As an alternative to attaching OSA triplets to your 2nd level system, use NICDEFs in your 2nd level system and couple them to your normal vswitch. In either case, you still need to make sure you don't create duplicate IP addresses. That's where a coupling to a new disconnected vswitch will
guarantee isolation -- a good feature when you are learning, playing, and testing. Brian Nielsen On Mon, 11 Oct 2010 09:56:34 -0400, Mark Pace <pacemainl...@gmail.com> wrote: >Just attach a set of OSA triplets to your 2nd level guest for the VSWITC H. > >On Mon, Oct 11, 2010 at 9:41 AM, Dave Jones <d...@vsoft-software.com> wrote: > >> Yes, it can, George, no problem. >> >> DJ >> >> >> On 10/11/2010 8:36 AM, George Henke/NYLIC wrote: >> >>> Can TCPIP and VSWITCH, though isolated from Level 1, (as indicated by >>> Brian Nielsen below) still be functional at Level 2 to shake down TCP IP >>> maintenance before rolling it up to Level 1? >>> >>> Everything at Level 2 is a clone except VOLSERs. >>> >>> Brian Nielsen wrote: >>> >>> For TCP, what happens depends on what you have set up for your L2 guest. >>>> >>> >>> If you gave it access to a real hardware (OSA, hipersocket, etc) the n you >>>> >>> >>> have other work to do to prevent IP conflicts. If instead you've given i >>>> t >>>> access to a disconnected VSWITCH and/or virtual LAN, then it won't cause >>>> >>> >>> any problems because it can't connect to anything. And, yes, defini ng >>>> >>> >>> GRAFs and using DIAL is a standard practice. >>>> >>> >>> >>> >>> >>> >>> George Henke/NYLIC >>> 10/06/2010 10:46 AM >>> >>> To >>> The IBM z/VM Operating System<IBMVM@LISTSERV.UARK.EDU> >>> cc >>> >>> Subject >>> Re: IPL VM/VM Issues >>> >>> >>> >>> >>> >>> I IPLed L2 COLD, because I did not have the extra SPOOL volumes yet t o >>> come up FORCE. >>> >>> So I do not have the SDF. >>> >>> I suppose I could SPXTAPE DUMP them from L1, except I do not have any >>> tape. >>> >>> Can I redirect SPXTAPE to disk to port the SDF to L2?. >>> >>> >>> >>> >>> Brian Nielsen<bniel...@sco.idaho.gov> >>> Sent by: The IBM z/VM Operating System<IBMVM@LISTSERV.UARK.EDU> >>> 10/05/2010 04:29 PM >>> Please respond to >>> The IBM z/VM Operating System<IBMVM@LISTSERV.UARK.EDU> >>> >>> >>> To >>> IBMVM@LISTSERV.UARK.EDU >>> cc >>> >>> Subject >>> Re: IPL VM/VM Issues >>> >>> >>> >>> >>> >>> >>> As others have pointed out, your L2 guest can't hurt your L1 system o r >>> >>> other guests unless L2 has access to those disks. Stating it from a >>> different angle - you should never give your L2 guest access to anything >>> >>> you don't want it to have access to. It's no different than keeping your >>> >>> z/OS and Linux guests from stepping on each other or your L1 system. Thi >>> s >>> guest just happens to be running VM. >>> >>> For your IPL, if you've DDR'd your SPOOL volumes from L1 to L2, there is >>> >>> no reason to do a COLD start. Do a FORCE start instead. Otherwise y ou >>> >>> need to rebuild or SPXTAPE LOAD the SDF files. >>> >>> For TCP, what happens depends on what you have set up for your L2 guest. >>> >>> If you gave it access to a real hardware (OSA, hipersocket, etc) then you >>> >>> have other work to do to prevent IP conflicts. If instead you've given i >>> t >>> access to a disconnected VSWITCH and/or virtual LAN, then it won't cause >>> >>> any problems because it can't connect to anything. And, yes, definin g >>> >>> GRAFs and using DIAL is a standard practice. >>> >>> Something you might want to look into is setting up your system so it >>> recognizes whether it's at L1 or L2 or at DR so that it does the "rig ht >>> >>> thing" for that situation. There's several ways to do that, but you' ve >>> >>> gone down the path of diverging your L2 system from your L1 system. At m >>> y >>> site I have it setup so that I can DDR my L1 system into my L2 guest, int >>> o >>> a couple different setups at my DR site, or into some unknown site, and i >>> t >>> comes up the way I want it to in each instance even if the others are >>> running. It makes life easier. What you're doing is a great learnin g >>> >>> experience, and eventually you'll see the value in making your multip le >>> >>> systems easier to manage. >>> >>> Brian Nielsen >>> >>> >>> On Tue, 5 Oct 2010 15:28:01 -0400, George Henke/NYLIC >>> <george_he...@newyorklife.com> wrote: >>> >>> L2 is a cloned L1 directory except for the volsers in the Directory and >>>> >>> CF >>> >>>> Parm files which have all been made unique, ie 540 becoming 54X. >>>> >>>> I plan to ipl as follows: >>>> >>>> system reset >>>> term conmode 3270 >>>> set mach esa >>>> I 125b clear loadparm 009 >>>> >>>> START COLD DRAIN >>>> >>>> To be safe, I suppose I should also add NOAUTO. >>>> >>>> L1 runs 5 z/OS machines and 3 Linuxes. >>>> >>>> They could be corrupted at L1 if I tried to bring them up in L2 at t he >>>> same time without GRS, MIM, or some other serialization product. >>>> >>>> I doubt TCPIP will work at L2 without some reconfiguring. >>>> >>>> So I should define some GRAFs and dial them. >>>> >>>> Not sure if my L2 entry in the L1 Directory needs 54XRES, 54XPAG, 54XSPL >>>> >>> , >>> >>>> 54XW01, 54XW02 or whether I can just specify the IPL vplume, 54XRES, and >>>> >>> >>> CP finds the rest from the Parm disk. >>>> >>> >>> >>> >>> > > >-- >Mark D Pace >Senior Systems Engineer >Mainline Information Systems >
