| Title: | CSGUK_SYSTEMS |
| Notice: | No restrictions on keyword creation |
| Moderator: | KERNEL::ADAMS |
| Created: | Wed Mar 01 1989 |
| Last Modified: | Thu Nov 28 1996 |
| Last Successful Update: | Fri Jun 06 1997 |
| Number of topics: | 242 |
| Total number of notes: | 1855 |
"Our Goal is to have the very best UNIX ... and support it heartily"
- Ken Olsen
Jun 1988.
One evening this week an engineer called into the building at about
17:30 and asked for some help with retrieving an Ultrix error log.
I gather that he was passed around and eventually ended up talking
to the one person who had stayed late to work on in the Ultrix TSC.
As it later turned out, all he wanted to know was that you have to
type:
uerf -o full
to see the contents of the Ultrix error log.
So that this doesn't happen again the replies to this note are:
.1 a crib sheet (copy to whoever you like) that provides the
available options to uerf.
.2 A detailed overview of how Ultrix error logging works - including
how you configure it and the files which are involved.
If you need any more info on Ultrix please come down and see Alwyn
Bradley or myself. The last time we bit the heads off any babies was
weeks ago........
I guess we need to get this Ultrix presentation arranged....
Regards
Alan Trevennor
PTG (Ultrix)
| T.R | Title | User | Personal Name | Date | Lines |
|---|---|---|---|---|---|
| 45.1 | How to get errorlogs from Ultrix (V2.0 or later). | COMICS::TREVENNOR | A child of init | Wed Jul 05 1989 12:39 | 97 |
ULTRIX UERF crib sheet.
uerf command options - examples in the comment field on the right.
The error log report formatter is called uerf - anyone can run it
(ie you don't have to be god on the system).
You provide a number of 'options' to the command to make it display
selectively. An option is a '-' sign followed by one or more letters.
/etc/uerf -x # negate all other supplied options
# EXAMPLE: /etc/uerf -D -x
# means display all errors except disks
-R # Display newest errors first
-n # Display errors as they occur
-h # show available uerf options
-H hostname # Display errors from node hostname
# EXAMPLE: /etc/uerf -H PERCY
# Means display all errors which have been
# logged which originated from node PERCY.
-A # Include hardware adapter errors
aie # BVP controller (BI disk controller).
aio # BVP controller (BI disk controller).
bla # BI LESI adaptor
bua # BI to UNIBUS adaptor
nmi # Nautilus memory interconnect
uba # VAX UNIBUS adaptor
# EXAMPLE: /etc/uerf -A bua,bla
# means display all bua and bla errors.
-c # Select specific classes of errors
err # Selects error reports from h/w or s/w
oper # Selects only operational events
-D # Includes disk errors.
rd54 # Include disk errors on RD54 type drives
ra80 # Include disk errors on RA80 type drives
ra60 # Include disk errors on RA60 type drives
etc # Any other DSA compatible drive type
# EXAMPLE: /etc/uerf -D ra60
# means display all errors logged on ra60's.
# EXAMPLE: /etc/uerf -D -o full
# means display all disk errors in full format
-f filename # Allows you to display errors from any
# current or archived error log file.
-M # requests that only errors to do with the
cpu # central hardware elements of the machine
mem # be displayed. The memory and the CPU..
# EXAMPLE: /etc/uerf -M mem -o full
# means display all memory errors in full
-o # Selects output format. brief is the
terse # default. Beware non-selective use of
brief # full, it can generate a long listing!
full #
-O # Operating system detected errors.
aef # Arithmetic exceptions (can be h/w or s/w)
ast # async trap exception faults (h/w or s/w)
bpt # Break point instruction fault (often h/w)
cmp # Compatibility mode faults (h/w possibly s/w)
pag # Page fault exceptions (h/w)
pif # Privileged instruction faults (h/w or s/w)
pro # protection faults (h/w or s/w)
ptf # Page table faults (h/w or corrupt s/w)
raf # Reserved address Faults (h/w or corrupt s/w)
rof # Reserved Operand Faults (corrupt s/w usually)
scf # System Service Call Exception faults (either)
seg # Segmentation Faults (h/w usually)
tra # Trace exception faults (h/w or s/w)
xfc # Xfc instruction failed (h/w)
# EXAMPLE: /etc/uerf -O seg -o full
# Means display full details of all
# segmentation faults.
-r n # Selects errors by record type - see manual.
-s nn-nn # Each error is assigned a number when it
# is logged. This option selects a specific
# number or range of numbers to be reported
-T # Includes tape errors in report
tk50 # Include errors which occurred on tk50
tu81 # Include errors which occurred on tu81
etc # Any other Digital tape drive type
-t # Selects by absolute or range of time.
| |||||
| 45.2 | Ultrix error logging description and overview. | COMICS::TREVENNOR | A child of init | Wed Jul 05 1989 12:40 | 423 |
Ultrix Error logging
Ultrix error logging.
A system which cannot provide information to assist hardware
engineers in fault diagnosis does little to enhance customer
confidence. Consequently, considerable effort was put into providing
full error logging for version 2.0 of Ultrix-32.
The components of Ultrix which are capable of detecting hardware
errors are principally the device drivers, but the fault detection
features built into VAX hardware can also cause an error to be logged
by means of an interrupt. An example of this is a hardware detected
memory error.
Figure one shows the overall concept of error logging under Ultrix
and shows how the three major programs in the errorlogging suite
interact together. These three programs are:
/etc/elcsd
/etc/uerf
/etc/eli
The software components of the system which can detect or declare
errors place packets of binary information detailing the error (or
system event) into the kernel errorlog buffer. Lets look at these three
programs in more detail.
The elcsd daemon.
The file /etc/elcsd (ErrLog Copy/Server Daemon) is the Ultrix error
logging daemon. It is run at system startup time via an entry in the
system startup script "/etc/rc". It takes errorlog packets placed in a
memory area known as the kernel errorlog buffer (identified by the
character special file /dev/errlog) and writes them into the error log
file of your choice.
Using the system timer mechanism, elcsd wakes up at regular
intervals and scans the errorlog buffer for new packets. Any new
packets found are appended to the currently selected error log file
(see later). The error log file need not be resident on the local node,
it can be on a disk attached to another node in the same network.
Configuring error logging.
So, how do you set up your error logging to best effect? The
operational parameters for error logging are specified by the contents
of a file called
/etc/elcsd.conf
The contents of this file are the parameters which define the error
logging daemon's aims in life. Ultrix allows you to centralise on one
node error logging for all the Ultrix nodes in a network. Implicit in
this is an ability of any node to log errors; locally, from any other
node in the network, or to send details of errors to any other node.
Figure two shows a network of three VAX machines connected via an
ethernet. Let's assume that the node GORDON is a VAX 8350, and the node
JAMES is a microVAX II and that the node THOMAS is a VAXstation 2000
series machine. To simplify the task of keeping tabs on the errorlogs
the person who is responsible for managing the machines in this set up
designates GORDON as the error logging node.
To effect this, the network manager can select the option of having
details of all errors logged into the standard error log file on GORDON
which is:
/usr/adm/syserr.GORDON
Alternatively a separate file can be created on GORDON's file system to
contain the log for each node. These will be:
/usr/adm/syserr.JAMES
/usr/adm/syserr.THOMAS
The way that this works is via an entry in the file:
/etc/elcsd.conf
When the elcsd daemon in each node starts up, it reads the contents of
this file and takes the third parameter in it as the pathname to the
primary error log file. The seventh parameter in the file specifies the
name of the node where this file exists. If the seventh parameter is
blank (as it is just after installation of Ultrix) then the file is
assumed to be on the local file tree.
Let's now look in detail at the eight parameters in the elcsd
configuration file and what they mean.
1) Status parameter one is a bitwise integer with a value from 0-7
where the bits may be set in any sensible combination. The bits are
used as follows:
Status parameter one is a bitwise integer with a value from 0-7 where
the bits may be set in any sensible combination. The bits are used as
follows:
+-------+-------+-------+
| 4 | 2 | 1 | <Bit weight.
+-------+-------+-------+
2 1 0 <Bit number.
| | |
/ \ \
/ \ \
/ \ \
If set, log errors \ \___________ If set then errors occurring
to an errorlog | on the local node are to be
file on a remote | logged.
system named in |
parameter 7. |
|
If set specifies that
errors from remote
machines are to be logged.
Examples: Set to 5 means errors occurring on this machine are to be
logged to both the local error log file and the remote one
named in parameter 7.
Set to 2 then only errors received from elscd daemons on
remote machines are to be logged.
2) Parameter two is the maximum size (in bytes) to which you wish to
limit the error log file. If you leave this parameter blank then the
error log will expand forever, or at least until the file system is 98%
full! The size you choose should reflect how often you intend to check
and clear the error log for your system or network.
3) The error log directory path. This is the pathname to which details
of errors will be sent for later retrieval and output. The pathname you
specify here can be anywhere on your file tree, but remember that when
you run the error log report formatter it assumes by default that
error logs are in the file:
/usr/adm/syserr.HOSTNAME
4) Ultrix allows you to specify a secondary - or backup - error log
file to be used in the event that the primary file (specified in
parameter 3) is not usable. If, for example, you have a two drive
system and your primary error log file is on one disk you could put
your backup error log file on the other drive. This would ensure that
if the first disk ever became inaccessible, the details of the error
would still be logged on the other disk.
5) This parameter provides the pathname of a file to which errors are
logged when the system is in single user mode. This is different to the
file specified in parameter 3, because you need to ensure that the file
system will be mounted in single user mode. When single user mode is
exited any errors logged into this file are appended to the main file
specified in parameter 3.
6) This parameter defines where incoming error logs from remote systems
are to be logged. Note that this is a pathname, not a file name. The
file will always be HOSTNAME. So if this parameter were set to:
/etc/errlogs
then error logs from node THOMAS will be placed in the file:
/etc/errlogs.THOMAS
7) If you are sending details of errors logged on your system to a
central logging node, then parameter 7 should be set to the name of the
machine - in our previous example this will be GORDON.
8) If your node is logging errors on behalf of other nodes then the
names of these should appear as parameters 8 through n. On the same
line as each node name should appear a colon, followed by either a
letter S or a letter R. R means that a file called:
/etc/syserr.REMOTES
is to be used as the errorlog file (so that all remote node errorlogs
can be grouped into one file. S means to create a separate file on the
central node for each node whose errors are being logged. In our
example network these files would be:
/etc/syserr.PERCY
/etc/syserr.THOMAS
and parameter 8 will look like this:
PERCY:S
THOMAS:S
The elcsd.log file.
It is quite possible for the elcsd daemon to encounter problems whilst
going about its business. For example, a file system may become full,
or a disk may suddenly drop offline due to a hardware fault. Under such
exceptional circumstances the continuation of error logging may become
impossible. In order to attempt to notify the system manager of a
cessation of normal error logging, elcsd will try to place an
appropriate error message into:
/usr/adm/elcsdlog
which is elcsd's log file. This file is cleared at every system reboot.
The messages also go directly to the terminal of any user whose name
appears in the file:
/etc/syslog.conf
(reference syslog(8).) The contents of the file are split into two
areas, separated by a blank line. The entries above the blank specify
to which files the system event logs will be copied. The lower half of
the file consists of a list of users names. Each user named will
receive system messages which fall into the "alert" category. Almost
all the messages which elcsd issues on its own behalf have system alert
status. An example of a message issued by elcsd is:
exceeded error log file limit size, error not logged to disk!
When you get messages directly from elcsd you should take prompt and
effective action. If there is a severe hardware problem preventing disk
access then data may be being lost. If the errorlog file is full then
you may be logging lots of recoverable errors on a rapidly degrading
disk, which you ought to be trying to take a tape dump of before total
failure occurs. In any case elcsd is usually saying that system error
events are not being logged, and that what may later prove to be vital
information is being lost.
Having looked at how the details of errors are logged by Ultrix,
let's now look at the methods available to collate and display this
information into reports. These reports will be mainly of use to your
field service engineer who has to rectify the problem which occurred.
Obtaining error log reports.
The program used to read the binary contents of the error log file
and turn it into a report is called uerf (Ultrix Error Report
Formatter). You do not need to be superuser to use uerf. By default
uerf lives in:
/etc/uerf
The command:
/etc/uerf -h
provides a summary of the options available for use with uerf. Also in
the /etc directory are three supporting files. These are:
/etc/uerf.bin # database of event information
/etc/uerf.hlp # a text file giving brief help on uerf
/etc/uerf.err # uerf's private error log (see later)
If required, you can move uerf and the three associated files to
another place on your file system and still have it work successfully,
because it uses the following search list to locate its supporting
files:
1) The directory where you ran uerf from
2) The /etc directory.
3) The pathname specified in your shell PATH variables - see setenv(8)
Vanilla flavour uerf is invoked by typing:
/etc/uerf
and then you get the brief format report output to your terminal. If
you are on a softcopy terminal then use the command:
/etc/uerf | more
the see the output a screen at a time. Often the default report
obtained using these commands will be sufficient to give an idea of
what has gone wrong. But what if your error log file is full of tape
retry logs from yesterday and you now want to see the details of an
uncorrectable memory error? This is where the numerous options
available with uerf are used.
Figure three is a quick reference to the options and option
specifiers you can use with uerf. The following are examples of uerf
commands, refer to figure three for all the options.
uerf command examples.
/etc/uerf -T tk50 -t s:01-jan-1988,00:00 e:05-feb-1988,23:59
# Display all tk50 errors logged between midnight
# 1st January 1988 and one minute before midnight
# on 5th Febuary 1988. (If either time specifier
# omitted with the -t option defaults to today).
/etc/uerf -D -t s: e:
# Display all disk errors logged today in default
# brief format.
/etc/uerf -M cpu
# Display all errors detected on the VAX Central
# Processing Unit.
/etc/uerf -O cmp,seg -t s:01-aug-1988 -o full
# Display full details of all operating system detected
# compatibility mode or segmentation errors which have
# been logged between the 1st August 1988 and today.
/etc/uerf -s 233-299 -o terse
# display terse details of errors whose sequence numbers
# are between 233 and 299
The eli utility.
The /etc/eli program provides a user interface to the error logger,
allowing the system manager to change the error logging parameters
during time sharing. Its primary use is to restart error logging after
changing the contents of the configuration file. To do this type:
/etc/eli -a
which stops elcsd and restarts it, thus forcing it to read the new
configuration parameters. The most useful of the other options are
shown below, use the -h option to obtain a full list.
/etc/eli -d # Disables error logging by stopping elcsd
/etc/eli -e # Enables error logging when the system is in multi user
# mode.
/etc/eli -i # Flushes the kernel error log buffer
/etc/eli -n # Disables error logging to disk.
/etc/eli -s # Enables single user mode error logging
F I G U R E T H R E E
Fig 3. uerf command options.
/etc/uerf -x # negate all other supplied options
# EXAMPLE: /etc/uerf -D -x
# means display all errors except disks
-R # Display newest errors first
-n # Display errors as they occur
-h # show available uerf options
-H hostname # Display errors from node hostname
# EXAMPLE: /etc/uerf -H PERCY
# Means display all errors which have been
# logged which originated from node PERCY.
-A # Include hardware adapter errors
aie # BVP controller (BI disk controller).
aio # BVP controller (BI disk controller).
bla # BI LESI adaptor
bua # BI to UNIBUS adaptor
nmi # Nautilus memory interconnect
uba # VAX UNIBUS adaptor
# EXAMPLE: /etc/uerf -A bua,bla
# means display all bua and bla errors.
-c # Select specific classes of errors
err # Selects error reports from h/w or s/w
oper # Selects only operational events
-D # Includes disk errors.
rd54 # Include disk errors on RD54 type drives
ra80 # Include disk errors on RA80 type drives
ra60 # Include disk errors on RA60 type drives
etc # Any other DSA compatible drive type
# EXAMPLE: /etc/uerf -D ra60
# means display all errors logged on ra60's.
# EXAMPLE: /etc/uerf -D -o full
# means display all disk errors in full format
-f filename # Allows you to display errors from any
# current or archived error log file.
-M # requests that only errors to do with the
cpu # central hardware elements of the machine
mem # be displayed. The memory and the CPU..
# EXAMPLE: /etc/uerf -M mem -o full
# means display all memory errors in full
-o # Selects output format. brief is the
terse # default. Beware non-selective use of
brief # full, it can generate a long listing!
full #
-O # Operating system detected errors.
aef # Arithmetic exceptions (can be h/w or s/w)
ast # async trap exception faults (h/w or s/w)
bpt # Break point instruction fault (often h/w)
cmp # Compatibility mode faults (h/w possibly s/w)
pag # Page fault exceptions (h/w)
pif # Privileged instruction faults (h/w or s/w)
pro # protection faults (h/w or s/w)
ptf # Page table faults (h/w or corrupt s/w)
raf # Reserved address Faults (h/w or corrupt s/w)
rof # Reserved Operand Faults (corrupt s/w usually)
scf # System Service Call Exception faults (either)
seg # Segmentation Faults (h/w usually)
tra # Trace exception faults (h/w or s/w)
xfc # Xfc instruction failed (h/w)
# EXAMPLE: /etc/uerf -O seg -o full
# Means display full details of all
# segmentation faults.
-r n # Selects errors by record type - see manual.
-s nn-nn # Each error is assigned a number when it
# is logged. This option selects a specific
# number or range of numbers to be reported
-T # Includes tape errors in report
tk50 # Include errors which occurred on tk50
tu81 # Include errors which occurred on tu81
etc # Any other Digital tape drive type
-t # Selects by absolute or range of time.
| |||||
| 45.3 | Which version of Ultrix am I running? | COMICS::TREVENNOR | A child of init | Wed Jul 05 1989 12:44 | 17 |
Q: How do I find out what the version of my Ultrix system is?
A: Log into the system and type the command:
strings /vmunix | grep Ultrix
This says, produce a list of all the printable strings you can
find in the currently running Ultrix Kernel, and show me any of
them which contains the word 'Ultrix'. Strings is the command to
scan a file for strings, and grep is the equivalent of the search
command in VMS.
Alan T.
| |||||
| 45.4 | Auto - booting Ultrix from other than DU0 on a uVAX. | COMICS::TREVENNOR | A child of init | Wed Jul 05 1989 12:46 | 61 |
Problem:
The customer has a dual drive microVAX2 system. He wishes
the system contained in partition A of drive one to be
automatically booted when the system restarts. Currently,
the system attempts to boot drive zero and fails because
there is not an Ultrix system on it.
BACKGROUND:
The key to this problem is understanding how the code in the
MicroVAX ROM boots the system.
The Virtual Memory Boot program (VMB.EXE) is used to bootstrap
the operating system from one of the disk or tape storage
devices on the system. The version of VMB which resides in the
MicroVAX bootstrap ROMs differs from the version which is loaded
into memory from a console storage device. ROM resident VMB
contains extra routines which search for a bootable
storage device and, when a device is found to be ready, the
first block is loaded from it, and then usually executed.
Although it is not a formal standard, certain conventions are
followed in Digital bootblocks for VAX disks. This makes it
easy for VMB to examine a loaded bootblock to see if the disk
looks to be bootable. If VMB cannot see any indication of a
valid bootstrap in Logical Block Number (LBN) zero of a disk
it will abort its attempt to boot that disk, and proceed to
the next disk.
NOTE: This behaviour of VMB is only valid for MicroVAX
machines. The version of VMB which is loaded from
console or system media (eg TU58, RX50 or RA
Disks) require you to specify a device to boot
and halt the machine if that device is not
bootable.
SOLUTION:
The following proceedure is NOT supported by Digital. It is
suggestion as to how an individual customer could approach the
solution to a particular problem.
If a disk does not contain a bootable Ultrix system the first
16 blocks are never used. On a bootable disk these will contain
the bootblock in LBN 0, and the program vaxboot in LBN 1-15.
The disk initialisation utility supplied with Ultrix always
writes a valid bootblock - although the pointers within that
bootblock are set to absolute default values.
To make a disk 'look' unbootable to the 'sniffer' boot routines
of ROM-based VMB all that is required is to fill the bootblock
with all zeroes, using some system calls in a privileged program
to do the work.
Extreme care is required when doing this, because if anything
after LBN 15 is overwritten, the file systems on the disk will
be invalidated and possibly lost.
| |||||
| 45.5 | The All New(?) DS5800. | KERNEL::EDMUNDS | Tue Oct 10 1989 11:18 | 9 | |
I recently attended an "ISIS"(DS5800) pilot course in Paris.
This is the new RISC Ultrix(only) system and the instructor has
forwarded a very useful lookup guide to me.
I am including this in the next reply,if you have any questions
I will be happy to answer them/play dead as appropriate.
Cheers
Steve Ed.
| |||||
| 45.6 | DS5800/Ultrix info. | KERNEL::EDMUNDS | Tue Oct 10 1989 11:20 | 664 | |
Kernel debugging on ULTRIX/RISC
by
Al Delorey
ULTRIX(tm) Engineering Group
abyss::afd
Copyright (c) 1989 by
Digital Equipment Corporation, Maynard, MA
RISC debug - afd
Address Space
The system is always in virtual address mode (no physical address mode)
Address spaces
KSEG0 - not mapped, cached; for kernel text
virtual: 8000 0000 -> 9fff ffff (512 MB)
KSEG1 - not mapped, not cached; for I/O space
virtual: a000 0000 -> bfff ffff (512 MB)
KSEG2 - mapped, cached; for stacks, kernel mallocs
virtual: c000 0000 -> ffff ffff (1 GB)
KUSEG - mapped, cached; for user space
virtual: 0 -> 7fff ffff (2 GB)
Stacks
No interrupt stack, only Kernel and User stacks [idle stack added in 4.0]
Startup Stack - at 0x8001f7ff (0x8002ffff in 3.0/RISC) -
grows downward and is used during system startup,
until a kernel stack is available.
Kernel Stack - starts at 0xffffe000 (kseg2 space)
User Struct - starts at 0xffffc000 (kseg2 space)
Per cpu data base - starts at 0xffff8000 (kseg2 space) [as of 4.0 - smp]
User Stack - starts at 0x7ffff000 (kuseg space)
(one guard page 0x7ffff000 to 7fffffff)
RISC debug - afd
Using nm
For a system crash that gives an Exception PC (EPC) on the console,
you can use nm(1) to determine what routine was executing.
nm -n /vmunix
This command will display the name list (symbol table), in numerical
order, of the vmunix image. Find the address that is closest to (but
less than) the given EPC from the crash. That address is the starting
address of the routine that was executing.
You can then subtract the start address of the routine from the
faulting PC, to determine the offset from the beginning of the
routine where the error occured. Then using dbx the offending
instruction can be found.
Sample nm output
----------------
First Kernel text address: 8003,0000 (192k bytes above 8000,0000)
80030000 T start
80030000 T eprol
800300ac T putstr
80030148 T lputc
8003018c T cn_reset
...
First Kernel data address: is approximately 8011,0000
80112030 D Sysmap
8011c830 D Usrptmap
8011f920 D camap
8011f930 D kmempt
8011f930 D ecamap
80123930 D Forkmap
RISC debug - afd
Overall system memory map (in sys area see mips/entrypt.h)
physical kseg1 use
-------- ----- ---
0x00030000 0xa0030000 upward Ultrix kernel text, data, and bss
0x0002ffff 0xa002ffff
to additional Prom Space (64K)
0x00020000 0xa0020000
0x0001ffff 0xa001ffff
to 1K netblock (host/client net boot info)
0x0001fc00 0xa001fc00
0x0001fbff 0xa001fbff
to 1K Ultrix Save State area (NEW in 3.1)
0x0001f800 0xa001f800
0x0001f7ff 0xa001f7ff downward 1K Ultrix temporary startup stack
| (at 0x2ffff in 3.0; here in 3.1)
v
0x0001f400 0xa001f400
0x0001f3ff 0xa001f3ff downward dbgmon stack (a few K less than 64K)
|
V
^
|
0x00010000 0xa0010000 upward dbgmon text, data, and bss
0x0000ffff 0xa000ffff downward prom monitor stack
|
V
^
|
0x00000500 0xa0000500 upward prom monitor bss
0x000004ff 0xa00004ff
to restart block
0x00000400 0xa0000400
0x000003ff 0xa00003ff
to general exception code
0x00000080 0xa0000080 (note cpu addresses as 0x80000080)
0x0000007f 0xa000007f
to utlbmiss exception code
0x00000000 0xa0000000 (note cpu addresses as 0x80000000)
RISC debug - afd
Some Usefull dbx Commands
Command (alias)
---------------
alias [name[(args)]cmd] show all aliases or define an alias
assign (a) var=value assign a value to a program variable
stop at (b) set a breakpoint at a given line
cont (c) continue after breakpint
delete (d) delete the given item from the status list
down move down an activation level in the stack
dump print variable info for current routine
dump . print global variable info for all routines
file what is the current file
file filename set the current file to 'filename'
func (f) set context to specified func name (selects the file)
history (h) print history list
status (j) show status list, shows breakpoints (journal)
list (l) list the next 10 lines of source code
l line:range list 'range' lines of code, starting at 'line'
next (n or S) step specified # of lines (don't stop in calls)
nexti (ni) step specified # of assembly lines (don't stop in calls)
print (p) print the value of the specified expr or var
pd print the value of the specified expr or var in decimal
po print the value of the specified expr or var in octal
px print the value of the specified expr or var in hex
note: Can't print register variables.
pr print values of all registers
quit (q) exit dbx
run [args] run the program with specified cmd line args
rerun (r) rerun the program with same arguments
return finish executing the function and stop back in caller
set show setting of dbx variables
set $var=value set a value to a dbx var (can define a new variable)
step (s) step specified # of lines (stopping in calls)
stepi (si) step specified # of assembly lines (stopping in calls)
stop at (b) set a breakpoint at a given line
stopi at <addr> set a breakpoint at a given assembly instruction addr
u list the previous 10 lines
unset $var unset a dbx variable
up move up an activation level in the stack
w list 5 lines before and after current line
W list 10 lines before and after current line
where (t) where are we & how did we get here (stack trace)
this can also be done when stopped at a breakpoint
(this will show where/how a system crashed)
whatis <var> show the variable/symbol definition
whereis <var> show all versions of the specified variable
which <var> print the default (current) version of the variable
/<regex> search ahead in the source code for the regular expr
?<regex> search back in the source code for the regular expr
!<history-item> specify a cmd from the history list (by number or str)
line edit there is an emacs like line edit capability. To enable
it you must set the shell environment variable LINEEDIT
(setenv LINEEDIT)
^A Move cursor to start of line
^B Move cursor back one char
^D Delete char at the cursor
^E Move cursor to end of line
^F Move cursor ahead one char
^H,delete Delete char preceding cursor
^N Move ahead one cmd line in hist list
^P Move back one cmd line in hist list
RISC debug - afd
signals
-------
catch list signals that dbx will catch
catch SIGNAL add a signal to the catch list
ignore list signals that dbx will ignore (not catch)
ignore SIGNAL add a signal to the ignore list
History
-------
Ultrix/RISC dbx is (based on) MIPS dbx.
Both Ultrix/VAX dbx and MIPS dbx are based on BSD dbx.
Mips did a lot of work on dbx: enhanced it to work on the kernel.
There is no adb for Ultrix/RISC
RISC debug - afd
Using dbx to debug the kernel
dbx -k vmunix.n vmcore.n
t get a back trace (to show where/how the system crashed)
routine-name/ni dump out n instructions from given routine
&<symbol>/<fmt> print address and contents of a symbol
<address>/<cnt><mode> print the contents of image at the given address
valid modes are:
d,D short, long decimal
o,O short, long octal
x,X short, long hex
c a byte as a char
s a null-terminated string
f single precision real
g double precision real
i machine instructions
example:
If the system crashes and reports an EPC of 0x8000dead, then dbx can be
used to determine where in the kernel that PC is located.
0x8000dead/9i decode 9 instructions (& show line #s) @ 0x8000dead
Beware: that code that's ifdef'ed out will not count
in dbx's line numbering
p gnode[n] print the gnode struct n in the gnode table
p text[n] print the text struct n in the text table
set $pid=n set processs context to given pid
can then do trace, p *up, p *up.u_procp, etc. on proc
p *up print the u_area (of current proc)
p *up.u_procp print the proc struct of the current pid ("$pid")
Using dbx on running vmunix
---------------------------
dbx -k /vmunix
&<symbol>/<fmt> print address and contents of symbol
a <symbol>=<value> to change the value of a symbol (must be root)
RISC debug - afd
Examining the Exception Frame
All error traps & interrupts (except cache parity errors) generate
an "exception condition".
Exception conditions trap to VECTOR(exception) in locore.s.
Exception routine saves state in the exception frame (on stack).
For interrupts, VECTOR(VEC_int) is called, which saves additional
state on the exception frame, & calls intr() (in trap.c).
intr() calls the specific interrupt handler thru "c0vec_tbl".
For traps, the individual trap routine is called thru the "causevec",
these routines (VEC_addrerr, VEC_ibe, VEC_dbe) in
turn call VECTOR(VEC_trap), which saves additional state on the
exception frame, and calls trap() (in trap.c).
A pointer to the exception frame (ep) is passed as an argument
to the following routines:
trap, intr, tlbmod, tlbmiss, syscall
Thus by using dbx to get a trace, you can find the address of the
exception frame (the ep argument). You can then dump out the exception
frame with a dbx command like:
dbx> 0xffffnnnn/41X
(cont on next page)
RISC debug - afd
Examining the Exception Frame (cont)
The offsets within the exception frame are defined as follows (see mips/reg.h):
#define EF_ARGSAVE0 0 /* arg save for c calling seq */
#define EF_ARGSAVE1 1 /* arg save for c calling seq */
#define EF_ARGSAVE2 2 /* arg save for c calling seq */
#define EF_ARGSAVE3 3 /* arg save for c calling seq */
#define EF_AT 4 /* r1: assembler temporary */
#define EF_V0 5 /* r2: return value 0 */
#define EF_V1 6 /* r3: return value 1 */
#define EF_A0 7 /* r4: argument 0 */
#define EF_A1 8 /* r5: argument 1 */
#define EF_A2 9 /* r6: argument 2 */
#define EF_A3 10 /* r7: argument 3 */
#define EF_T0 11 /* r8: caller saved 0 */
#define EF_T1 12 /* r9: caller saved 1 */
#define EF_T2 13 /* r10: caller saved 2 */
#define EF_T3 14 /* r11: caller saved 3 */
#define EF_T4 15 /* r12: caller saved 4 */
#define EF_T5 16 /* r13: caller saved 5 */
#define EF_T6 17 /* r14: caller saved 6 */
#define EF_T7 18 /* r15: caller saved 7 */
#define EF_S0 19 /* r16: callee saved 0 */
#define EF_S1 20 /* r17: callee saved 1 */
#define EF_S2 21 /* r18: callee saved 2 */
#define EF_S3 22 /* r19: callee saved 3 */
#define EF_S4 23 /* r20: callee saved 4 */
#define EF_S5 24 /* r21: callee saved 5 */
#define EF_S6 25 /* r22: callee saved 6 */
#define EF_S7 26 /* r23: callee saved 7 */
#define EF_T8 27 /* r24: code generator 0 */
#define EF_T9 28 /* r25: code generator 1 */
#define EF_K0 29 /* r26: kernel temporary 0 */
#define EF_K1 30 /* r27: kernel temporary 1 */
#define EF_GP 31 /* r28: global pointer */
#define EF_SP 32 /* r29: stack pointer */
#define EF_S8 33 /* r30: callee saved 8 */
#define EF_RA 34 /* r31: return address */
#define EF_SR 35 /* status register */
#define EF_MDLO 36 /* low mult result */
#define EF_MDHI 37 /* high mult result */
#define EF_BADVADDR 38 /* bad virtual address */
#define EF_CAUSE 39 /* cause register */
#define EF_EPC 40 /* program counter */
RISC debug - afd
Examining any Process in the System
ps -axlk vmunix.n vmcore.n - Flags (see ps(1))
-a All processes (not just your own)
-x Even processes w/ no tty
-l Long format (more info given)
-k Kernel files given
- get <pid> of process that you want to look at
Back in dbx...
set $pid=n set processs context to given pid (in dbx)
Can then do t (trace), p *up, p *up.u_procp, etc. on the process
The process' stored registers in the u_area are in "exception frame format"
and can be obtained as follows:
px up.u_ar0[n]
RISC debug - afd
Forcing a Panic (not hung)
dbx -k /vmunix /dev/mem - run as root to write
a ln_softc=0
- will panic on next network interrupt
(even works in single user mode)
(Don't do this if system is diskless or
it won't dump)
a gnodeops=0
- will also panic the system
Note: don't bash the proc struct, or then dbx can't work on the image
Note: don't bash the console structs or you won't see the panic messages
RISC debug - afd
Forcing a Memory Dump on DS3100/2100 when hung
If you set the bootmode to 'r' (restart), then when the restart
button (on a DS3100) is pressed, the system will do a memory dump,
and then a reboot, as opposed to halting and clearing memory.
Note that the dump may be silent, so be patient.
To set the bootmode to restart use the console command:
>>> setenv bootmode r
If the system "hangs" or drops into console mode without doing
a memory dump, the memory dump routine can be started manually.
If a DS3100 "hangs", you can press the reset button to enter console
mode. The default action on the DS3100 is for the reset to re-initialize
memory. To prevent this (preserve memory), set the bootmode to debug
by typing the following command in console mode (prior to debugging a
"hang" situation):
>>> setenv bootmode d
With bootmode set this way, if the system is "hung" you can press the
reset button to enter console mode (with memory contents preserved).
The crash dump code can then be run by typing the "go" command with
a special address (the kernel start address + 8) that will call the
memory dump routine. In Ultrix V3.0/3.1 the kernel start address is
0x80030000, so the dump routine is started by:
>>> go 0x80030008
If the system was in multi-user mode when the reset button was
pressed, then the dump will occur silently, no messages will be
printed. The memory dump will take several minutes, then the
console prompt will re-appear. After the dump is completed, you
can re-initialize the system and reboot as follows:
>>> init
>>> auto
Note: When bootmode is set to 'd' it is important to type "init" before
"boot" or "auto" when the system has been shutdown to console mode,
or reset to console mode. Failure to use the init command may
cause the system boot to fail.
RISC debug - afd
Forcing a Memory Dump on DS5400
Set the break enable switch up (the dot in the circle).
Press the break key to get the console prompt.
The crash dump code can then be run by typing the "go" command with
a special address that will call the memory dump routine:
>>> go 0x80030008
RISC debug - afd
Forcing a Memory Dump on DS5800
Set the break enable switch up (the dot in the circle).
Press the break key to get the console prompt.
The crash dump code can then be run by typing the "go" command with
a special address that will call the memory dump routine:
>>> go 0x80030008
RISC debug - afd
Debugging "hung" systems
(Finding the real kernel stack)
When you force a dump from a "hung" system, the standard back trace done
by dbx will not be useful for the currently active process.
Dbx will get the process context out of the u_area, which is old.
That is, the u_area will have the process context for the last time
that the process was context switched out.
The kernel stack for each process in the system is located at virtual
address 0xffff,e000 in KSEG2 space. The system has an array of NPROC
u_areas that are 8k bytes each. Even though each user process has its
u_area and kernel stack at the same virtual address in KSEG2 address
space, each uarea/kstack maps to a unique physical address.
On context switches the first 2 entries in the TLB ("safe entries")
are set up to map the u_area and kernel stack for that user process.
Kernel Stack: 0xffff,e000 +-------+ higher addresses
| . |
| . |
| . | 8 K bytes for kernel stack
| v | and u area
| | In Kseg2 space (see param.h)
|_______|
| ^ |
| | |
| | |
U area: 0xffff,c000 +-------+ lower addresses
Within dbx, you can dump out the kernel stack with a command such as:
0xffffd000/1028X
This will dump the kernel stack from low to high memory (most recent events
to oldest events).
RISC debug - afd
Examining stack frames with dbx
odump(1)
The utility odump(1) can be used to get a symbol table dump of vmunix.n
odump -P vmunix.n > vmunix.syms
See /usr/include/sym.h (struct runtime_pdr) for the format of the
runtime procedure descriptor created by the loader.
The "fpoff" field as shown by odump is the frame size for the particular
procedure entry.
The general format of the stack (stack frames) is:
high memory +---------------+
| arg n | Space for all args, even though
| . | first 4 args passed in regs
| . |
| . |
virtual frame | arg 1 |
ptr -> /|---------------|\
frame < | local vars | \
offset \|---------------| \
|saved R31 (ret)| \
|...............| \
|more saved regs| > framesize
| 16-23, 30 | /
|---------------| /
| arg passing | /
| area | /
stack ptr -> |---------------|/
(framereg) | . |
| . |
| . |
| |
low memory +---------------+
Using this information, you should be able to work your way back up
the call history on the stack.
Examples of usage are in libexc: unwind.c, exception.c, exception.h
RISC debug - afd
More on Examining stack frames with dbx
You may find it equally productive to start at the top (high memory) end
of the kernel stack and look for the return address of VEC_syscall on the
stack. This is where VEC_syscall calls "syscall" and the stack frame for
entry into "syscall" has the return address of VEC_syscall saved on the
stack. Using the following dbx command will show the instructions in
VEC_syscall, and in particular where "syscall" is called, thus you can
see the return address that will be on the stack.
(dbx) VEC_syscall/30i
[VEC_syscall, 0x800c3868] ori r5,r16,0x1
[VEC_syscall:590, 0x800c386c] mtc0 r5,sr
[VEC_syscall:591, 0x800c3870] sw r2,20(sp)
[VEC_syscall:592, 0x800c3874] sw r3,24(sp)
[VEC_syscall:593, 0x800c3878] move r5,r2
[VEC_syscall:594, 0x800c387c] move r6,r16
[VEC_syscall:595, 0x800c3880] jal syscall
[VEC_syscall:595, 0x800c3884] nop
**> [VEC_syscall:596, 0x800c3888] bne r2,r0,0x800c3810
[VEC_syscall:596, 0x800c388c] nop
The return address will be: 0x800c3888
Using dbx in this way and the dump of the kernel stack, you can pick and
guess your way down the stack to find where the system went.
RISC debug - afd
Disassemble Utility
dis -p routine image-file - will disassemble a routine in the image file
see dis(1)
RISC debug - afd
Some Usefull Console Commands
Dump
dump -w -x ADDR#CNT - dump the contents of memory, starting
at given ADDR & dumping CNT locations
(long words in hex format)
dump -w -x ADDR:ADDR - dump the contents of memory, starting
and ending at given ADDRs
(long words in hex format)
dump -w -x 0x8001f400:0x8001f800 - dump the startup stack
Examine
e [-(b|h|w)] ADDR - examine byte, halfword, word;
ADDR is a virt addr; to examine
physical loc 0 use 0x80000000
Go
go [pc] - transfer control to given entry point
Help
help [cmd]
? [cmd] - if no cmd given, display cmd menu
Printenv
printenv [evar] - display current value of specified
environment variable
Setenv
setenv EVAR STRING - set the specified environment
variable to the given string
environment variable
Unsetenv
unsetenv EVAR - remove the environment variable
from the environment variable table
Test
t a - test all components and subsystems
Booting
auto - use environment variable "bootpath"
to boot to multiuser: DS3100/2100 only
boot - use environment variable "bootpath"
boots to singleuser on DS3100/2100
boots to multiuser on other systems
boot -s - boot to single user (this cmd option
not on DS3100/2100)
boot -f rz(CTRL,UNIT,PART)vmunix- boot the specified image to singleuser
boot -f mop() - boot from the network to singleuser
boot ... memlimit=<#bytes of mem> - to artificially reduce memory size
RISC debug - afd
References For Further Info
Header files:
/sys/h/
proc.h
user.h
/sys/machine/mips
entrypt.h
frame.h
pcb.h
pte.h
reg.h
Crash(1M)
Sys V "crash" program that is only partially converted to
understand the ULTRIX kernel data structures.
| |||||
| 45.7 | More info on ULTRIX online. | COMICS::TREVENNOR | A child of init | Wed Feb 07 1990 11:10 | 8 |
In:
COMICS::SYS$PUBLIC:ULTRIX*.*
are a number of ULTRIX info files which may be of interest/use.
Alan T.
| |||||
| 45.8 | ULTRIX_CSC notes file. | COMICS::TREVENNOR | A child of init | Wed Feb 07 1990 11:16 | 34 |
If anyone is CSG is supporting ULTRIX and would like to become a member
of the ULTRIX_CSC notes file (charter attached) please let me know.
You asked for it, and here it is. The ULTRIX_CSC
conference is intended as a members-only forum for use by
two groups of people within Digital:
1) People who work with ULTRIX in a technical problem solving
environment - usually in a Customer Support Centre (CSC).
2) People who work in engineering functions associated with
ULTRIX, or those who provide high-level support.
The primary reason for having the conference members-only is
to keep the volume of notes to a usable number, but also to have
a forum where CSC and engineering staff can pool their expertise.
The spin-off benefits are that unannounced information can be
freely discussed, and that you can make contact with people doing
a similar job in different geographies.
CONFERENCE RULES?
1) Be relevant and accurate (If you are guessing, say so).
2) Be as brief as you can.
3) Use KEYWORDS, and make note titles descriptive of the contents
of your note.
4) Do not ask for membership for people with a job description which
does not fit the above profile.
5) If you discover an answer to a problem please post it here.
6) NEVER directly quote or send copies of notes in this conference to
a customer
7) Its okay to say the U*** word 8-)
| |||||
| 45.9 | KERNEL::MOUNTFORD | Wed Feb 07 1990 14:55 | 153 | ||
Moved by mod with permission.
<<< KERNEL::DISK$APD1:[NOTES$LIBRARY]CSGUK_SYSTEMS.NOTE;1 >>>
-< CSGUK_SYSTEMS >-
================================================================================
Note 93.0 ultrix news 1 reply
KERNEL::WIBREW 144 lines 6-FEB-1990 09:48
--------------------------------------------------------------------------------
From: 20050::POEGEL "31-Jan-1990 1425" 31-JAN-1990 21:27:27.26
To: @UONE
CC:
Subj: ULTRIX NEWS SUMMARY
U L T R I X U P D A T E
January 26, 1990
System Software Marketing, Field Programs
Editor: Lynne Poegel XIRTLU::POEGEL ZKO3-3/Y25
*** INTRODUCTION ***
It is our goal to send out the ULTRIX mailing biweekly. The
frequency may change depending on the flow of information. If
you would like to be added to the distribution list, please
send an electronic mail message with your name and node to
XIRTLU::POEGEL. Comments and suggestions for topics are
always welcome.
+-----------+
| Contents: |
+-----------+
PRODUCT INFORMATION
+Phase 0 Opened for DECnet/SNA ULTRIX Programming Toolkit
+Update of Support for TSV05 and the DECsystems 5400
MARKETING INFORMATION
+New RISC/ULTRIX DECwindows Demos Now Available
CUSTOMER EVENTS
+ List of Upcoming UNIX Tradeshows
Digital Company Confidential, For Internal Use Only
UNIX is a registered trademark of American Telegraph & Telephone Company
PRODUCT INFORMATION:
--------------------
PHASE 0 OPENED FOR DECNET/SNA ULTRIX PROGRAMMING TOOLKIT:
Phase 0 has opened for the DECnet/SNA ULTRIX Programming Toolkit.
The purpose of this release is to provide the ULTRIX user a 3270 Data
Stream Programming Interface. Releases could include an LU6.2
programming interface.
If you wish to input into the product requirements, please log into
SSGBPM Infobase through the ACCESS Sales Communication Vehicle or
through the VTX library and look under SSGBPM ULTRIX Liaison Program
for a copy of the Product Requirements Input Form.
---
UPDATE OF SUPPORT FOR TSV05 AND THE DECSYSTEMS 5400:
"Support for the TSV05 on the DECsystem 5400 appears with
V3.1C of ULTRIX. The qualification was not complete in time
for the TSV05 to be incorporated into the Q3 Systems and
Options Catalog and SPD 3.1C, but it will be a supported option."
MARKETING INFORMATION:
----------------------
NEW RISC/ULTRIX DECWINDOWS DEMOS NOW AVAILABLE:
LDP/Science has developed and documented 10 new demonstrations.
Use these demos to showcase the performance of Digital's
RISC/ULTRIX platforms and the DECwindows user interface to
your customers.
These demos illustrate several Cooperative Marketing Program
(CMP) partner applications as well as an integrated DECwrite
solutions.
For a listing of all the available demos, enter the SSGBPM
Infobase through the ACCESS Sales Communication Vehicle or enter
the VTX library and look under SSGBPM ULTRIX Liaison Program.
---
CUSTOMER EVENTS:
----------------
EVENT DATE LOCATION CONTACT
----- ---- -------- -------
UNIXEXPO May 19-11 Los Angeles,CA Judy Izzi
/Spring
DECUS/Spring May 7-11 New Orleans, LA Judy Izzi
Xhibition May 20-25 San Jose, CA Judy Izzi
Usenix June 11-15 Anaheim, CA Judy Izzi
******************************************************************************
* If you know individuals who should receive this newsletter or be deleted *
* from the distribution list, please send mail to XIRTLU::POEGEL with your *
* name, location, and node. *
******************************************************************************
| |||||
| 45.10 | KERNEL::MOUNTFORD | Wed Feb 07 1990 14:55 | 74 | ||
<<< KERNEL::DISK$APD1:[NOTES$LIBRARY]CSGUK_SYSTEMS.NOTE;1 >>>
-< CSGUK_SYSTEMS >-
================================================================================
Note 93.1 ultrix news 1 of 1
COMICS::TREVENNOR "A child of init" 66 lines 7-FEB-1990 11:06
-< Don Zereski (VP CS) on ULTRIX. >-
--------------------------------------------------------------------------------
From: NAME: DON ZERESKI
FUNC: CUSTOMER SERVICES
TEL: 276-9625 <ZERESKI.DON AT A1 at CSSE at OGO>
To: See Below
I recently attended the UNIFORUM Show in Washington, D.C. with
a few other members of CSMC. We then had a long session with
some of our UNIX support people and salespeople. It was truly
an eye opener. We're doing a very poor job convincing our
customers that we are in fact serious about our UNIX strategy.
There are a whole host of items which will soon come out in the
form of some minutes which we'll utilize as a set of individual
items to help guide us during the creation of a UNIX support
strategy.
We will have a single topic meeting to focus on our UNIX
support strategy. Tom Karpowski, along with his ASDS group,
will host the meeting and help establish the process. At the
very minimum we will outline roles and responsibilities and the
action items necessary to solve the existing list of problems,
but more importantly we've got to outline a complete plan for
the long term support of UNIX.
Currently, we have two cultures within Customer Services. One
that's had the last 10-12 years to grow, namely the VAX/VMS
culture and a new culture which is the RISC UNIX culture.
Given the size of the organization and it's time to imbed a
culture, the VAX/VMS culture is well entrenched throughout the
entire organization.
We treat customers who have made a choice to move to UNIX and
RISC as if they were foreigners. We continually try to show
them that they made a very bad decision by forcing them to use
all of our VMS front end tools to access the problem solution
data bases and our CSC's for RISC and UNIX. We continuously
remind them that our choice for support of UNIX will be much
less adequate than our tools and position for VMS. We won't
allow them to submit SPR's via Internet but instead require
paper inputs. Our tools for supporting UNIX and diagnosing
UNIX systems, are well behind the times. We support UNIX, a
kernel in Atlanta, but if they have a UNIX network problem with
TC/PIP we force them to go to Colorado. Why am I mentioning
all of the above? Clearly these are symptomatic of one very
large culture suppressing a new and novel one that more and
more customers are choosing every day, and must be allowed to
be nurtured and grown such that it has the same level of
efficiency and competence as we display with VMS. Clearly,
if we are to be the service vendor of choice our tools and
our performance for UNIX must be equal to those for VMS.
I believe to approach this whole topic as a major cultural
change and perhaps even an organizational change within
Customer Services, we've got to make the necessary investments
and ensure once a customer has made a choice, we provide equal
and adequate support.
We will make our one day single topic meeting meet the vision
that we've laid out for the 90's in terms of our heterogeneous
distributed support.
Regards,
Don
| |||||
| 45.11 | KERNEL::MOUNTFORD | Tue Feb 27 1990 15:17 | 136 | ||
From: COMICS::TREVENNOR "Ultrix Interest Mailing list: 27-Feb-1990 1435" 27-FEB-1990 14:37:13.18 To: @DIST:ULTRIXERS CC: TREVENNOR Subj: UK Ultrix Support Group: Mailshot #19 2 Items in this mailshot. +---+ +-------------------------------------------------------+ | 1 | | UK ULTRIX Support Strategy - latest version available | +---+ +-------------------------------------------------------+ The latest version of the UK ULTRIX Support strategy is available on line on COMICS::SYS$PUBLIC:ULTRIX_STRAT.PS Please pull it over and read through it, and try to have your manager read it too. +---+ +-------------------------------+ | 2 | | Project Athena. One mans view | +---+ +-------------------------------+ [Forwardings deleted....] From Tim Hicks of the North East Sales group. --------------------------------------------- Fellow DWT-ees: For those of you who weren't at the Big East Workstation Training Event in Waterville Valley, NH, I'd like to share with you something that may make you angry, or depressed, or shocked, depending on how much you care about workstation sales for Digital. Suppose I told you that there exists today a software technology for workstations that creates, for workstation users, almost undreamed-of levels of network transparency and ease-of-use of distributed computing resources? This software: - eliminates the needs to know physical locations for anything (ie. no more system:/foo/bar/this.that or VAX::DUA1:[FOO]BAR.TXT) - user's physical files are stored anywhere on the net, but are available transparently to the user REGARDLESS OF WHAT WORKSTATION S/HE LOGS-IN TO. - mail services route mail to the user REGARDLESS OF WHAT WORKSTATION S/HE LOGS-IN TO (no need to know the users location or home system a la: SYSTEM::USER) - Notification of actions and events are routed to the user REGARDLESS OF WHAT WORKSTATION S/HE LOGS-IN TO. - print services for any user are always located on nearby convenient printers, regardless of the users ID and REGARDLESS OF WHAT WORKSTATION S/HE LOGS-IN TO. - all of the above carried out in a COMPLETELY SECURE ENVIRONMENT which utilizes a state-of-the-art user identification/authentication technology. - implements a degree of client-server computing which goes significantly beyond current UNIX-type systems. Do you know any MIS managers that care about user services, client-server computing or network security? Does anyone ever ask you about managing distributed workstation environments, maintaining control and reducing costs? The technology mentioned above PROVIDES SOLID ANSWERS AND POSITIVE BENEFITS that would make any MIS person's heart jump for joy. Now your probably thinking, "yeah, sounds great, but it probably belongs to some third party and isn't ported to DECstations." If so, your wrong! THIS SOFTWARE RUNS ON DECSTATIONS TODAY (with OSF/Motif thrown in to boot)!!!! Its fully documented and has been in production for a year, in a network of 1200+ workstations (growing to 10,000)!!! *** NOW COMES THE SAD PART, FOLKS *** The software described above was built by none other than Digital Equipment Corporation in conjunction with Mass Institute of Technology. Its called Project Athena, and its REAL LIVE SOFTWARE THAT IS FULLY IMPLEMENTED ON RISC/ULTRIX. Maybe you've heard of Project Athena before, and thought that it was academic stuff with no practical purpose. Nothing could be further from the truth! Several of your fiercest workstation competitors are working feverishly to get this product ported to their platforms. But the way things look right now, IBM, Sun and HP will be selling this software on their platforms while you're just finding out about it!!!!! THAT'S RIGHT, FOLKS!!! IBM, Sun and HP will have the competitive advantage that DEC has spent $50MM building!!! Oh, someone in-the-know will tell you that we might be putting PARTS of this technology into ULTRIX in some dim-and-distant future release, in the form of some "toolkits" that the customer can use to build their own applications (does this sound familiar?) In other words, the ingenious customer could roll their own with some pieces from DEC, or get it out-of-the-box from our competition!!!!!! The fact is, in the ultimate expression if Not-Invented-Here syndrome, some of our software architects pooh-pooh this technology because, although it was built by DEC, IT WASN'T BUILT IN THE RIGHT PART OF DEC!!! (Can you say "Spit Brook Road, the fount of all software goodness"?) I don't know about you, but, like most of us in the DWTs, I'm breaking my buns trying to survive in the workstation wars, and I don't know what makes me more angry and depressed: - knowing that Athena software is built and runs on DECstations but Digital will neither market, sell nor support it. - knowing that our competition will have it. - knowing that someone in charge somewhere in DEC made this decision. Well, I've done it now, and I'll get arrogantly-toned hate mail from people and places I've never heard of; however, if this keeps going on its current course, I won't want to live with myself knowing that I could have said something at a time when it might have made a difference. (Pardon the run-on sentence). <<< Tim Hicks >>> | |||||
| 45.12 | KERNEL::MOUNTFORD | Thu Apr 19 1990 15:19 | 1386 | ||
More from Alan....
From: COMICS::TREVENNOR "Ultrix Interest Mailing list: 18-Apr-1990 0915" 18-APR-1990 09:22:54.70
To: @DIST:ULTRIXERS
CC: TREVENNOR
Subj:
UK Ultrix Support Group:
Mailshot #21
4 Items in this mailshot.
1 - DECstation 5000 maintenance manuals available.
2 - What IS a DECstation 5000?
3 - TURBOCHANNEL - introductory details.
4 - ULTRIX is Digital's fastest growing software product - official.
+---+ +---------------------------------------------------------+
| 1 | | DECstation 5000 model 200 maintenance manuals available |
+---+ +---------------------------------------------------------+
In this mailshot we cover a lot about the DECstation 5000.
For all you micro's guys and gals who'd like to get ahead of the game a little
bit we have a LIMITED number of DS5000 maintenance manuals available (10
copies). These can be had by calling the Product & Technology group secretary
Tina Tillbrook on 7833 3919. First come first served - and there are no more
copies available when these are gone!
+---+ +-------------------------------+
| 1 | | So what IS a DECstation 5000? |
+---+ +-------------------------------+
ANNOUNCING THE DECstation 5000 MODEL 200 WORKSTATION AND
THE DECsystem 5000 MODEL 200 SERVER
Mike Savello
Product Manager
RANCHO::SAVELLO
DTN: 543-6588
DIGITAL OFFERS HIGHEST PERFORMANCE RISC/ULTRIX WORKSTATION
============================================================================
| |
| o 60% higher performance (18.5 SPECmarks) than the DECstation 3100 |
| |
| o ECC memory capacity from 8 MB to 120 MB in a slim, desktop package |
| |
| o Range of competitive graphics offerings including color frame buffer,|
| 2D accelerator and high-performance 3D options |
| |
| o Open, high-performance TURBOchannel I/O interconnect supporting a |
| range of Digital-offered options and available via no-charge |
| licensing to option vendors as well as system vendors |
| |
| o Future industry standard VME-bus expansion |
| |
| o Future FDDI interface with industry leadership performance |
| |
| o New 1.0 GByte (1000 MByte) 5.25" external SCSI disk drive available |
| |
| o New 1.2 GByte 4mm Digital Audio Tape (DAT) available for high |
| capacity backup |
| |
| o Entry level configurations start at just $14,995! |
| |
| o Available for immediate delivery |
| |
| o Full binary compatability ensures immediate availability of today's |
| existing portfolio of RISC/Ultrix applications (500+) |
| |
============================================================================
The new DECstation 5000 Model 200 and DECsystem 5000 Model 200 utilize the
MIPS R3000 chip set running at 25 MHz and are the fastest Digital workstations
available as well as the most expandable desktop systems in the industry.
PRODUCT DESCRIPTION
The DECstation/DECsystem 5000 Model 200 is actually a family of systems
comprised of the following:
o DECstation 5000 Model 200CX - 8 plane frame buffer offers
grayscale or color capability
o DECstation 5000 Model 200PX - 8 plane 2D accelerator
o DECstation 5000 Model 200PXG - 8 or 24 plane 3D accelerator
o DECstation 5000 Model 200PXG Turbo - 24 plane high performance 3D
o DECsystem 5000 Model 200 - Server system
Each system includes the following:
o MIPS R3000/R3010 CPU/FPU @ 25 MHz
o 128 KB Cache - 64 KB Instruction/64 KB Data
o 8 MB memory expandable in 8 MB increments to 120 MB
o Integral SCSI - supports up to 7 external devices
o Integral ThinWire Ethernet controller
o Two 25-pin RS-232 asynchronous ports - each with full modem control
DECstation 5000 Model 200CX
- ---------------------------
The DECstation 5000 Model 200CX utilizes an 8 plane color frame buffer which is
comparable to the color versions of the DECstation 3100. This means that there
is no special hardware to accelerate graphics performance and that all
operations are performed by the CPU. However, the fact that the CPU is rated
at 24+ MIPS and that the color frame buffer interfaces to the very high
performance TURBORchannel I/O interconnnect, the achieved graphics performance
of the CX option is very good. It will often exceed not only frame buffer
performance of competitive systems but often their accelerated performance as
well.
CX Option Specifics:
Module size Requires one TURBOchannel slot
Planes 8
Resolution 1024 x 864
Refresh 60 Hz
Supported Monitors 19" Mono (VR262), 16" Color (VR297),
19" Color (VR299)
When used Cost sensitive applications, no need
for 1280 x 1024
Applications 2D CAD, CASE, Tech Pubs, Financial
DECstation 5000 Model 200PX
- ---------------------------
The DECstation 5000 Model 200PX utilizes a custom graphics co-processor called
PixelStamp which accelerates 2D graphics operations such as vectors and
polygons by 200-300% compared to the CX option.
PX Option Specifics:
Module size Requires one TURBOchannel slot
Planes 8 with 8 plane double buffer
Resolution 1280 x 1024
Refresh 66 Hz
Supported Monitors 19" Color (VRT19)
When used Customers who need high performance 2D
or low cost 1280 x 1024
Applications ECAD, MCAD, Tech Pubs, Earth Sciences
DECstation 5000 Model 200PXG
- ----------------------------
The DECstation 5000 Model 200PXG utilizes the same PixelStamp graphics engine
as the PX, plus an Intel i860 microprocessor which operates at 33 MHz. The
i860 acts as a geometry accelerator and performs all of the 3D transformations.
The PXG is available as either an 8 plane configuration with an 8 plane double
buffer or as a 24 plane configuration with a 24 plane double buffer. A
customer can have an 8 plane PXG upgraded to 24 planes in the field by Digital
Field Service. An optional 24-bit Z-buffer can be added to either
configuration for use by applications requiring fast hidden-line and hidden-
surface removal. The 24-bit Z-buffer can be added either in the factory by
Digital manufacturing or in the field by the customer.
PXG Option Specifics:
Module size Requires two TURBOchannel slots
Planes 8 with 8 plane double buffer or
24 with 24 plane double buffer
Field upgradeable from 8 to 24 planes
Resolution 1280 x 1024
Refresh 66 Hz
Supported Monitors 19" Color (VRT19)
When used Customers who need low cost 3D or true
color 3D with Z-buffer plus an open
TURBOchannel slot
Applications MCAD, Earth Sciences, True Color
Imaging
DECstation 5000 Model 200PXG Turbo
- ----------------------------------
The DECstation 5000 Model 200PXG Turbo utilizes two PixelStamp graphics
engines that operate in parallel plus an Intel i860 microprocessor which
operates at 40 MHz. The combination of the additional PixelStamp engine plus
the faster i860 geometry accelerator helps to boost 3D performance by as much
as 100% over the PXG option for certain graphics operations. The PXG Turbo
option comes standard with 24 planes plus a 24 plane double buffer, 24-bit
Z-buffer, and 24 planes of additional offscreen memory which can be used by
the X server for pixmaps. The PXG Turbo option is the only option which
consumes all three DECstation 5000 Model 200 TURBOchannel slots.
PXG Turbo Option Specifics:
Module size Requires three TURBOchannel slots
Planes 24 with 24 plane double buffer, 24-bit
Z-buffer, and 24 plane pixmap memory
Resolution 1280 x 1024
Refresh 66 Hz
Supported Monitors 19" Color (VRT19)
When used Customers who need highest performance
3D
Applications MCAD, Earth Sciences, Molecular Modeling
DECstation 2D COMPARISON CHART
DS2100 DS3100 DS5000 DS5000
200CX 200PX
------- ------- ------- -------
Processor R2000 R2000 R3000 R3000
Clock Rate (MHz) 12.5 16.67 25 25
SPECmarks 8.3 11.3 18.5 18.5
Dhrystone MIPS 11.8 14.9 24.2 24.2
Linpack FP (SP/DP) 2.8/1.2 4.0/1.6 6.4/3.7 6.4/3.7
2D Vec/Sec 60K 80K 130K 300K
Price (19"C,8MB) $11,950 $14,900 $19,000 $21,500
DECstation 5000/200 3D COMPARISON CHART
DS5000 DS5000 DS5000 DS5000
200CX 200PX 200PXG 200PXG Turbo
------- ------- ------- -------
3D Linked Vec/Sec 60K 90K 330K 440K
Ind. Triangles/Sec* 8K 20K 50K 60K
Linked Triangles/Sec* 9K 20K 70K 115K
* Numbers represent Gouraud shaded, Z-buffered triangles/sec except CX and PX
DECsystem 5000 MODEL 200
The DECsystem 5000 Model 200 utilizes the same technology as the DECstation
5000 Model 200 without the graphics options. This allows the DECsystem 5000
Model 200 to be utilized as a compute server, a file server, a peripheral
server, a future gateway to an FDDI backbone, etc. Since the DECsystem 5000
Model 200 does not include graphics, all three TURBOchannel slots are available
for a variety of options from additional SCSI controllers to Thickwire Ethernet
options to VME and FDDI adapters. It is possible that the DECsystem 5000/200
could be configured with 3 additional SCSI options, each supporting 7 devices,
for a system maximum of 28 devices. Since one of these SCSI peripherals is
a 1.0 GByte disk, one can easily see the advantages that a DECsystem 5000
Model 200 could offer.
The DECsystem 5000 Model 200 is an effective competitor against IBM's new
RS6000 POWERserver series, offering comparable functionality to the deskside
POWERserver 520 at a price competitive with the POWERserver 320. See the
DECstation/DECsystem 5000 Model Competitive Update article for more information.
FEATURES/BENEFITS
Feature Benefit
------- -------
o Based on 25 MHz R3000 o 60% higher performance with full
binary compatibility with all Digital
RISC-based systems
o ECC memory up to 120 MB o Highest memory capability of any
desktop system in industry plus
the reliability of ECC
o TURBOchannel I/O Interconnect o Fastest desktop I/O interconnect in
industry fully open to system and
option vendors
o Future VME-bus expansion o Bus of choice for technical markets
Leadership performance
o Future FDDI network interface o Desktop access to 100Mb/sec FDDI
networks - Leadership performance
o New 1.0 GByte hard disk o State-of-the-art storage capacity
at lowest price/MB
o New 1.2 GByte 4mm DAT backup o 4mm is emerging standard for high
capacity tape backups
DECstation/DECsystem 5000 MODEL 200 BUS STRATEGY
The overall strategy for the DECstation and DECsystem product line is to offer
a choice of open I/O interconnects. TURBOchannel is an innovative, open,
high-performance I/O interconnect integral to the DECstation/DECsystem 5000
Model 200 and will also be implemented in future DECstations and DECsystems.
Industry-standard VME bus will be supported on the DECstation/DECsystem 5000
Model 200 as well as future workstations that implement the TURBOchannel.
Futurebus+ will be supported via an adaptor to TURBOchannel in 1992 as the
natural follow-on to VME. SCSI will continue to be supported as an industry
standard, low performance I/O interconnect for disks, tapes, printers, etc.
For more information on TURBOchannel or VME I/O interconnects, please refer
to their associated article in this Sales Update issue.
ORDERING INFORMATION - REFER TO SOC FOR COMPLETE INFORMATION
All DECstation/DECsystem 5000 Model 200 systems can be configured using one
of two system configuration guides - Packaged or Custom. Both methods are
designed to provide you with the best system to fit your needs. Please refer
to the Systems and Options Catalog (SOC) for Custom Systems ordering
information.
Packaged Systems
- ----------------
The Packaged System concept simplifies ordering by offering many different
system designs in ready-to-ship configurations. Packaged systems have faster
delivery times and lower costs than comparable Custom systems.
Depending on which Packaged system you choose, the system will include a
graphics controller, monochrome or color monitor, and 8, 16, or 24 Mbytes of
memory. The server systems (DECsystem 5000/200) do not include a graphics
controller or a monitor. A VT-style terminal should be ordered as a console
device for the server systems.
It is important to note that only field-installable options can be ordered with
Packaged systems, no additions or substitutions can be made to the packaged
systems. If an option is needed, either order a field-installed option and
have it installed after the Packaged system arrives at your site or order a
system with the option off the Custom menu.
Standard on all Packaged systems is:
- U.S. keyboard (120-V systems only, workstations only)
- Mouse (workstations only)
- SCSI controller
- ThinWire Ethernet port with a T-connector and two terminators
- 6-foot power cord (wall socket to system box, 120-V systems only)
- 3-foot convenience power cord (monitor to system box)
- 10-foot video cable (workstations only)
- 10-foot keyboard/mouse cable (workstations only)
- English language user documentation
- ULTRIX Software Licenses
- PHIGS Runtime License is included in DECstation 5000/200PXG and PXG Turbo
To order a Packaged DECstation/DECsystem 5000 Model 200 system, simply order a
system from Step 1. 120-V systems include a U.S. keyboard and all required power
cords. Order software from Step 2 if necessary.
- --------------------------------------------------------------------------------
Step 1 - Packaged Systems
- --------------------------------------------------------------------------------
Diskless With 332-Mbyte disk
- --------------------------------------------------------------------------------
DECstation 5000 Model 200CX 19", Greyscale, 2 free TURBOchannel slots
8 Mbyte, 120 V/240 V/S. Hemi. PM361-BD/BE/BF PM361-MD/ME/MF
- -------------------------------------------------------------------------------
DECstation 5000 Model 200CX 16", Color, 2 free TURBOchannel slots
8 Mbyte, 120 V/240 V/S. Hemi. PM361-BG/BH/BJ PM361-MG/MH/MJ
- -------------------------------------------------------------------------------
DECstation 5000 Model 200CX 19", Color, 2 free TURBOchannel slots
8 Mbyte, 120 V/240 V/S. Hemi. PM361-BK/BL/BM PM361-MK/ML/MM
- -------------------------------------------------------------------------------
DECstation 5000 Model 200PX 19", 2D Accelerator, 2 free TURBOchannel slots
8 Mbyte, 120 V/240 V/S. Hemi. PM362-BK/BL/BM PM362-MK/ML/MM
16 Mbyte, 120 V/240 V/S. Hemi. PM362-DK/DL/DM PM362-PK/PL/PM
- -------------------------------------------------------------------------------
DECstation 5000 Model 200PXG 19", 8-plane 3D, 1 free TURBOchannel slot
16 Mbyte, 120 V/240 V/S. Hemi. PM363-DK/DL/DM PM363-PK/PL/PM
- -------------------------------------------------------------------------------
DECstation 5000 Model 200PXG 19", 24-plane 3D, 1 free TURBOchannel slot
16 Mbyte, 120 V/240 V/S. Hemi. PM364-DK/DL/DM PM364-PK/PL/PM
- -------------------------------------------------------------------------------
DECstation 5000 Model 200PXG TURBO 19", High-performance, 24-plane 3D, no slots
24 Mbyte, 120 V/240 V/S. Hemi. PM365-EK/EL/EM PM365-RK/RL/RM
- -------------------------------------------------------------------------------
DECsystem 5000 Model 200 Server, 3 free TURBOchannel slots
16 Mbyte, 120 V/240 V PM369-PY/PZ
- --------------------------------------------------------------------------------
Step 2 - Software Media and Documentation
An ULTRIX s/w kit must be ordered for the first DECstation/DECsystem 5000/200
on site.
ULTRIX kits include documentation, ULTRIX single-user kit, and DECwindows
software on TK50 or compact disc.
QA-VV1AA-H5 ULTRIX WS (RISC) SW UPD TK50 for DECstation 5000/200
QA-VV1AA-H8 ULTRIX WS (RISC) SW UPD CDROM for DECstation 5000/200
QA-VYVAA-H5 ULTRIX (RISC) SW UPD TK50 for DECsystem 5000/200
QA-VYVAA-H8 ULTRIX (RISC) SW UPD CDROM for DECsystem 5000/200
The following kits serve DECstation 5000 Model 200 networked systems from a
RISC or VAX server.
QA-VV1AB-H5 ULTRIX WS (RISC) SW SERVER (TK50)
QA-VV1AB-H8 ULTRIX WS (RISC) SW SERVER (CDROM)
The following PHIGS kits and licenses are available. Note that
DECstation 5000 Model 200PXG and PXG Turbo 3D systems include DEC PHIGS
Runtime license.
QL-VW7AA-AA DEC PHIGS Runtime License for DECstation/DECsystem 5000/200
QL-VW6AA-AA DEC PHIGS Development License for DECstation/DECsystem 5000/200
QA-VW7AA-H5 DEC PHIGS Runtime s/w (TK50)
QA-VW6AA-H5 DEC PHIGS Development s/w (TK50)
- --------------------------------------------------------------------------------
DECstation/DECsystem 5000 Model 200 Options and Upgrades
- --------------------------------------------------------
In addition to what can be ordered from the Packaged and Custom menus, options
and upgrades can be added to a system unit after it has been installed at the
customer site. They allow the customer to grow the system as needs change over
time. All ordering steps are optional.
- --------------------------------------------------------------------------------
Step 1 - TURBOchannel and Other Options
PMAD-AB Thickwire Ethernet TURBOchannel option card. Requires one
TURBOchannel slot.
PMAZ-AB Additional SCSI TURBOchannel option card. Supports 7 external
devices. Requires one TURBOchannel slot.
Note: - The following PMAG-xB part numbers are graphics module upgrades for
the DECstation 5000/200. If upgrading from a DECstation 5000/200CX
color frame buffer, a new VRT19 monitor must be ordered in addition
to a graphics module. The upgrades are not available at present for
DECsystem 5000/200.
- With the exception of PMAG-GB, all other options are customer
installable/upgradeable.
PMAG-CB DECstation 5000/200 PX upgrade (2D accelerator). Require one
TURBOchannel slot.
PMAG-DB DECstation 5000/200 PXG 8-plane upgrade (low-end 3D). Requires
two TURBOchannel slots.
PMAG-EB DECstation 5000/200 PXG 24-plane upgrade (mid-range 3D).
Requires two TURBOchannel slots.
PMAG-FB DECstation 5000/200 PXG Turbo upgrade (high-end 3D). Requires
three TURBOchannel slots.
PMAG-GB DECstation 5000/200 8-to-24-plane upgrade for PXG. Requires
Field Service installation. Doesn't require a TURBOchannel slot
PMAG-HA 24-bit optional Z-buffer for DECstation 5000/200PXG only. Does
not require a TURBOchannel slot.
- --------------------------------------------------------------------------------
Step 2 - Mass Storage
Note: - 240-V systems require an additional power cord for each external
expansion box.
- xA = 120 V. x3 = 240 V.
- Maximum of 7 external devices can be supported by the base SCSI
controller. If additional SCSI devices are required, a SCSI
TURBOchannel option card must be added (customer installable).
- Each RZ5X, TK50Z, TLZ04, and RRD40 expansion device includes the
following cables:
BC19J-1E 18-inch, 50-pin to 50-pin SCSI cable
RZ5X-CA/C3 332-Mbyte disk drive in a BA42 expansion box.
RZ55-UK 332-Mbyte disk upgrade for BA42
RZ5X-FA/F3 665-Mbyte disk drive in a BA42 expansion box.
RZ56-UK 665-Mbyte disk upgrade for BA42
RZ5X-HA/H3 1.0-Gbyte disk drive in a BA42 expansion box.
RZ57-UK 1.0-Gbyte disk upgrade for BA42
RZ5X-GA/G3 Two 1.0-Gbyte disk drives in a BA42 expansion box.
TLZ04-FA 1.2-Gbyte 4mm digital audio tape (DAT) in external expansion box
TK50Z-GA/G3 95-Mbyte streaming tape in a BA40 expansion box.
RRD40-FA/F3 600-Mbyte compact disc drive in an expansion box.
- --------------------------------------------------------------------------------
Step 3 - Memory
MS02-AA 8-Mbyte, ECC memory module. System supports up to 120 Mbytes
- --------------------------------------------------------------------------------
Step 4 - Optional Input Devices
Note: - The tablet is used in place of the mouse
- The Lighted Programmable Function Keyboard (LPFK) and Programmable
Function Dials (PFD) can be ordered as a pair or seperately. The
LPFK and PFD packages listed below include a Peripheral Control
Module (PCM) which provides multiplexing of both LPFK and PFD into
a single RS232 port. In addition, each package includes a power
supply, cables, and user documentation.
VSXXX-AB 11-inch x 11-inch Tablet with a 2 button stylus and a 4 button
puck
VSX10-AA Combination LPFK and PFD Package - 120 V
VSX10-A3 Combination LPFK and PFD Package - 240 V
VSX20-AA LPFK Package - 120 V
VSX20-A3 LPFK Package - 240 V
VSX30-AA PFD Package - 120 V
VSX30-A3 PFD Package - 240 V
RZ55-BASED PACKAGED SYSTEMS
In addition to the diskless Packaged Systems available, there are also a number
of Packaged Systems that include an RZ55 in an expander box at a $2000 discount
off of sum-of-the-pieces. The reason for offering this package is to be able
to have very attractively priced, standalone configurations available. The
competition can sell standalone systems with 2x100MB disks vs. DECstations
which require an RZ55 to support standalone operation. By selling a standalone
Packaged System which competes directly with competitive systems with two 100MB
disks, are total product offering becomes much more competitive.
PRICING INFORMATION
Packaged Systems, no disks Model Number U.S. MLP
- -------------------------- -------------- -------
DS5000/200CX,19"grs,8MB,no disk PM361-BG/BH/BJ $14,995
DS5000/200CX,16"clr,8MB,no disk PM361-BG/BH/BJ $16,500
DS5000/200CX,19"clr,8MB,no disk PM361-BK/BL/BM $19,000
DS5000/200PX,19"2DA,8MB,no disk PM362-BK/BL/BM $21,500
DS5000/200PX,19"2DA,16MB,no disk PM362-DK/DL/DM $27,100
DS5000/200PXG,19"3D 8pl,16MB,no disk PM363-DK/DL/DM $29,100
DS5000/200PXG,19"3D 24pl,16MB,no disk PM364-DK/DL/DM $36,100
DS5000/200PXG Turbo,19"3DH,24MB,no disk PM365-EK/EL/EM $56,200
Packaged Systems w/bundled RZ55 Model Number U.S. MLP
- ------------------------------- -------------- -------
DS5000/200CX,19"grs,8MB,RZ55/exp PM361-MG/MH/MJ $18,495
DS5000/200CX,16"clr,8MB,RZ55/exp PM361-MG/MH/MJ $20,000
DS5000/200CX,19"clr,8MB,RZ55/exp PM361-MK/ML/MM $22,500
DS5000/200PX,19"2DA,8MB,RZ55/exp PM362-MK/ML/MM $25,000
DS5000/200PX,19"2DA,16MB,RZ55/exp PM362-PK/PL/PM $30,600
DS5000/200PXG,19"3D 8pl,16MB,RZ55/exp PM363-PK/PL/PM $32,600
DS5000/200PXG,19"3D 24pl,16MB,RZ55/exp PM364-PK/PL/PM $39,600
DS5000/200PXG Turbo,19"3DH,24MB,RZ55/exp PM365-RK/RL/RM $59,700
DSYS5000/200 server,16MB,RZ55/exp PM369-PY/PZ $24,095
Ala Carte Systems, no monitor, no disk Model Number U.S. MLP
- -------------------------------------- -------------- -------
DS5000/200CX,color frame buffer,8MB PM371-BY/BZ $14,600
DS5000/200PX,2D accelerator,8MB PM372-BY/BZ $17,100
DS5000/200PXG,8pl 3D,8MB PM373-BY/BZ $19,100
DS5000/200PXG,24pl 3D,8MB PM374-BY/BZ $26,100
DS5000/200PXG Turbo,high perf. 3D,8MB PM375-BY/BZ $40,600
DSYS5000/200 server,8MB PM379-BY/BZ $14,995
Option Decsciption Model Number U.S. MLP
- ------------------ -------------- -------
DS5000/200 8MB ECC memory MS02 -AA $ 5,600
DS5000 Thickwire Ethernet option PMAD -AA/AB $ 400
DS5000 SCSI Controller option PMAZ -AA/AB $ 1,500
DS5000 8Pl 2D Accelerator Upgrade PMAG -CB $ 3,500
DS5000 8Pl 3D Upgrade PMAG -DB $ 5,500
DS5000 24Pl 3D Upgrade PMAG -EB $12,500
DS5000 High performance 3D Upgrade PMAG -FB $25,000
24Pl upgrade for DS5000 8Pl 3D PMAG -GB $ 7,000
24-bit Z-buffer option for DS5000/PXG PMAG -HA/HB $ 3,000
WARRANTY
DECstation and DECsystem 5000 Model 200 customers have four warranty options
to choose from that provide the flexibility to satisfy specific customer
support needs. The recommended warranty level is the STANDARD Warranty
Support option which will provide DECstation and DECsystem 5000 Model 200
customers with full hardware and software support for the duration of the
warranty period, one year. The spectrum of warranty offerings available to
DECstation and DECsystem 5000 Model 200 customers include: Product Foundation
Warranty, Basic Hardware Support, Basic System Support (Standard), and
DECsystem Support 9/5. (Refer to the preface of the USPL for warranty
information.)
All add-on options, including TURBOchannel upgrades and internal memory, for
prepackaged/preconfigured systems are customer installable with the exception
of the PMAG-GB 24 plane upgrade for PXG.
Post-Warranty Services
- ----------------------
For support beyond the first year, Digital offers a comprehensive range of
services that appeals to a diverse customer base--from customers who are
very price-sensitive to customers who want the highest value-added service.
Customers have the same flexibility as during the warranty with additional
options for specialized service programs to meet their business needs. For
more information on any of the warranty offerings or continuation services,
contact your local Field Service Sales support office.
AVAILABILITY
o Start placing orders April 3, 1990
o DECstation 5000 Model 200CX ready April 6, 1990
for shipment w/UWS V2.1d
o DECsystem 5000 Model 200 ready for April 6, 1990
shipment w/Ultrix V3.1d
o DECstation 5000 Model 200PX ready for May, 1990
shipment w/UWS V4.0
o DECstation 5000 Model 200PXG and August, 1990
DECstation 5000 Model 200PXG Turbo ready
for shipment w/UWS V4.0
o RZ57 and TLZ04 devices See associated
articles this issue
o VME and FDDI adaptors See associated
articles this issue
o All other options listed in Ordering April 6, 1990
Information above
Graphics Upgrade Availability
- -----------------------------
The DECstation 5000 Model 200CX will be available for customer delivery at
announcement and the DECstation 5000 Model 200PX will be available within
approximately 30 days of announcement. The PXG and PXG Turbo 3D configurations
will be available in August, 1990. For customers who want to purchase PXG
or PXG Turbo systems, they can start with a PX and then upgrade when the 3D
options become available by purchasing an add-on module. This way they get
to use the same VRT19 monitor that they purchased with their PX system.
And since PHIGS is the same on the 2D CX and PX as it is on the 3D PXG and
PXG Turbo, 3D applications developed on the CX and PX will migrate easily to
the PXG and PXG Turbo.
COMPETITIVE SUMMARY
DESKTOP: ENTRY-LEVEL 2D, 19" COLOR, 8MB, DISKLESS
=================================================
DS5000 DS3100 DS2100 SUN H-P IBM
M200 SPCst1 834CH RS6000*
M320
Processor R3000 R2000 R2000 SPARC HP-PA IBM
Clock Rate (MHz) 25 16.67 12.5 20 15 20
SPECmarks 18.5 11.3 8.3 8.3 9.5 22
Dhrystone MIPS 24 14.9 10.8 13 14 27
MFLOPS (Linpack-DP) 3.7 1.6 1.2 1.4 2.0 7.4
Price $19K $14.9K $12K $12.5K $22K $16.3K
DESKTOP: ACCELERATED 2D/ENTRY 3D, 19" COLOR, 8MB, DISKLESS
==========================================================
DS5000/200 VS3520/ SGI SUN H-P IBM
PX and PXG 3540 4D/25 SPCst1GX 825CHX RS6000*
825SRX M320
Processor R3000 VAX R3000 SPARC HP-PA IBM
Clock Rate (MHz) 25 31 20 20 12.5 20
VUPS 20 5-10 11 10 22
Dhrystone MIPS 24 16 13 8 27
MFLOPS (Linpack-DP) 3.7 0.4 1.4 1.4 0.7 7.4
2D Vec/sec 300K 100K 85K/200K 400K 70K 90K
3D Vec/sec 330K 100K 85K/200K 175K 70K(SRX) 90K
3D Triangle/sec 70K 7-10K 5K/20K n/a 8K 10K
Price PX: $21.5K $29.5K $18.5K/ $15K $42.5K(CHX)
PXG:$23.5K $25.5K $56.5K(SRX) $18.1K
Basic/Turbo
HIGH-END 3D (DOUBLE BUFFER, Z-BUFFER, FASTEST GRAPHICS),19" COLOR, 8MB,DISKLESS
===============================================================================
DS5000/200 VS3520/ SGI SUN H-P IBM
PXG Turbo 3540 4D/50GTB SPCstn 835SRX RS6000*
330GXP Turbo M730
Processor R3000 VAX R2000 SPARC HP-PA IBM
Clock Rate (MHz) 25 31 8 25 15 25
VUPS 20 5-10 7 14 30(est)
Dhrystone MIPS 24 16 14 34
MFLOPS (Linpack-DP) 3.7 0.4 0.7 2.6 2.0 10.9
2D Vec/sec 440K 100K 300K 240K 990K
3D Vec/sec 440K 100K 300K 90K 240K 990K
3D Triangle/sec 90K 7-10K 25K 5.5K 38K 120K
Price $45K $40.5K $60K $35K $63.5K $75K+
SERVER, 16MB, 300MB DISK
========================
DSYS5000/200 MIPS SUN H-P IBM
RC3240 SPCsrv 815S RS6000*
330 M320
Processor R3000 R3000 SPARC HP-PA IBM
Clock Rate (MHz) 25 ? 25 10 25
Cache Size 128KB ? 128KB 16KB 40KB
Memory min/max 8/120MB ? 8/40MB 8/96MB 8/32MB
VUPS 20 ? 14 22
Dhrystone MIPS 24 18 16 7 27
MFLOPS (Linpack-DP) 3.7 ? 2.6 0.6 7.4
VME slots 9/opt ?/std 5/std 7/std none
Price $24.1K ~$40K $35.9K $38.8K $20.4K
* All IBM diskless configurations include a 120MB page/swap disk due to the
fact that AIX on IBM does not support diskless operation
QUESTIONS AND ANSWERS
CAN A CUSTOMER UPGRADE FROM A DECstation 2100 OR DECstation 3100 TO A
DECstation 5000 MODEL 200?
At this time, a customer with a DECstation 2100 or DECstation 3100 can re-use
and external disks, tapes, or other external SCSI peripherals that he has
attached to his existing system. We are presently exploring the possibility
of offering a DECstation 5000 Model 200CX without a monitor so that DECstation
2100 and DECstation 3100 customers could re-use the monitor, keyboard, and
mouse as well. If you have a customer interested in such an upgrade, please
contact Worksystems Product Management.
WILL 4Mbit DRAMS BE AVAILABLE IN THE FUTURE
It is our goal to offer 4Mbit DRAMS in the future. This would boost memory
capacity to 480 MB. It may not be possible to mix-and-match 8MB and 32MB
modules in the future. Look for this capability some time in late 1991 or
early 1992.
CAN A CUSTOMER CONFIGURE A SERVER OR WORKSTATION WITH MULTIPLE SCSI OR
ETHERNET OPTIONS?
Yes, it is conceivable that a DECsystem 5000 Model 200 could support up to
4 total SCSI controllers or 4 total Ethernet controllers.
WHEN SHOULD I RECOMMEND A 2D ACCELERATOR (PX) OVER A COLOR FRAME BUFFER (CX)
We expect the 2D accelerator (PX) option to be very popular - especially in
the ECAD marketplace. However, the graphics performance of the 8 plane frame
buffer is 60% faster than the DECstation 3100 and offers very competitive
graphics performance adequate for many applications. Customers requiring very
fast 2D vector performance (300K) should move to the PX option.
WON'T VME PERFORMANCE BE SLOW DUE TO THE FACT THAT IT'S AN OPTIONAL ADAPTOR?
Due to the fact that our VME adaptor interfaces to our very high performance
TURBOchannel I/O interconnect, we expect to achieve excellent throughput to
VME options. In fact, the DECstation/DECsystem 5000 Model 200 may end up
having the fastest VME interface available in the industry. See related
Sales Update article in this issue.
WILL DIGITAL BE OFFERING OTHER TURBOCHANNEL OPTIONS IN THE FUTURE?
It is our intention to offer whatever is necessary to make our product offering
as robust as possible. There are many options still required. We will be
working with third parties to satisfy some needs, and where necessary will be
developing our own solutions. If you have a customer with a specific option
need not satisfied today, contake Worksystems Product Management.
ISN'T IT A PROBLEM THAT THE HIGH PERFORMANCE PXG TURBO SYSTEM DOESN'T HAVE ANY
OPEN TURBOchannel SLOTS?
The DECstation 5000 Model 200PXG Turbo will only be sold to those few select
customers that require the absolute best 3D performance. In most cases, that
means that they have a very specialized application and don't require optional
expandability. For those customers who do require expandability, we can offer
the DECstation 5000 Model 200PXG in a 24 plane configuration with an optional
24-bit Z-buffer and still have a TURBOchannel slot open for expansion.
+---+ +------------------------------------------+
| 3 | | Turbochannel Q&A - introductory details. |
+---+ +------------------------------------------+
1. GENERAL TURBOchannel QUESTIONS:
Q: What is TURBOchannel and what are you trying to do with it?
A: TURBOchannel is Digital's new, OPEN, high performance I/O
interconnect integral to the DECstation 5000 Model 200, DECsystem
5000 Model 200, and future DECstations. It is "open to the
industry", including option, system, and chip vendors. This means
that the TURBOchannel design is available for license to anyone;
the license is no-charge/royalty-free.
TURBOchannel is differentiated from other buses by:
a. industry-leading, proven high performance: 100 MB/sec peak
transfer rate, with 93 MB/sec achieved on the DECstation
5000/200.
b. designed for simplicity, longevity and scalability
c. completely open -- including option, system, and chip
vendors; no-charge/royalty-free license
Digital is working with leading semiconductor manufacturers to
make interface ASICs and ASIC design services available to
independent option vendors that are developing TURBOchannel
options.
The TRI/ADD Program, a new program residing in Palo Alto with the
RISC Workstations Engineering group, provides: recruiting,
technical support, and marketing support for all independent
hardware vendors (TURBOchannel, VME, and SCSI). Vendors can call
the TRI/ADD Program at 1-800-678-OPEN immediately [as of April
3rd announcement] to receive an information package that contains:
a technical data sheet with enough information to determine if
they might be interested in designing a TURBOchannel product; a
TURBOchannel overview; information about the free technical and
marketing support available; and a combined TURBOchannel license
and TRI/ADD Program registration form.
Digital is making available a comprehensive technical documentation
kit which includes the TURBOchannel hardware specification, mechani-
cal drawings, information on writing a device driver, and system
parameters. This TURBOchannel Developers Kit is targetted at both
option and system vendors and will be available in June 1990 for a
$65 charge (essentially covers printing costs). Additional docu-
mentation will be phased in over time: option and system firmware
specifications, option and systems designers guides, and system
mechanical drawings.
Q: Why is Digital offering another "proprietary" bus? [trick Q]
A: Digital developed TURBOchannel but is sharing the technology with
the industry on a no-charge, no-royalties basis. There are no
restrictions on who can use the TURBOchannel design to develop
TURBOchannel products. There is strong interest among the option
and system vendors we have talked with for the high-performance
and concise design of TURBOchannel, with the fact that TURBO-
channel is available today and has been designed for the long-term,
and the fact that there are no charges, no royalties, and no
restrictions on who can use the TURBOchannel design. End users
also benefit from TURBOchannel's low cost since it is integral to
the system and no external expansion box is necessary.
TURBOchannel was chosen as the integral I/O interconnect for the
long-term because it has much higher performance than existing
open buses:
proven performance
TURBOchannel 93 MB/sec
VME ~37 MB/sec
SBus ~27 MB/sec
MCA ~13 MB/sec
TURBOchannel brings industry-leading I/O performance to the
desktop TODAY, with an architecture that allows options developed
today to run on future DECstations and other vendors' systems that
incorporate TURBOchannel.
Q: Is TURBOchannel available in all of Digital's RISC workstations?
What is Digital's long-term commitment to TURBOchannel?
A: TURBOchannel is currently available only in the DECstation 5000/200
workstation and DECsystem 5000/200 server, however it is planned
for at least the next several generations of DECstations. TURBO-
channel will not be retrofitted on the DECstation 2100 or 3100
platforms. Digital has protected customers' investments in these
first members of the DECstation workstation family by ensuring
that applications developed on the DECstation 2100 and 3100 run
on subsequent family members; monitors, keyboards, SCSI peripherals,
and mice purchased for the 2100 and 3100 platforms can be used
with the DECstation 5000/200.
Q: Will TURBOchannel also be available in VAX workstations?
A: The TURBOchannel design is under review for inclusion in future
VAX workstations and other vendors' systems, although there are
no commitments to report. The TURBOchannel architecture is not
limited by any specific processor architecture.
Q: Will Digital be offering a Qbus adapter for TURBOchannel?
A: Digital has no current plans to develop a Qbus adapter. However,
TURBOchannel's design is available to the industry so nothing
prevents another vendor from developing such an adapter.
Q: What options are being developed to connect to TURBOchannel?
A: Digital-developed TURBOchannel options available on DECstation
5000/200 and number of TURBOchannel slots used:
- 200/CX: 1024x864, 8-plane frame buffer, operating at
60 Hz [1 slot]
- 200/PX: accelerated 2D graphics option supporting 1280x1024,
8 planes of color, full-page double-buffering, at
66 Hz; with PixelStamp rendering chipset [1 slot]
- 200/PXG: accelerated 3D graphics option supporting 1280x1024,
up to 24 planes of color, full-page double-buffering,
at 66 Hz; PixelStamp; Intel i860 for geometric
transformations; optional 24-bit Z buffer [2 slots]
- 200/PXG Turbo: advanced 3D graphics option adds to the PXG
features an additional rendering processor, an addi-
tional 24 planes image memory, faster i860 [3 slots]
- additional SCSI controller [1 slot]
- DEC LANcontroller 700: additional Ethernet controller (Thickwire)
[1 slot]
- VMEbus adapter: program announced to be available in 1990
[1 slot]
- DEC FDDIcontroller 700: FDDI adapter; program announced to be
available in the future [1 slot]
Third parties will provide the majority of TURBOchannel options.
As part of the April 3rd announcement, the following companies
announce their intent to develop TURBOchannel options:
- CSPI:
SuperCard/TURBO(tm) vector processor
- Data Translation:
50 MHz data acquisition board
color image processing boards for real-time image analysis
- RasterOps Corporation:
high resolution 24-plane true-color frame buffer
- Sky Computers, Inc.:
SKYbolt(tm) application accelerator
2. OPENNESS AND FIT WITH DIGITAL'S OPEN BUS STRATEGY
Q: How open is TURBOchannel?
A: TURBOchannel is "open to the industry." This means that anyone
can use the TURBOchannel design that's specified in the TURBO-
channel Hardware Specification to design a TURBOchannel option,
develop a system that includes TURBOchannel, or design a TURBO-
channel interface chip. The TURBOchannel design is licensable
to anyone on a no-charge, royalty-free basis. End users benefit
from its openness by the availability of options and systems
developed by both Digital and other vendors. In addition,
TURBOchannel is the mechanism for providing access to industry-
standard VME and Futurebus+ on DECstations
Among the free services that the new TRI/ADD Program offers in
support of TURBOchannel product design efforts are: hardware and
software technical support, marketing support, an equipment
program and an online communications mechanism for vendors.
Comprehensive technical documentation is available for a $65
charge which essentially covers our printing costs. The kit
includes: the TURBOchannel hardware specification, option
mechanical drawings, information on writing a device driver, and
system parameters. Other documents will be phased in over time:
option and system firmware specifications, option and system
developers guides, system mechanical drawings. Digital has put
a lot of emphasis on making TURBOchannel, and TURBOchannel tech-
nical expertise, truly open to the industry.
Q: Do I need a license to design to TURBOchannel?
A: Yes. Digital owns the technology, but anyone can be licensed
to use the TURBOchannel design to develop TURBOchannel options,
systems, and chips. The license itself is very concise (1/2 page)
and is offered on a no-charge, royalty-free basis. TURBOchannel
is open to the industry.
Q: Do I have to buy any parts from Digital to design to TURBOchannel?
Where and how fast can I get a copy of the spec?
A: Anyone can receive a free TURBOchannel technical data sheet
by calling the TRI/ADD Program at 1-800-678-OPEN. It provides
enough of the essential information about TURBOchannel to decide
if it is something you want to design to: protocol description,
timing diagrams, option mechanical drawing, performance data, and
an overview ROM description. It is part of a free information
package that also includes a description of the no-charge tech-
nical and marketing support offered by the TRI/ADD Program, a
TURBOchannel Overview, and a combined TURBOchannel license and
TRI/ADD Program registration form. The TURBOchannel Developers
Kit contains the spec and other technical information to help an
option, system or chip vendor to design to TURBOchannel. This
kit is available in June for $65 which essentially covers our
printing costs. We are developing a fast-order system for the
Developers Kit.
In addition to offering technical documentation, Digital is
working with semiconductor manufacturers to supply interface
ASICs and ASIC design services to assist vendors in their
development efforts.
Q: How does the announcement of TURBOchannel relate to Digital's
recent open bus announcement [February 14; BUSCON/West] of VMEbus
and Futurebus+ support? Why are you providing TURBOchannel when
you've just announced a commitment to VMEbus and Futurebus+?
A: TURBOchannel is a critical part of Digital's recently-announced
open bus strategy. Beginning with the DECstation 5000 Model 200,
Digital provides a CHOICE of I/O interconnects so that customers
have flexibility and openness:
a. TURBOchannel, an OPEN, high-performance I/O channel for appli-
cations such as imaging and graphics that require the maximum
possible bandwidth (93 MB/sec peak DMA proven on DECstation
5000/200); with TURBOchannel, Digital offers industry-leading
I/O performance TODAY.
b. industry-standard, general purpose buses:
- VME for applications that require an industry-standard
solution and access to existing VME applications. (Peak
VME architectural performance is 40 MB/sec.). A VME
adapter connects directly into TURBOchannel; an external
expansion enclosure is required.
- Futurebus+ is being defined by an IEEE standards committee
as the natural follow-on to VME. Estimates are that it
will be available in the 1992 timeframe in a *deskside*
form factor. Futurebus+ will be supported by future
DECstations through a TURBOchannel adapter.
A number of option vendors have announced [April 3rd] their intent
to develop TURBOchannel boards for applications that require
maximum possible bandwidth today. Some of these option vendors
also offer VME boards for applications requiring industry-
standard solutions. Other option vendors have announced [April
3rd] VME boards. TURBOchannel, VME and Futurebus+ are
complementary buses.
Q: When should a developer use a non-standard bus like TURBOchannel
rather than an industry standard bus like VMEbus or Futurebus+?
A: TURBOchannel is the best solution for high-performance, low-
cost, desktop solutions today. TURBOchannel offers industry-
leading performance: TURBOchannel's peak architectural DMA is
100 MB/sec while VME's peak DMA is 40 MB/sec. With the DECstation
workstation family, TURBOchannel brings this high performance to
the desktop today. No external expansion boxes are required
(as with VME) which results in lower-cost solutions. Futurebus+
is still under definition by an IEEE standards committee and
applications are not expected to be available for another several
years. Futurebus+ is also being defined as a deskside implementa-
tion. TURBOchannel is available today in a desktop form factor
(integral to the DECstation 5000/200) with the flexibility to move
unchanged into the future. Board vendors and system vendors with
whom we've shared information about TURBOchannel are very excited
about it and how open we've made it.
Q: What standards bodies are evaluating TURBOchannel to endorse it
as a standard?
A: Digital developed TURBOchannel and is choosing to share this
technology, without restrictions, with the industry to make its
performance and cost benefits widely available. No formal
standards bodies are evaluating TURBOchannel to endorse it.
Standards can emerge either from a standards body, such as IEEE,
or from market demand. Digital believes TURBOchannel will
succeed based on market demand, and efforts are underway to
maximize this demand.
Q: If Futurebus+ was selected by Digital to provide boundless
throughput capability for the long-term, doesn't this mean that
TURBOchannel has a very limited life? Why bother with TURBO-
channel?
A: The Futurebus+ protocol currently under definition by its IEEE
standards committee defines a deskside ("profile B") general
purpose bus architecture, with a peak DMA transfer rate expected
to be in the range of 200MB/sec - 1.7GB/sec in 1992 (for 32-bit -
128-bit packet systems respectively), across the entire bus.
Since Futurebus+ is being defined by a standards committee, we
expect it will become a standard quickly. Digital supports
Futurebus+ as a standard and a natural follow-on to VME.
TURBOchannel brings industry-leading high performance to the
desktop today (100MB/sec peak DMA, 93MB/sec proven on DECstation
5000/200). Because it is defined as an asymetrical I/O archi-
tecture rather than a general purpose bus architecture, the
CPU and system memory are defined separately from the TURBO-
channel architecture. This allows the same TURBOchannel
architecture available today to be used in future higher-
performance systems where the CPU-to-memory interconnect is
optimized and where each option can have access to the entire
TURBOchannel bandwidth. Such a hypothetical future system can
have 800 MB/sec effective I/O performance, and more, without
changing TURBOchannel at all. Also, as proven with other buses
in the marketplace, market life is longer than technological life.
TURBOchannel is an optimized solution -- optimized around
performance and simplicity -- for today and the long-term.
Q: How does TURBOchannel compare to VAXBI and Qbus?
A: TURBOchannel is OPEN and Digital is licensing it to the industry
on a no-charge, royalty-free, no-restriction basis. This degree
of openness, and the lack of licensing restrictions, contrasts
with VAXBI. TURBOchannel's ability to bring new technologies to
the desktop and to allow current solutions to be supported in the
future compares favorably to Qbus. TURBOchannel's openness, high
performance, and design simplicity -- coupled with excitement
from the vendors we have shared the TURBOchannel design with to
date -- positions it as a solution for the long-term. Ultimately,
the industry will determine how "standard" TURBOchannel becomes
and how favorably it compares with the longevity of Qbus.
3. COMMON TECHNICAL QUESTIONS
Q: How many open slots are available on the DECstation 5000/200?
Will this be changed in future platforms?
A: There are 3 TURBOchannel slots on the DECstation 5000 Model 200
for add-in boards. Boards can be double-, or triple-width to
accommodate sophisticated circuitry that requires more real
estate than is provided by a single-width option. Double- and
triple-width options also guarantee electrical power to drive
complex integrated circuits requiring more than 26 watts per
slot. In addition, the TURBOchannel mechanical specification
provides a sufficient spatial envelope to allow for double-side
surface-mount boards. The underside of a board can accommodate
passive components and low-profile active components. Since the
TURBOchannel option and socket sizes will not change in future
DECstations, the number of TURBOchannel slots in future DEC-
stations is constrained by internal box dimensions. Other
system vendors using TURBOchannel as the integral I/O inter-
connect may choose to offer different numbers of slots.
Q: IBM offers 4 MCA slots on their RS/6000 Model 320 yet we're
saying that the DECstation 5000 Model 200 is more flexible.
Please explain.
A: Of the 4 MCA slots, 2 are taken up by Ethernet and SCSI boards.
The DECstation 5000/200 has integral Ethernet and SCSI. After
considering these 2 options, there are 2 open MCA slots and
3 open TURBOchannel slots. We also offer industry standard
VME expansion through an adapter into TURBOchannel; IBM does
not offer this.
Q: Is TURBOchannel's 100 MB/sec peak performance shared across all
options or does each option get 100 MB/sec?
A: In the DECstation 5000/200 platform, there is a single TURBOchannel
across all options. Future platform architectures can assign
each option its own dedicated TURBOchannel so that each option
can have access to the entire 100 MB/sec bandwidth. Also, since
TURBOchannel is defined separately from the system memory-to-CPU
interconnect, future systems can increase total system performance
by optimizing that connection without changing TURBOchannel.
Q: Will you be evolving the design of TURBOchannel? Will TURBOchannel
boards or systems that I design or buy today be obsoleted by a
new and improved Digital design?
A: Digital's commitment to TURBOchannel is strong; TURBOchannel was
designed for longevity and scalability. The address space upper
limit today is 16GB while competing buses are in the 256MB (SBus)
- 4GB (MCA, NuBus, VME, EISA) range. Minimums stated in the
TURBOchannel specification are committed minimums, such as power
(26 watts/slot) and air flow (150 LFM). Physical features will
remain unchanged: board dimensions (4.6" x 5.675"), connectors
(96-pin DIN), clock speed (12.5 MHz - 25 MHz), etc. Architectural
features will remain unchanged: 4MB-512MB address space per slot,
1- to 128-word DMA bursts, etc. Digital is committed to protecting
investments by end users, board vendors (including Digital!),
systems vendors, and chip vendors.
Q: Can you get more expansion slots by connecting an expansion box
to TURBOchannel?
A: There is no architectural reason why an adapter to a TURBOchannel
expansion box could not be implemented since the address space
is defined as 4MB-512MB. Such an adapter would share the slot
address space so this is not possible with the DS5000/200's 4MB
slot address space.
Q: Would a customer have problems interfacing non-SCSI storage to
TURBOchannel? Is booting from such a storage device supported?
A: A customer would *not* have problems interfacing non-SCSI
storage to TURBOchannel, and booting from such storage devices
*is* supported.
Q: Would non-SCSI storage offer increased performance?
A: Because TURBOchannel is open to the industry, vendors can
develop any interfaces to it. With IPI or ESDI, for example,
the performance inherent in such interfaces would be realized.
Q: Does TURBOchannel enhance my SCSI disk performance?
A: No. Disk performance is limited by the throughput of SCSI on
the DS5000/200, however faster disk controllers could be
developed for TURBOchannel.
Q: What will be the throughput of a VMEbus option connecting through
the VME adapter to TURBOchannel? reads and writes?
A: Throughput is limited by the slowest transfer rate. VME's peak
architectural transfer rate is 40 MB/sec whereas TURBOchannel's
peak transfer rate is 100 MB/sec. In this case, VME is slower
than TURBOchannel so VME options would never have faster throughput
(reads or writes) than VME. Support for Digital's VME adapter
to TURBOchannel is a program announcement on April 3rd. Per-
formance characterizations of this VME option is not available
yet.
Q: How difficult is TURBOchannel to design to?
A: TURBOchannel is designed to be concise. The concise protocol
defines only 2 types of transactions: I/O reads and writes,
and DMA reads and writes. TURBOchannel has 44 signal pins,
compared with other buses: SBus, 82; VME, 106; MCA, 136; EISA,
153; Futurebus+, 91-343. A TURBOchannel Developers Kit is
available and provides a concise specification (under 20 pages),
and additional hardware and software technical documentation.
Slave options are easiest to design. Initial feedback from
board vendors that have seen the TURBOchannel specification
is that porting a board from SBus would be easiest (one estimate
was 2 weeks to port either a synchronous/asynchronous serial
line or a parallel port). Imaging boards require synchronizing
signals with the X Server so the effort is greater and is
weighted towards software. The TRI/ADD Program provides free
technical support (hardware and software) to vendors to help
optimize their design efforts. In terms of interface costs,
estimated costs of interface silicon are: ~$20 for TURBOchannel,
~$25 for NuBus; ~$40 for SBus; ~$50 for VME; ~$200 for Futurebus+.
Digital is working with leading chip vendors to make interface
ASICs and ASIC design services available to option vendors
developing TURBOchannel options.
Q: What software is necessary for a TURBOchannel option? Who
provides this software?
A: Option vendors will provide their own device driver software.
For applications that require synchronizing with X server,
option vendors will provide additional software.
Q: If a Futurebus+ adapter connects directly to TURBOchannel on
DECstations, how do I get full performance from Futurebus+?
A: The Futurebus+ architectural peak performance is defined as
200 MB/sec for its 32-bit implementation, whereas TURBOchannel
(defined as a 32-bit architecture) has a peak architectural
performance of 100 MB/sec. Futurebus+ options connected to a
single, dedicated TURBOchannel can have access to the entire
100 MB/sec bandwidth of that TURBOchannel. Because Futurebus+
is being defined as a general purpose bus, all options connected
to a single Futurebus+ will share its bandwidth.
4. COMPETITIVE POSITIONING
Q: How does TURBOchannel compare with Sun's SBus?
A: SBus has a peak DMA equivalent to TURBOchannel (100 MB/sec)
but much lower proven performance (27 MB/sec vs. 93 MB/sec).
SBus boards are slightly smaller than TURBOchannel boards
(123 square cm vs. 168 square cm). Athough not a constraint of
SBus itself, SPARCstation enclosures have much less head-room
in them than the DS500/200. Some options, therefore, must take
up more slots on SPARCstations than DECstations. TURBOchannel
options in the DECstation 5000/200 can be double surface-mounted
and can accommodate daughter cards, which minimizes the number
of slots required by a particular option.
Both SBus and TURBOchannel are open buses and the designs are
licensable to anyone. Sun charges $300 for their technical
documentation kit; Digital charges $65 to cover our costs, and
includes additional documents such as 1:1 mechanical drawings
and system parameters.
Technical support programs are offered by Sun (Catalyst) and
Digital (TRI/ADD Program). Feedback from board vendors is that
very little, and poor quality, support has actually been provided.
These same vendors had a much higher expectation of support from
Digital and were very pleased with our description of the range
of TRI/ADD Program services and Digital's commitment to support.
Q: How does TURBOchannel compare with IBM's MCA (Micro Channel
Architecture?
A: IBM's "new" MCA, announced on February 15th, has a peak archi-
tectural DMA of 40 MB/sec, with a proven DMA of ~13 MB/sec
(confirmed by MCA board vendors). This is much less than
TURBOchannel's peak architectural DMA of 100 MB/sec and achieved
DMA of 93 MB/sec.
The RS/6000 Model 320 (desktop workstation) has fewer available
MCA slots than the DECstation 5000/200, although the RS/6000/320
actually has 4 MCA slots compared with 3 TURBOchannel slots on
the DECstation 5000/200. Since the DS5000/200 includes integral
Ethernet and SCSI, and these options on the RS/6000 Model 320
consume 1 slot apiece, the DS5000/200 effectively has 3 available
slots to the RS/6000/320's 2. The RS/6000/320 is IBM's only
desktop system in their RS/6000 family of RISC machines.
MCA's power per slot is 12.6 watts, compared with 26 watts per
slot for TURBOchannel.
MCA is licensable to system vendors, however, the licensing
policies for system vendors are very restrictive and expensive:
a system vendor must pay very high royalties on future MCA-system
sales as well as a tax on past systems that used the AT-bus.
TURBOchannel is open to system vendors on a no-charge/royalty-free
basis.
5. SUPPORT FOR TURBOchannel
Q: What do I have to do to get technical support for TURBOchannel?
A: Call the TRI/ADD Program at 1-800-678-OPEN. This program was
developed to provide technical and marketing support for
TURBOchannel option, system, and chip vendors. It is open
to anyone and is staffed to provide comprehensive hardware
and software technical support, including online communication
with other vendors involved in designing TURBOchannel products.
To help us help you, you just need to fill out the short
program registration form (1/2-page) included with the TRI/ADD
Program information package we sent you. In addition to
TURBOchannel support, the TRI/ADD Program provides technical
support for VME and SCSI option development.
Q: What documentation does Digital provide to option and system
designers?
A: A free TURBOchannel technical data sheet is available immediately
after announcement by calling 1-800-678-OPEN. It includes
excerpts from the TURBOchannel Hardware Specification to
assist in an initial assessment of TURBOchannel. It is
provided as part of the TRI/ADD Program's information package
which also includes: TURBOchannel Overview, description of
the TRI/ADD Program, and a combined TURBOchannel license and
TRI/ADD Program registration form.
The TURBOchannel Developers Kit is one kit targetted at both
option and system designers, with both hardware and software
technical information. The initial kit, available in June
for $65, includes:
- hardware specification (under 20 pages)
- system parameters
- information on writing TURBOchannel device drivers
- option mechanical drawings (7 C-sized drawings)
Additional documentation will be phased into the kit over time:
option and system firmware specifications, option and system
designers' guides, system mechanical drawings. As a point of
reference, Sun charges $300 for their developers kit which does
not include actual-size mechanical drawings and is targetted
just at option vendors. No technical documentation was widely
available from Sun until 5 months after announcement of an open
SBus in April 1989.
Q: What other support does Digital provide?
A: The TRI/ADD Program also provides marketing support, an equip-
ment program, an online technical conference for vendors, and
assistance in investigating a possible business relationship with
Digital whereby Digital could service a vendor's TURBOchannel
products.
+---+ +------------------------------------------------------+
| 4 | | Ultrix is Digital's fastest growing software product |
+---+ +------------------------------------------------------+
ULTRIX FINANCE OVERVIEW
APRIL 1990
* The ULTRIX S/W business will be a $100 million dollar business
by the end of FY'90. This is close to a doubling of FY'89's
NOR of $54 million and includes revenue from license sales,
updates, support, service, and education. If the UNIX market is
growing at 30-40% we're gaining; not losing market share.
* ULTRIX is the fastest growing of the top 20 DEC S/W products.
* Through Feb. of FY'90, ULTRIX License MLP is 114% ahead of
Feb. FY'89 YTD. This is 15% ahead of the OSG FY'90 plan and
10% below the corporate plan (which was driven by the PBU
hardware volumes rather optimistic RISC projections.)
* VAX based Ultrix license MLP is running 100% above the OSG plan while
RISC based is running 30% below the OSG plan and 50% below the
the corporate plan. The RISC business has been a disappointment
while the strength of the VAX business has been a surprise. The
Vax based Unix business has been relatively stable the past 6
quarters and the RISC business is gradually improving.
* The transition from VAX based license sales to RISC based has
begun. In Q1'FY90 30% of the business was RISC and 70% was VAX.
In Q2'FY90 the ratio had shifted to 50/50. By Q4'FY90 we think
the actuals will be 70% RISC based.
* ULTRIX based units as a percent of total DEC unit sales have
increased from a 9% share one year ago to a 14% share as of Q2'FY90.
The penetration of ULTRIX on total worldwide workstation license
units is currently 20%. In the USA, this penetration is 26% a
full 10% points above the foreign penetration of 16%.
* The investment in OSG engineering continues to grow. FY'90 Spending
is up approximately 25% over last years level, which was 40%
above FY'88. The OSG organization continues to expand with the
addition of the OZIX group last year and the VAX System V group this
year. The total OSG headcount has doubled in the past two years and
now numbers approximately 450.
| |||||
| 45.13 | KERNEL::MOUNTFORD | Thu Jun 14 1990 15:10 | 273 | ||
From: COMICS::TREVENNOR "Ultrix Interest Mailing list: 13-Jun-1990 1911" 13-JUN-1990 19:15:07.31
To: @DIST:ULTRIXERS
CC: TREVENNOR
Subj:
UK Ultrix Support Group:
Mailshot #22
4 Items in this mailshot.
+---+ +---------------------------------+
| 1 | | 1989 Figures on the UNIX market |
+---+ +---------------------------------+
LEADING UNIX SYSTEMS MANUFACTURERS - 1989
($M VALUE OF SHIPMENTS)
1989 SHARE
---- -----
H-P 1820 13%
SUN 1810 13%
DIGITAL 1060 8%
CRAY 700 5%
AT&T 650 5%
IBM 640 5%
OTHERS 7000 51%
TOTAL $13680 100%
LEADING PC/WS SYSTEMS MANUFACTURERS - 1989
($M VALUE OF SHIPMENTS)
1989 SHARE
---- -----
SUN 1780 29%
HP 1470 24%
INTERGRAPH 380 6%
IBM 300 5%
DIGITAL 290 5%
OTHER 1940 31%
TOTAL $6160 100%
LEADING UNIX SMALL SCALE SYSTEMS MANUFACTURERS - 1989
($10K TO $100K, 2 TO 16 USERS)
($M VALUE OF SHIPMENTS)
1989 SHARE
---- -----
DIGITAL 370 10%
AT&T 320 8%
NCR 310 8%
IBM 230 6%
UNISYS 230 6%
OTHER 2390 62%
TOTAL $3850 100%
LEADING UNIX MEDIUM SCALE SYSTEMS MANUFACTURERS - 1989
($100K TO $1M, 17 TO 128 USERS)
($M VALUE OF SHIPMENTS)
1989 SHARE
---- -----
DIGITAL 400 17%
HP 270 11%
AT&T 200 8%
SEQUENT 140 6%
CONVEX 120 5%
PYRAMID 100 4%
OTHER 1180 49%
TOTAL $2410 100%
LEADING UNIX LARGE SCALE SYSTEMS MANUFACTURERS - 1989
(OVER $1M, OVER 128 USERS)
($M VALUE OF SHIPMENTS)
1989 SHARE
---- -----
CRAY 700 56%
AMDAHL 160 13%
IBM 100 8%
OTHER 300 24%
TOTAL $1260 100%
SOURCE OF ALL DATA ABOVE: IDC, APRIL 1990
+---+ +-----------------------------------------------+
| 2 | | Part Numbers for Ordering ULTRIX Support Kits |
+---+ +-----------------------------------------------+
Base Operating System
---------------------
V3.1 Support Kit part numbers:
media and documentation: BR-VEYAB-HW
documentation only: BR-VEYAB-3Z
V3.1B Support Kit part number: BR-VEYAK-25
V3.1C Support Kit part number: BR-VYVAM-25
ULTRIX V4.0 Support Kit: Same part number as for the two V3.1 kits. These
are not yet available until V4.0 FCS. I suspect that
FCS will not be until early July in the USA, it is
currently scheduled for 20-JUN-90.
Workstations
------------
UWS V2.2 Support Kit part numbers:
VAX media and documentation: BR-OJQAA-HW
RISC media and documentation: BR-VV1AA-HW
VS3520: BR-OJQAF-HW
VS3520 Update TK50: BR-OJQAG-HW
ULTRIX/WS V4.0 Support Kit: Same part number as for the two V2.2 VAX and RISC
kits. These are not yet available until V4.0 FCS.
Electronic V4.0 Kits: V4.0 kits are available for copying from the network.
Currently you can only get field test 2 software and
documentation. The FCS software and documenation will
also be made available this way.
To get information on obtaining the electronic kits
send mail to:
FTCODE::FTREGISTRAR
SUBJECT: HELP
You will be sent an electronic application form to be
filled out and returned. You will then receive a reply
giving you information on how and where to get the kit.
+---+ +--------------------------------------------------+
| 3 | | Programming in C from ULTRIX Course in July 1990 |
+---+ +--------------------------------------------------+
***********************************************************
*** ***
*** >>>>>>>>>>>>>>>>> ANNOUNCING <<<<<<<<<<<<<<<<<< ***
*** ***
*** From a screen play by Roy Schweiker ***
*** ***
*** Based on an original idea by Ken Thompson and ***
*** Dennis Ritchie ***
*** ***
*** >>>>>> Utilising ULTRIX Features From C <<<<<< ***
*** ***
***********************************************************
Prerequisites: 1) A knowledge of ULTRIX or UNIX(tm) to utilities
-------------- and command level (with subsequent practical
experience)
2) Proficiency in the "C" programming language
Target audience:1) (System and application) programmers wishing
---------------- to utilise ULTRIX system calls within (C) programs
2) System designers and system managers wishing to
gain insight into the features of ULTRIX
3) Those planning to attend an "ULTRIX
Internals" course
This course will be open to customers and employees
---------------------------------------------------
Description: This course covers the use of ULTRIX system calls
------------ and data structures for purposes such as: File and
Terminal IO; Controlling and creating processes;
and interprocess communication using a range
of techniques.
A fuller topic list can be found at the end
of this note
Location: Shire Hall, Reading, UK
& date 23-27 July 1990
Bookings: DTN 830-8020
(EY 2286 - Using ULTRIX Features From C)
For further information on course content, contact
Michael Kerrisk DTN: 830-4079
Mail: DOOZER::KERRISK
------------------------------------------------------------------------
Course Outline
==============
1. Compiling C programs
The preprocessor, compiler, assembler and linker
2. Program development tools
Object libraries, make, sccs
3. Debugging
With the dbx debugger
4. Allocating Memory
5. Controlling and inspecting the process envvironment
UIDs and GIDs, Process Ids, Process Groups
6. Ordinary Files
System calls for file IO, file and record locking
7. Creating and Controlling Processes
fork(), exec(), wait() and exit()
8. Signals
9. Terminal IO
setting terminal characteristics; non-blocking and signal
driven IO
10. Pipes
11. Sockets
In the UNIX and INTERNET domain: stream and datagram sockets
12. System V Interprocess Communication Facilities
Shared Memory, Semaphores, Message Queues
| |||||
| 45.14 | KERNEL::MOUNTFORD | Wed Jun 27 1990 23:21 | 71 | ||
UK Ultrix Support Group:
Mailshot #22
1 Item in this mailshot.
+---+ +--------------------------------------------+
| 1 | | Geoff Shingles on the need for UNIX growth |
+---+ +--------------------------------------------+
<Forwards deleted>
From: NAME: GEOFF SHINGLES
FUNC: BOARD OF MANAGEMENT
TEL: (7)830-3238 <SHINGLES AT A1_CHEFS AT SUBURB AT REO>
Date: 25-Jun-1990
Posted-date: 26-Jun-1990
Precedence: 1
Subject: UNIX
To: See Below
CC: See Below
** PLEASE COULD YOU PASS A COPY OF THIS NOTE TO ALL SALES, EIS AND **
CUSTOMER SERVICES STAFF
I just wanted to take a couple of minutes of your time to tell you how
I feel about UNIX as I know it is a topic of major interest to
everybody. It is today's biggest growth area!
I think that the Corporation has now really got its act together. Its
put one of our best people, Dom Lacava, in the driver's seat. He is
enthusiastic and active. He came and chatted to us at the European
Management Team meeting in Geneva and I was very impressed. He says
that our present plans are not nearly aggressive enough and won't get
us there. He believes that we must have at least a third of our
business coming out of UNIX within the next two or three years and he
is hell bent on doing that. This means doubling our market share by
the end of FY91.
This is being supported in Europe by putting one of the best
Marketeers we have in charge of the UNIX programme. This is Barry Nay
and his new role has just been announced.
I have charged Tony with the responsibility for getting one of the
very best people that it is possible to get to run the UNIX efforts
for us here in the UK. Incidentally, UNIX was put at the top of our
list of Marketing Programmes for FY91, so its funding is assured and
its focus is assured by having it at the head of the Marketing
priority and investment list. We will invest in major retraining
efforts and do all we need to protect our existing UNIX skills.
I am committed. We have started to change. We have started to get
some very good orders coming in. We have a long long way to go but we
have good products on which to build our efforts. During the next
twelve months every one of us has to put in the right amount of effort
to make sure that everybody understands inside and outside Digital
that we mean business.
This is our number one Marketing issue! That's what our customers
will see at DECworld.
Regards,
To Distribution List:
<deleted>
| |||||
| 45.15 | KERNEL::MOUNTFORD | Wed Jul 25 1990 08:46 | 116 | ||
Latest from Alan Trevennor:
osf osf unix posixposix xopenosf osf osf osf
osf osf svid posixposix osfwopen osf unix unix
osf osf unix svid osf osf osf unix unix
osf osf svid unix osfosf osf posixposix
osfosfosf unixunix svid osf osf osf svid svid
xopenunix xopenosf unix osf unix osf osf osf
xopenunix xopenosf unix osf unix osf osf osf
UK Ultrix Support Group:
Mailshot #24
1 Items in this mailshot.
+---+ +----------------------------+
| 1 | | Ultrix V4 summary and info |
+---+ +----------------------------+
From: NAME: Juliet Oshea
FUNC: Product & Technology
TEL: (7)833 3786 <OSHEA_J AT A1_KERNEL @THESUN @UVO>
To: See Below
ULTRIX-32 V4.0 is a functional enhancement and replacement for ULTRIX
V3.1. This information describes the new features of ULTRIX V4.0.
The support for ULTRIX V4.0 remains in the ULTRIX SSD Unit. If you
require any further information regarding this new version, please
contact Juliet O'Shea,(833-3786) Alan Trevennor (833-3286) or Alwyn
Bradley (833 3922).
----------------------------------------------------------------------
Digital ******* Company Confidential ****** Product & Technology
** New Version **
ULTRIX-32 V4.0 Ref:UKSP/jo/20-Jul-90 Rev A
FCS: 3-Aug-90
Description: V4.0 is an enhancement release of the ULTRIX-32
(UNIX-based) operating system. V4.0 functionality
includes SMP, LMF, VM changes for memory
utilisation, C2 security, NCS RPC, and support for
1159. It also includes NFS performance
enhancements; support for 96 RA devices; MIT Athena
Network Services -- Kerberos, Hesiod, Ntp; SNMP
Network Management Agent; Update Installation
option; Multi-architecture DMS and RIS, and
Consolidated Distribution. Languages, Tools &
Utilities will be enhanced for I18N compatibility,
and performance. For support, there will be a
multi-architecture crash analysis tool, DBBR for
SCSI, and the updated Kernel Messages Manual will
ship with the customer documentation set. Hardware
supported in this release includes: Calypso/Rigel,
ISIS, Mipsfair, 3MAX/R3000, and Teammate II.
Supported mass storage devices are: ESE20, TA90,
TQ70L, HSX, XSA, and RZ56. New communication and
printer support includes LPS40 and LN0*R postscript
and software features.
Impact: High
Class: ULTRIX, UNIX
Volume: For UK Sales assume 1/3 EUR.
FCS
QTR Q+1 Q+2 Q+3 Q+4 Q+5
EUR. 76 76 118 118 118 118
Logist/ESDC: WMOD206
Support Flow: CSC -> ULTRIX SSD ->
=> Corporate CSSE
Service: Standard Offerings Expected
PTG PFE: Alwyn Bradley, Alan Trevennor
PTG Mgr: Mike Smith
Starter Kits: Planned
Part Numbers: Q*-VEY**-**. SPD # 26.40.XX
--------------- L o c a l I n f o r m a t i o n ----------------
CSSE: Linda Trowbridge
Comments: ULTRIX V4.0 is the largest, most funtionally rich
ULTRIX release developed to date. More CPUs,
peripherals, standards, and features are supported
in Version 4.0 than ever before. It is the result
of 24 months of planning, engineering and testing.
When V4.0 was announced back in April, the Press
was unanimous in applauding the software efforts
that Digital (OSG) undertook and the results of our
efforts. The announcement, coupled with the new
RISC product line announcement, caused the leading
analysts to change their Digital stock ratings from
hold, to buy. This is something to be proud of,
their change in position was based primarily on new
ULTRIX and RISC products. Version 4.0 coupled with
our RISC H/W platform will keep Digital as a leader
in the open systems marketplace for the near
future. The cumulation of our efforts is already
paying off; sales are considerably up and more ULTRIX
orders have been booked than ever before.
| |||||
| 45.16 | Olsen on VMS on RISC | COMICS::TREVENNOR | A child of init | Tue Aug 14 1990 18:00 | 67 |
From the front page of the latest DEC Computing - mostly accurate.
Alan T.
Olsen confirms plans to put VMS on Risc
DEC president Ken Olsen laid out plans for the next generation of
VAX's in an interview given at the DECworld show in Boston last week.
He claims a Risc version of the VMS machines and improved operating
system and compiler technology will enable DEC to maintain its annual
price/performance growth rates for years to come.
Olsen said that even the VAX 9000 "is very close to being superseded".
DEC is clearly looking at a VAX architecture based on its own Risc
technology, and at an open - if not portable - version of VMS. The
machine dubbed the E-VAX by analysts will be based on multiprocessing
Risc CPUs and will appear to the user just like existing VAX machines,
but will offer similar price/perfomance to a Unix workstation.
But Olsen's commitment to Unix is, as ever, ambiguous and he now seems
to be thinking in terms of opening up VMS rather than gunning
wholeheartedly for Unix. "The industry focuses too much on operating
system kernels," he said. "The kernel, in one sense, you could say is
unimportant. If you follow the OSF standards for any operating system
you should have transportable software."
Olsen claims DEC has rethought its commitment to Unix as a commercial
operating system, and that while it is "beautiful" for its traditional
scientific purposes, it will never be as robust as VMS. "CPUs run
faster with Unix, but we believe VMS will catch up next year."
Olsen is also sceptical of the trend towards ever larger desktop
workstations, and remains respectful of the traditional terminal. "I
think mainframes and servers will have a lot fewer mips because
bandwith problems become so expensive to solve," he commented. "Power
is also in disk speed, number of mounted disks and bandwidth as well
as mips. I think most desktop computers will be simple dumb
terminals, maybe with windows."
With client/server architectures being DEC's theme for the show, Olsen
said: "The minicomputer is dead. We have servers now. Actually,
they do the same thing. We've always had servers." But he admitted
that the VAX 4000 is the first DEC machine specifically designed as a
server.
Olsen also confirmed that DEC would offer PCSA software for the
DECstations this autumn.
| |||||
| 45.17 | KERNEL::MOUNTFORD | Fri Dec 14 1990 11:12 | 71 | ||
More goodies from Alan.........
From: COMICS::TREVENNOR "Ultrix Interest Mailing list: 13-Dec-1990 1914" 13-DEC-1990 19:18:39.84
To: @DIST:ULTRIXERS
CC: TREVENNOR
Subj:
osf osf unix posixposix xopenosf osf osf osf
osf osf svid posixposix osfwopen osf unix unix
osf osf unix svid osf osf osf unix unix
osf osf svid unix osfosf osf posixposix
osfosfosf unixunix svid osf osf osf svid svid
xopenunix xopenosf unix osf unix osf osf osf
xopenunix xopenosf unix osf unix osf osf osf
UK Ultrix Support Group:
Mailshot #25
1 Item in this mailshot.
+---+ +-------------------------------------------+
| 1 | | Dom Lacava (Customer Services VP) on UNIX |
+---+ +-------------------------------------------+
Digital - UNIX "earthquake" shakes industry
{Livewire, 27-Nov-90}
At the UniForum exhibition held recently in London, Digital's commitment to
UNIX was underscored in Dom LaCava's keynote address:
UNIX "earthquake" shakes industry
Computer companies will need to undertake radical restructuring in order to
satisfy the market demand for open computing. In a keynote speech delivered at
UniForum's Open System's '90 conference in London, Dom Lacava, vice president,
UNIX-based Systems and Software, explained the ways in which vendors such as
Digital are adapting to the changing demands of the industry. He stated that
in order to offer truly open systems, computer vendors are having to undergo
significant change, which will undoubtedly affect corporate culture.
"Today UNIX is driving one of the most important changes in computing,"
commented Dom. "It's rumbling through the computing industry like an
earthquake. Users want UNIX and more importantly, open systems. Vendors can
see it as the tremendous opportunity it is, or they can deny that it's
happening. Part of seizing the opportunity is adapting the businesses to the
changing demands."
According to Dom, the immediate impact of open systems on vendors is the
need to create an organizational structure which will keep pace with the
rapidly changing market. For Digital this meant combining all the individual
parts of the company which were working on RISC and UNIX into one worldwide
organization, investing heavily in technical training and support, and
building UNIX resource centers in key cities to help customers develop and
port applications.
"Internationally, nationally and within our industry, the old ways are being
swept away faster than we ever imagined possible," he added. "Today, the
computer industry is a completely new game. Many of the rules which applied
only a few short years ago no longer apply, and every vendor is struggling
with the changes this has brought about. The vendors who will thrive are the
ones who, like Digital, can recognize the need to change the way they work.
They will apply the resources it takes to capitalize on the opportunities
offered by the promise, and move to open systems."
---
| |||||
| 45.18 | KERNEL::MOUNTFORD | Fri May 03 1991 09:25 | 130 | ||
From: COMICS::TREVENNOR "Ultrix Interest Mailing list: 02-May-1991 1255" 2-MAY-1991 13:00:19.63
To: @DIST:ULTRIXERS
CC: TREVENNOR
Subj:
osf osf unix posixposix xopenosf osf osf osf
osf osf svid posixposix osfwopen osf unix unix
osf osf unix svid osf osf osf unix unix
osf osf svid unix osfosf osf posixposix
osfosfosf unixunix svid osf osf osf svid svid
xopenunix xopenosf unix osf unix osf osf osf
xopenunix xopenosf unix osf unix osf osf osf
UK Ultrix Support Group:
Mailshot #26
1 Item in this mailshot.
+---+ +-----------------------+
| 1 | | Training opportunity |
+---+ +-----------------------+
Please find attached the course description for 'Technical Introduction
to UNIX and Open Systems Environments'. EY-G242, as announced in UCA
171 (March).
There are at least 10 slots available for the course which starts on
May 7th in Reading.
The course will not be run again in the South UK before Q2FY92.
The course aims to give a basic technical understanding and positioning
of Digital's capabilities in the Open Systems environment.
It is appropriate to Sales, EIS and Customer Services, so please pass
this on to your colleagues.
COURSE DESCRIPTION
******************
COURSE TITLE: Technical Introduction to the UNIX and Open Systems
Environment
COURSE OBJECTIVES:
The main focus will be on the practical business of responding to a
UNIX ITT and the knowledge necessary to position Digital's Unix
strategy, strengths and offerings at customer presentations.
To provide the "high level" technical awareness necessary to enable
Unix ITTs and proposals to be understood and processed. Candidates
completing the course should be able to handle a significant
percentage of the UNIX based work and planning locally. The aim of the
course is NOT to provide an in depth technical training in the subject
areas.
The course will contain presentations by Specialists and Consultants
with everyday experience of working in the UNIX environment.
PREREQUISITE
The candidates will be EIS Technical and Account specialists tasked
with supporting Sales on Digital's Unix strategy, positioning and
offerings. Unix awareness would be an advantage but not essential.
Sales specialists may also wish to attend.
COURSE DESCRIPTION
UNIX (Operating system and Networking)
- Customers view of the UNIX world
- their typical positions and misconceptions,
- how to position the Digital difference.
- Ultrix Version 4 training.
- Unix Networking & Communications.
- ITT experience
- Case studies
- example configurations.
- SUK Common Unix Environment using the version 4
features of Bind and Hesiod.
OFFICE
- Uniplex.
PC INTEGRATION
- What is an integrated PC?
- The Main Contenders in the market.
- LAN Manager - What is it and why is it important.
- Digital's Solutions.
- Graphical User Interfaces.
- Where to go for in-depth information and knowledge.
CASE
- Cohesion (the UNIX fit).
- UNIX CASE products (to include FUSE and VUIT).
DATABASE
- Overview of Database architecture and terms.
- ULTRIX/SQL V1.0 and V2.0.
- What you buy from Ingres.
- Distributed Access.
RISC PRODUCTS
- Products - Workstations, Servers, Graphics Options.
WINDOWS and GRAPHICS
- X Windows, MOTIF, DECwindows, Competition
- Graphics - GKS and PHIGS
OPEN SYSTEMS
- OSF1, Positioning and Major Differences
- Generic Open Systems and Standards
- Digital Offerings
UNIX ITT EXPERIENCE and CASE STUDIES
INTRUDER OVERVIEW and COMPETITION
| |||||
| 45.19 | Colourbook crashes at UC of Wales. | KERNEL::JAMES | Alan James CSC Basingstoke | Thu Jul 18 1991 10:34 | 63 |
Below is info on how to recognise Colourbook Software problem on
University College of Wales 5800. The present version of Colourbook
is field test and known to cause crashes.
REM> t
Entering TRANS mode [protocol off] (^A to exit)...
[?1l>
[?1h=
[K
aberdb 09:23:21 # Ov [?1l> !!!!! ALREADY LOGGED IN
[?1h=
[K
aberdb 09:23:40 # cd /usr/adm/crash[?1l> !!!! GO TO CRASH DIRECTORY
[?1h=
[K
aberdb 09:26:07 # ls -l vm*[?1l> !!!!! LOOK AT CRASH FILES
-rw-r--r-- 1 root 2940123 Jul 10 16:21 vmcore.83.Z
-rw-r--r-- 1 root 2887025 Jul 13 15:08 vmcore.84.Z
-rw-r--r-- 1 root 134180864 Jul 17 21:00 vmcore.85 !!!! 85 IS LATEST
-rw-r--r-- 1 root 1548443 Jul 6 14:27 vmunix.81.Z
-rw-r--r-- 1 root 1548443 Jul 10 16:19 vmunix.83.Z
-rw-r--r-- 1 root 1548443 Jul 13 15:06 vmunix.84.Z
-rw-r--r-- 1 root 3227444 Jul 17 20:58 vmunix.85 !!!!
[?1h=
[K
aberdb 09:26:21 # dbx -k vmunix.85 vmcore.85[?1l> !!!! GO TO DBX
dbx version 2.0
Type 'help' for help.
reading symbolic information ...
[using memory image in vmcore.85]
(dbx) t !!!!! GET STACK TRACE
> 0 boot(paniced = 0, arghowto = 0) ["../../machine/mips/machdep.c":544, 0x8011be04]
1 subr_prf.panic(s = 0x80182e7f = "trap") ["../../sys/subr_prf.c":1165, 0x8009d810]
2 kn5800_trap_error(ep = 0x801190e4, code = 2148639496, sr = 2148635012, cause = 805306376, signo
= 0xffffdc1c) ["../../machine/mips/kn5800.c":1655, 0x80118ea0]
3 trap.trap(ep = 0xffffdc40, code = 8, sr = 63508, cause = 805306376) ["../../machine/mips/trap.c
":447, 0x80124e14]
4 VEC_trap() ["../../machine/mips/locore.s":781, 0x8011a464]
5 .block580 ["../../netiso/net_uif.c":900, 0x80105e80]
6 net_read(dev = -2144354320, uio = 0xffffdec8) ["../../netiso/net_uif.c":900, 0x80105e80]
7 isoread(dev = 1, uio = 0xffffdec8) ["../../netiso/iso_uif.c":209, 0x800fb0e4]
^^^^^^^
!!!!! ISOREAD ONLY USED BY COLOURBOOK SOFTWARE AT PRESENT.
8 spec_rwgp(gp = 0x802d5d0c, uiop = 0xffffdec8, rw = UIO_READ, ioflag = 0, cred = 0xc96fd800) [".
./../fs/specfs/spec_vnodeops.c":289, 0x80081208]
9 rwgp(gp = 0x802d5cd0, uio = 0x8013c030, rw = UIO_READ, ioflag = 0, cred = 0xc96fd800) ["../../f
s/gfs/gfs_gnodeops.c":267, 0x8003f1d4]
10 gno_rw(fp = 0x802e245c, rw = -2146610140, uio = 0xf801) ["../../fs/gfs/gfs_gnodeops.c":204, 0x8
003efe0]
11 rwuio(uio = 0xffffdec8, rw = UIO_READ) ["../../sys/sys_generic.c":290, 0x8009e974]
12 read() ["../../sys/sys_generic.c":108, 0x8009e4c4]
More (n if no)?
13 syscall(ep = 0xffffdf5c, code = 3, sr = 1113916, cause = 32) ["../../machine/mips/trap.c":1070,
0x801261f8]
14 VEC_syscall() ["../../machine/mips/locore.s":836, 0x8011a4ec]
15 entry.start() [0x4154d4]
(dbx) quit
[?1h=
[K
aberdb 09:29:29 #
| |||||