DASD, what IO's are 
started? Is it 1 start for a query of the label and 1 start for the actual 
write of data? Or does DDR issue a start for every block of data?

Regards, Berry.

Reply via email to