[comp.sys.mac.misc] Mac Plus server viability

phil@waikato.ac.nz (10/26/90)

In article <1990Oct24.161907.22685@panix.uucp>, alexis@panix.uucp (Alexis
Rosen) writes:
> 
> BTW, using a MacPlus as a server is a singularly poor choice. If you have the
> option, even an SE (or Classic) will perform much better.
> 
Could you explain why?  The people I know who run them as such have had no
problems whatsoever (one serves a lab of 25 SEs and has about 3 SCSI devices
attached to it).
-- 
Phil Etheridge (phil@waikato.ac.nz)                 /\  /\   -+-,--, .--, ._
Computer Services/Mathematics & Statistics         /  \/  \  / /--< /--  /
University of Waikato, Hamilton, NZ.              /        \/ /__.) \_  /

bmug@garnet.berkeley.edu (BMUG) (10/26/90)

In article <2077.272816ec@waikato.ac.nz> phil@waikato.ac.nz writes:
>In article <1990Oct24.161907.22685@panix.uucp>, alexis@panix.uucp (Alexis
>Rosen) writes:
>> 
>> BTW, using a MacPlus as a server is a singularly poor choice. If you have the
>> option, even an SE (or Classic) will perform much better.
>> 
>Could you explain why?

Well, I can think of at least two reasons why an SE is better than a Plus:

1. If you ever decide to Ethernet your network, it's much easier doing it
with an SE server (because of the internal slot).

2. The SE has a fan and more robust power supply.  I like my servers cool
and reliable.

Point (2) goes for the Classic, as well.

John Heckendorn
                                                             /\
BMUG                      ARPA: bmug@garnet.berkeley.EDU    A__A
1442A Walnut St., #62     BITNET: bmug@ucbgarne             |()|
Berkeley, CA  94709       Phone: (415) 549-2684             |  |

alexis@panix.uucp (Alexis Rosen) (10/27/90)

In article <2077.272816ec@waikato.ac.nz> phil@waikato.ac.nz writes:
>In article <1990Oct24.161907.22685@panix.uucp>, alexis@panix.uucp (Alexis
>Rosen) writes:
>> BTW, using a MacPlus as a server is a singularly poor choice. If you have the
>> option, even an SE (or Classic) will perform much better.
>Could you explain why?  The people I know who run them as such have had no
>problems whatsoever (one serves a lab of 25 SEs and has about 3 SCSI devices
>attached to it).

Sure. The fact is, SCSI on a Mac Plus is _much_ slower than on an SE or
Classic. Since LocalTalk is an absolute bottleneck (it grabs the entire
processor whenever it's actually sending or receiving a packet), anything
else that takes time is also a bottleneck. Thus, the faster your disks,
the faster your server. (Of course the extra 20% speed on an SE helps a little
too. Processor speed is very important on an AppleShare server.)

In fact, both the Mac Plus and the SE are abysmally slow servers. The fact
that the SE's SCSI is about twice as fast as the plus's is doesn't mean that
the SE will serve twice as fast (net bandwidth, which is more important, is
a constant). But every little bit helps. On the other hand, users on a network
can't tell the difference between a Mac Plus and an SE as workstations (all
other things being equal)- I've checked. So it may well make sense for you
to take one of your SEs and swap it with the Plus.

To descend to the level of rumor now, I've heard that there is some problem
with the plus that makes it a bit more unreliable with LocalTalk than the
SE, even if they're both using the AppleTalk v52 (or v53) file in the System
Folder. I don't know if this is true, but there's some small evidence to
support it. (Some third-party network products don't work on a Plus, even
with AT v.52.)

Followups to comp.sys.mac.system.

---
Alexis Rosen
Owner/Sysadmin, PANIX Public Access Unix
{cmcl2,apple}!panix!alexis