Sun-Spots-Request@Rice.edu (William LeFebvre) (09/02/88)
SUN-SPOTS DIGEST Thursday, 1 September 1988 Volume 6 : Issue 213
Today's Topics:
Re: IconEditor problem / At problem
Re: Gnu Emacs under 4.0
package for printing on VAX printers from Suns
Those WREN-IV's again
Speeding up dumps over the network
SunOS 4.0 version of nnstat
SunOS 4.0 -- initgroups error
vi core dumps (illegal instruction)
Having trouble running standalone copy over the net
Yellow Pages problem
1/4" cartridge tape drives for Sun SCSI?
/usr/lib/sendmail vs /usr/lib/sendmail.mx?
Sun delivery times: HOW LONG?
Send contributions to: sun-spots@rice.edu
Send subscription add/delete requests to: sun-spots-request@rice.edu
Bitnet readers can subscribe directly with the CMS command:
TELL LISTSERV AT RICE SUBSCRIBE SUNSPOTS My Full Name
Recent backissues are available via anonymous FTP from "titan.rice.edu".
For volume X, issue Y, "get sun-spots/vXnY". They are also accessible
through the archive server: mail the request "send sun-spots vXnY" to
"archive-server@rice.edu" or mail the word "help" to the same address
for more information.
----------------------------------------------------------------------
Date: Wed, 31 Aug 88 16:51:05 EDT
From: Sharon O. Beskenis <sdo@phoebe.larc.nasa.gov>
Subject: Re: IconEditor problem / At problem
I reported the 'at' problem to Sun a few months ago. No, you are not
doing anything wrong. The solution to the problem is to replace
/usr/lib/atrun and /usr/bin/at with the 3.4 versions and everything will
run fine.
Sharon Beskenis
PRC Kentron
NASA/LaRC
MS 478
Hampton, VA 23665
(804)865-3535
------------------------------
Date: Wed, 31 Aug 88 15:46:17 BST
From: Paul Hudson <mcvax!moncam!paul@uunet.uu.net>
Subject: Re: Gnu Emacs under 4.0
Reference: v6n203
Whoops. GNU Emacs 18.50. As previously reported, fixed in 18.51. This
will be my last comment on GNU in SunSpots, honest :-).Paul Hudson
Snail mail: Monotype ADG Email: ...!ukc!acorn!moncam!paul
Science Park, paul@moncam.co.uk
Milton Road,
Cambridge,
CB4 4FQ
------------------------------
Date: Fri, 12 Aug 88 15:33:14 EDT
From: Operator <root@helios.physics.toronto.edu>
Subject: package for printing on VAX printers from Suns
I am enclosing the shell archive containing all the necessary bits and
pieces to make VAX printers accessible, via DECnet, from Suns. You must be
running the SunLink DNI product on at least one Sun in order to make use
of these routines.
As I mentioned in my earlier message, we have tested the package only on
our own machines, on which it works. We can make no guarantees as to its
functionality on your machines, nor can we take responsibility for any
undesired consequences it may have (where have I heard this before? :-) ).
We will not be issuing regular updates or bugfixes; however you are more
than welcome to make improvements yourself. A number of small adjustments
will be needed to adapt the programs to the printers and VAXes you have at
your site.
As a matter of interest, it should be fairly straightforward to adapt the
core routines in this package to make a kind of "vaxsh" program, to
execute VMS commands on a given VAX, e.g. "vaxsh node1 show users".
My apologies for taking so long to get this put together. I have been on
vacation, and some of our employees are leaving/have left, so it was
rather busy before and after I was away.
Enjoy!
Ruth Milner
Systems Manager
University of Toronto Physics
sysruth@helios.physics.utoronto.ca
P.S. As an aside to programmers making changes to the package: under the
current version we are running, DNI 5.0 on a Sun 3/180 with SunOS 3.4, if
you pass the wrong information to the DNI ioctl routine, it will crash the
machine. We have also uncovered problems with certain connections through
dnalogin which, if terminated incorrectly (i.e. killed) will crash the VAX
(!). This one hasn't happened often enough for us to pin it down.
[[ This has been placed in the archives under "sun-source" as
"dnaprint.shar". It is 49162 bytes long and it can be retrieved via
anonymous FTP from the host "titan.rice.edu" or via the archive server
with the request "send sun-source dnaprint.shar". For more information
about the archive server, send a mail message containing the word "help"
to the address "archive-server@rice.edu". --wnl ]]
------------------------------
Date: Wed, 31 Aug 88 16:57:18 BST
From: Paul Hudson <mcvax!moncam!paul@uunet.uu.net>
Subject: Those WREN-IV's again
(I've never had a one-line message generate so much mail!)
As promised to various people a summary of our experiences of WREN-IVs on
Suns.
WREN-IV disks are made by MPI (a subsidary of CDC?). They come in various
flavours. Ours give about 290 MB usable. They're cheap (esp. compared with
SUN pricing), fast, and, so far, reliable.
We have sucessfully run a 3/50 with 4.0 and one WREN-IV, and a 3/160 with
3.4 with 2 of them for some time now. We have also run a 3/60 briefly,
with no problems.
We paid around 1180 UK pounds, from Sintrom (a UK distributor). We added
our own power supply, box, and cables: nothing strange about these. If you
use two or more, remember to take the terminators off all but the last; if
you don't things very nearly work but not quite! I can't remember the
power requirements, and can't find the appropiate bit of paper, but they
don't use much. (Tape drives, however, are a different story ....)
Formatting using 3.4 requires telling diag that you have an Emulex
controller. The documentation onn the drive that we had appears not to be
quite right: it appears that the figures include a sector/cylinder always
reserved by the drive (tut, tut). The tracks and such seemed odd too, so
we assumed it was really a larger relative of a WREN-III, and came up with
1365 cylinders, 48 sectors/track 9 heads. We have found that some drives
can't even manage this but 1350 usable cylinders + 8 alternates has always
worked. Formatting on 4.0 worked fine.
Finally, with these drives, we didn't noticed any particular performance
change when upgrading the 3/50 to 4.0, unlike some previous comments.
Anyone know what's happening?
Paul Hudson
Snail mail: Monotype ADG Email: ...!ukc!acorn!moncam!paul
Science Park, paul@moncam.co.uk
Milton Road,
Cambridge,
CB4 4FQ
------------------------------
Date: 31 Aug 88 11:51:56 I (Wed)
From: truesdel@prandtl.nas.nasa.gov
Subject: Speeding up dumps over the network
At my previous job, I had done quite a bit of work speeding up our dumps.
We had a number of "big" Sun fileservers which needed to be dumped to tape
drives on vaxen (one on each of two separate networks). A number of
comments on what I had done.
1) I ported the 4.3Bsd dump to the Sun's. That didn't take an
excessive amount of effort. I also ifdef'ed the setsockopt, so
sendspace would be increased if possible.
2) I rewrote /etc/rmt to be multi-buffered in the same fashion as dump.
It would build up a ring of processes when starting reads and writes,
and tear it down when doing ioctl's, etc. Also, I left out the
setsockopt to change rcvspace, as you shouldn't do that unless you
*KNOW* that the sender *HAS* increased their sendspace. When reading
from a tape, it would (if possible) try to increase sendspace. Another
tweak was to make sure that the buffers malloc'ed were page aligned.
I'll try to get a diff together of the changes I made to rmt, and post
them. If you want to get a look at the inspiration for these changes,
look in your (I believe it was net.sources) archives for a program
called "stream" by Don Speck, the originator of the dump speedups.
dave truesdell (truesdel@prandtl.nas.nasa.gov)
BTW...
As a rule of thumb: The size of sendspace should be >= The size of
rcvspace. If rcvspace is significantly larger you will have to wait for a
timer to expire for the receiver to send an ACK, instead of sending one as
soon as the receive buffer has filled. This was covered in fair detail
back around the time 4.3 first came out.
------------------------------
Date: Tue, 30 Aug 88 10:10:09 EDT
From: steinmetz!grymoire!barnett@uunet.uu.net
Subject: SunOS 4.0 version of nnstat
I am in the middle of converting the nnstat programs from 3.X to SunOS
4.0.
Nnstat is a powerful network monitoring package available via anonymous
ftp from venera.isi.edu. It can monitor the ethernet in real time, and
collect statistics which can be gathered once a day, etc. You can specify
what type of information to gather, as it comes with a configuration
language.
The network interface tap (NIT) in SunOS 4.0 is STREAMS based, and allows
you to download a buffering package and/or a filtering package into the
NIT.
Anyhow, I have the code *almost* working. If anyone with more knowledge
and/or free time would like to collaborate, I would appreciate any leads.
Bruce G. Barnett <barnett@ge-crd.ARPA> <barnett@steinmetz.UUCP>
uunet!steinmetz!barnett
------------------------------
Date: Wed, 31 Aug 88 15:46:27 -0700
From: bien@aerospace.aero.org
Subject: SunOS 4.0 -- initgroups error
If a user is in more than 8 groups, they cannot log into a SunOS 4.0
system. Root also cannot su to that user id.
The error message you get is:
NFS lookup failed for server <where the home directory is>:
RPC: Authentication error
On 3.5 systems, the user gets the warning message:
initgroups: <user> is in too many groups
but can log in sucessfully.
This has been reported to Sun.
--Cheryl Bien
The Aerospace Corporation
bien@aerospace.aero.org
------------------------------
Date: Wed, 31 Aug 88 14:20:25 EDT
From: Felix Lee <flee@blitz.cs.psu.edu>
Subject: vi core dumps (illegal instruction)
Model: Sun 4/260S
OS: Sun/OS 4-4.0
Problem:
vi dumps core when my environment is around 1750 characters long. Here is
a script of someone else's session:
Script started on Wed Aug 31 11:49:45 1988
shire<1>% setenv X abcd
shire<2>% env | wc
22 52 1753
shire<3>% vi
[everything fine]
shire<4>% setenv X abcde
shire<5>% env | wc
22 52 1754
shire<6>% vi
Illegal instruction (core dumped)
shire<7>% exit
shire<8>%
script done on Wed Aug 31 11:50:08 1988
The range of failure is about 50 characters. Longer and shorter
environments work fine. The location of the range depends on random
unknown factors, but seems to be consistent for each csh.
The failure actually depends on the length of the environment plus the
length of the arguments passed to vi.
Using ex and then saying "open" or "vi" works as long as you're editing a
file. Running ex with no arguments and then saying "open" or "vi" will
fail.
"Illegal instruction" is a bit puzzling...
--
Felix Lee *!psuvax1!flee
------------------------------
Date: Wed, 31 Aug 88 16:42:22 EDT
From: mike@shogun.cc.umich.edu (Michael Nowak)
Subject: Having trouble running standalone copy over the net
I am trying to set up a Sun 3/50 with a disk but no tape. I was able to
boot the stand/diag program from another fully operational Sun 3/60 to
format and label the drive. Now I am at the stage where I need to run
standalone copy to copy the miniroot from the other Sun to the new one. I
am able to boot stand/copy (by typing "b le()stand/copy from the new Sun"
and then I am faced with the From: prompt. The miniroot is in a file
called /usr/minifs on the other Sun.
When I specify "le()minifs" at the From: prompt, it just sits there and
finally says "nd: no file server, giving up". I think I have the nd set
up correctly on the other Sun. The /etc/nd.local file looks like this:
# nd.local 1.2 86/10/07
#
# nd.local - net disk local initialization file
#
# Commands (see nd(4) for details):
#
# son
# soff
# user [ipaddr] [hisunit] [mydev] [myoff] [mysize] [mylunit]
# ether [ipaddr] [etheraddr]
# clear
# version [versionnumber]
#
clear
version 1
user ninja.cc.umich.edu 0 /dev/sd0g 0 -1 0
user ninja.cc.umich.edu 1 /dev/sd0b 0 -1 0
son
The /tftpboot directory on the other Sun looks like this:
total 111
drwxrwxr-x 2 root 512 Aug 31 10:00 ./
drwxr-xr-x 13 root 512 Aug 31 09:04 ../
lrwxrwxrwx 1 root 29 Aug 31 10:00 23C1F00C -> /tftpboot/ndboot.sun3.private*
-rwxr-xr-x 1 root 13616 Nov 12 1987 in.tftpd*
-rwxr-xr-x 1 root 31056 Nov 12 1987 ndboot.sun3.private*
-rwxr-xr-x 1 root 31056 Nov 12 1987 ndboot.sun3.pub0*
-rwxr-xr-x 1 root 31056 Nov 12 1987 ndboot.sun3.pub1*
lrwxrwxrwx 1 root 1 Aug 26 17:18 tftpboot -> ./
I have a directory /pub on the the root which is link to /usr. I can't
figure out what's wrong. I wish stand/copy gave a more specific error
message than ?. Any thoughts?
------------------------------
Date: Wed, 31 Aug 88 11:45:15 PDT
From: pyramid!versatc!tran@sun.com (Tony Tran)
Subject: Yellow Pages problem
Hi. When I do a make after changing Yellow pages password, the following
message appears:
updated passwd
Can't get an address for server .
Can't get an address for server .
pushed passwd
Does anybody know what it means? Was the change propagated from the master
server to all its slave ones?
Tony Tran
Versatec, Inc.
------------------------------
Date: Wed, 31 Aug 88 16:17:30 PDT
From: chaos@gojira.Berkeley.EDU
Subject: 1/4" cartridge tape drives for Sun SCSI?
According to the sales rep here Sun does not sell the SCSI 1/4" cartridge
tape drives by themselves, only with disks.
Delta Microsystems, (415) 449-6881, sells such a box. Does anyone have any
experience with it or them? Comments appreciated.
I am also looking for a used Sun 3 keyboard and a used color display board
for a 3/160. Any sources for this type of equipment in addition to Adcomp
and APEX?
Jim Crutchfield
Physics, UCB
(415) 642-1287
------------------------------
Date: 31 Aug 88 19:48:40 GMT
From: ehrlich@blitz.cs.psu.edu (Dan Ehrlich)
Subject: /usr/lib/sendmail vs /usr/lib/sendmail.mx?
I noticed after installing SunOS 4.0 on a 4/260s that there are two
flavors of sendmail. One is named 'sendmail' and the other is named
'sendmail.mx'. As no reference to 'sendmail.mx' has been located by me in
the SunOS 4.0 documentation set, I might assume that 'sendmail.mx' is a
variant that can deal with internet domain MX records. Is this a valid
assumption? If it is, what sort of magic incantations need to be made in
the sendmail.cf file to invoke this behavior? Any hints would be
appreciated. Thanks in advance.
Dan Ehrlich <ehrlich@blitz.cs.psu.edu>
The Pennsylvania State University
Department of Computer Science
University Park, PA 16802
------------------------------
Date: 31 Aug 88 18:14:35 GMT
From: mcvax!hafro.is!gunnar@uunet.uu.net (Gunnar Stefansson)
Subject: Sun delivery times: HOW LONG?
What is people's general experience with delivery times from Sun ? We
ordered a bunch of Suns from our local dealer back in January, and the
order still isn't complete.
Would you belive they sent us a Sun 4 without the operating system and
without the SCSI interface, both of which we ordered with the machine ?
This was 5 months after we placed the order. When we asked them to please
send the missing parts immediately, it seems they were taking an
inventory, so they plain refused to send the OS right away. It's now
nearly 3 months since we recieved the Sun 4 and not a word from Sun. Oh,
yes, we finally got the Sun 4 opererating system special delivery from
Sweden. No manuals, no disc interface, just 2 tapes. We still haven't
received SunOs 4.0 for the 3/50's, so we can't even install on the 4 over
the network. Of course this was also supposed to come with the Sun 4,
when they finally got around to sending *something*, back in June.
The worst part of our experience is that whenever we contact anyone, our
local dealer, Sun in Sweden or Sun in the US, we are always told that the
problem is totally unusual and will be fixed immediately.
Is this what other users normally experience, is it just the norm in
Europe or are we just plain unlucky ?
If this is the norm, it should be advertised, so customers can place their
order a year or two in advance. We placed the order back in January, to be
absolutely sure to get everything well before a meeting, scheduled in
July. Well, that wasn't long enough ahead. So, Sun guys, what is the true
delivery time, when you say 45 or 60 days ?
Maybe they meant weeks ?
Gunnar
PS Can anyone tell me what to do with a Sun 4/260 with no disc interface
and no operating system ? Too tall for a coffee table.
[[ That's tough. It doesn't even have a front panel on which to do a
light show! --wnl ]]
[gunnar@hafro.is - mcvax!hafro!gunnar]
------------------------------
End of SUN-Spots Digest
***********************