Path: utzoo!attcan!uunet!lll-winken!lll-tis!helios.ee.lbl.gov!pasteur!ucbvax!LBL.GOV!nagy%warner.hepnet From: nagy%warner.hepnet@LBL.GOV (Frank J. Nagy, VAX Wizard & Guru) Newsgroups: comp.os.vms Subject: Re: One piece to the LAVC/750 puzzle... Message-ID: <880609062727.256017ce@LBL.Gov> Date: 9 Jun 88 13:27:27 GMT Sender: daemon@ucbvax.BERKELEY.EDU Organization: The Internet Lines: 48 >> Some friends would like to set up a LAVC which includes 750s as satellite >> nodes that boot form a system disk on the LAVC boot node. Presumably, one >> would need either a boot PROM or a suitably configured TU58 with an >> appropriate boot program (such is used to boot 750s in a CI cluster). > ... I would think that it would be possible > to make a 750 look like a MicroVAX II or PDP/Gateway without too much sweat. > It may again be a very nontrivial task. ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ is > ... A cursory comparison of the exact > differences between 750 and MicroVAX II architecture should be an indicator. > [Yes, there are some *BIG* differences, but will it matter?] The reason the MicroVAX-II (or -3000) can be booted across the network is that the boot PROMs support the DEQNA as a boot device for the 'secondary' boot code. The primary boot code (known as VMB.EXE on *big* VAXes) lives in those ROMs on the MicroVAXen. There has to be explicit support in the boot ROMs for network boots. The 750 ROMs do not have the necessary code to boot from a DEUNA or DELUA. From some old conversation I had, I understand that the 750 boot ROMs are too small for such code to be added. It *might* (repeat: MIGHT) be possible to write a version of VMB.EXE to boot a 750 across the network. The 750 would then have to be setup to boot VMB from the TU58 (real *s*l*o*w*l*y*) AND again you'll probably have space problems since a TU58 does not have a great deal of space. In either case, this is *not* a trivial task. And without ROM support for network boots, 750's can't be LAVC satellites. However, you should be able to make multiple 750's be LAVC members by having each LAVC be a boot node with its own local system disk. You'll have to play with the cluster parameters (VOTES and QUORUM) to get it to work correctly and avoid a partitioned cluster, will have to understand what you are doing since what you do will be *UnSupported* by Digital (that means they'll not be too sympathetic or helpful if you have problems - it does not necessarily mean that it *won't* work, it just takes more expertise and the will to succeed - or the guts to bull ahead through the obstacles). = Frank J. Nagy "VAX Guru & Wizard" = Fermilab Research Division EED/Controls = HEPNET: WARNER::NAGY (43198::NAGY) or FNAL::NAGY (43009::NAGY) = BitNet: NAGY@FNAL = USnail: Fermilab POB 500 MS/220 Batavia, IL 60510