bradbury@oracle.UUCP (Robert Bradbury) (10/25/85)
In the September 23 issue of Electronics there is a discussion of why
Ultrix will not run on the uVaxII. Apparently there is a problem with
the memory chips that writing into memory chips which have not been
written into for 45 to 60 seconds causes the bits in the column adjacent
to the one being written to change. The article fingered the NEC memory
chips although DEC Corporate support would not confirm this.
DEC did indicate that the problem would be fixed with a component swap.
We have only been on the network for a short time, so pardon the questions
if they were hashed over in September.
1) Will anyone who knows (presumably at DEC) confirm or deny the article?
2) If the NEC chips are bad, whose chips are they replacing them with?
3) Why does the problem not bother VMS on the uVaxII? (Hypothesis: is
the VMS page handler so aggressive that pages get swapped out if they
are not accessed for 30 seconds?)
The article indicated that DEC had over 30 engineers working on the
problem at one point, if so why aren't they willing to tell customers
what the exact problem was? At least it would keep the negative comments
about Ultrix on the uVaxII from floating around on the net.
--
Robert Bradbury
Oracle Corporation
(206) 364-1442 {ihnp4!muuxl,hplabs}!oracle!bradburysnoopy@ecrcvax.UUCP (Sebastian Schmitz) (10/29/85)
Summary: Expires: References: <oracle.130> Sender: Reply-To: snoopy@ecrcvax.UUCP (Sebastian Schmitz) Followup-To: Distribution: Organization: European Computer-Industry Research Centre, Munchen, W. Germany Keywords: Re. 1. I have heard that memory was the problem. I do not know if the problem was exactly as you mentioned, but I do know it had something to do with memory loss. I do not work for DEC, nor am I their spokesperson and I know two people on the Net who are very happy about this (right A&J) :-). Re. 2. I don't know whose chips will be used to replace the duff ones. It may be good NEC ones, for all I care. Re. 3 The reason VMS was not bothered is simply that VMS will regularly access all the memory in a sort of "software refresh". I know nothing about VMS whatsoever so I don't know if this is true. This is one story I have heard. The other story I have heard is that Un(ltr)ix runs in low memory (which is where the duff chips were) and therefore got corrupted. VMS runs in high memory and therefore its kernel did not get mangled. The real answer is likely to be a combination of both. -- Love, Sebastian (Snoopy) "You haven't done it, till you've done it with pointers" \!mcvax\!unido\!ecrcvax\!snoopy /* N.B. valid csh address */