Path: utzoo!attcan!utgpu!jarvis.csri.toronto.edu!mailrus!ames!ucsd!orion.cf.uci.edu!mrichey
From: mrichey@orion.cf.uci.edu (Mike Richey)
Newsgroups: comp.dcom.lans
Subject: Re: ELS II/non-ded ctrl-alt-del at server/workstation
Message-ID: <2817@orion.cf.uci.edu>
Date: 20 Sep 89 15:33:17 GMT
References: <2762@orion.cf.uci.edu> <8195@megatest.UUCP>
Reply-To: mrichey@orion.cf.uci.edu (Mike Richey)
Distribution: usa
Organization: University of California, Irvine
Lines: 50

In article <8195@megatest.UUCP> palowoda@megatest.UUCP (Bob Palowoda) writes:
>From article <2762@orion.cf.uci.edu>, by mrichey@orion.cf.uci.edu (Mike Richey):
>> In article <8111@megatest.UUCP> palowoda@megatest.UUCP (Bob Palowoda) writes:
>> 
>>>  Why does Novell sell a non-dedicated mode and everyone complains it
>>>  works so poorly and pawns it off on bad marketing hype. I suspect 

>
>> is my experience. I managed four Novell file servers at my previous place
 of 
>>employment. I had 100% uptime on three servers for fourteen months til I left.
>> (read, NO DOWN time, except for standard maintenace)
>
>  I calculate my maintenace as downtime. I had a group of software
>engineers developing software and off coarse their would bound to

Okay, so we disagree on what maintenance is (vs down time). Let me say that the
servers did not fail. They had battery backup to be sure.
 
>> The fourth was a different story. This wasn't the fault of Netware, but
>>hard ware faults. I'm not advocating that Netware is flawless. Like I said,
>>this was my experience.

> Yeah I seen this, intermittent bad spots on disks. I still think Novel's
> surface scan is not as robust as unix's. 

Hotfix is now used in the advanced version of Netware. This perform a read
after write verify of the block written. If the verify is not successful,
the block is relocated. The need to run the surface scan in compsurf has
pretty much been elliminated. Now, without justifying all of this I'm sure
this is open to attack, but I'm not trying to convert anyone to Netware,
so I'll leave out the specifics as to why. It used to be that a 160 meg
drive would take around 48 hours to certify the disk for the recommended
3 passes. 

I didn't mention bad spots on the hard disk at all. This was your assumption.
Since you brought it up. I had a power supply fail (not the fault of the O/S)
A memory failure, which produced NMI interrupts, I will admit that Netware
does not take to kindly to memory failures where the kernel exists, but what
O/Ss for 80X86 products do. (Does Unix? I really don't know)
I had an Ethernet card fail in the server, and a disply adapter. None of these
were the cause of the O/S. But each time this server failed, I was sweating.

So what's the point, you don't like Netware? Okay. so be it.


Michael S. Richey    Internet: mrichey@orion.cf.uci.edu
Bitnet: MRichey@UCI  CompuServe: 71650,3132    Voice: (714) 856-8374
University of California, Irvine    Network and Telecommunication Services
342 Computer Science      Irvine,   CA  92717