*I* certainly  didn't post this  message to  LSTSRV-L, and can't  see the
relationship between CTC  problems and LINKSWT FILE, which  is a LISTSERV
data file defining those network lines that are particularly overloaded.
 
However you should note that:
 
- VM spool is  not driven by a  single virtual machine, but  by CP. Spool
  I/O are, however, synchronous, and  RSCS will therefore be 'held' until
  the spool I/O is  complete. I've always said that RSCS  ought to do its
  own private spooling.
 
- Throughput  is  much  better  under   MVS  mainly  because  of  the  XA
  architecture, which (for example) allows a  CCW to be processed by some
  subchannel and  the next one  by another  subchannel, if the  first one
  happens to be busy. This is  another explanation of why I/O can degrade
  considerably under S/370 when the I/O subsystem is heavily loaded.
 
- The RSCSV1 DMTVMC  line driver is MUCH more  efficient than DMTNJI/NJE.
  It  takes a  4k spool  buffer, sends  it 'as  is', etc.  It works  only
  between VM  system of course,  so it  is irrelevant in  this particular
  case. By doing some experimentation, I've found out that the speed of a
  CTC transfer  is limited  solely by the  DASD speed/load  using DMTVMC,
  whereas CPU contention can be observed with DMTNJI/NJE if the system is
  heavily loaded.
 
Eric