Going back ... I didn't follow this:
On 05/17/2018 08:03 AM, Paul Flint wrote, citing me: >> *IUCV* is best, if you have an IUCV terminal server running. You can SSH >> into the terminal server, hop to the ailing Linux guests via IUCV, and >> "everything is normal". (Well ... except that you're in single-user >> maint mode.) We really should make more use of IUCV in all ways, >> terminal server, things layered on that, and non-terminal traffic. It's >> awesome! > > While I have in my time incanted with IUCV I must petition the Troth > that this potent VM Magic (VMM) may not be the best for an initiate. > A wizard of your standing can hold power over the wild IUCV, transfer > files, edit minidisk content, run DDT (whose name will be spoken > here), or invoke new CMS, but I plea, this is a dangerous path! On both Linux (all platforms) and CMS, IUCV is available as an address family (not functional apart from z/VM) in the sockets interface. AF_IUCV is right there alongside AF_INET, AF_INET6, AF_UNIX (AF_LOCAL), and the rest. What I don't follow is why you say "this is a dangerous path". (Or maybe I misunderstood.) Dangerous? It's actually *less* dangerous than the networked address families because it's guaranteed to be local, as is AF_UNIX (AF_LOCAL). Have no fear! It's easy to use and as safe assocket() can be. The "IUCV console" support in the kernel is based on the same thing. At one time I wrote some code to connect to a Linux console from CMS and vice versa. Fun! -- R; <>< ---------------------------------------------------------------------- For LINUX-390 subscribe / signoff / archive access instructions, send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390 ---------------------------------------------------------------------- For more information on Linux on System z, visit http://wiki.linuxvm.org/
