Trailing-Edge
-
PDP-10 Archives
-
tops20_v6_1_tcpip_distribution_tp_ft6
-
6-1-documentation/tops20.bwr
There are 34 other files named tops20.bwr in the archive. Click here to see a list.
TOPS-20 Beware File
Field Test Tape 6
19 Jul 85
Version 6.1(6763)
Revision 3
COPYRIGHT (C) DIGITAL EQUIPMENT CORPORATION 1976, 1985. ALL RIGHTS
RESERVED.
THIS SOFTWARE IS FURNISHED UNDER A LICENSE AND MAY BE USED AND COPIED
ONLY IN ACCORDANCE WITH THE TERMS OF SUCH LICENSE AND WITH THE
INCLUSION OF THE ABOVE COPYRIGHT NOTICE. THIS SOFTWARE OR ANY OTHER
COPIES THEREOF MAY NOT BE PROVIDED OR OTHERWISE MADE AVAILABLE TO ANY
OTHER PERSON. NO TITLE TO AND OWNERSHIP OF THE SOFTWARE IS HEREBY
TRANSFERRED.
THE INFORMATION IN THIS SOFTWARE IS SUBJECT TO CHANGE WITHOUT NOTICE
AND SHOULD NOT BE CONSTRUED AS A COMMITMENT BY DIGITAL EQUIPMENT
CORPORATION.
DIGITAL ASSUMES NO RESPONSIBLITY FOR THE USE OR RELIABILITY OF ITS
SOFTWARE ON EQUIPMENT WHICH IS NOT SUPPLIED BY DIGITAL.
TOPS-20 Beware File, V6.1, Field Test Tape 6 Page 2
19 Jul 85
CONTENTS
of the
TOPS-20 Beware File, V6.1
V6-Series Field Test Tape #6
1.0 Function of this Document . . . . . . . . . . . . . 4
2.0 Critical Cautions and Corrections (Read these
carefully) . . . . . . . . . . . . . . . . . . . . . 4
2.1 LINK V6 Required . . . . . . . . . . . . . . . . . 4
2.2 Patches to the FT6 Monitor . . . . . . . . . . . . 4
2.3 LATSRV patch for stopping RESCHK BUGHLTs . . . . . 4
2.4 Patches to BASIC running on TOPS-20 V6.1 Field
Test Tape 6 . . . . . . . . . . . . . . . . . . . 5
2.5 RP20 Disks . . . . . . . . . . . . . . . . . . . . 6
2.5.1 Serial Numbers . . . . . . . . . . . . . . . . . 6
2.5.2 DX20 Microcode . . . . . . . . . . . . . . . . . 6
2.6 Executive DDT and MDDT now run in their own
section. . . . . . . . . . . . . . . . . . . . . . 6
2.7 MASSBUS Disk - both ports to same KL . . . . . . . 6
2.8 CI Disk Configurations . . . . . . . . . . . . . . 7
2.9 Breaking Away from the Cluster . . . . . . . . . . 7
2.10 %Drive forced offline because a running system
hasn't joined cluster . . . . . . . . . . . . . . 7
2.11 MSCP Server . . . . . . . . . . . . . . . . . . . 7
2.12 Power Fail Restart . . . . . . . . . . . . . . . . 8
2.13 Running V5.1 on a machine with a CI20 or NIA20
Installed . . . . . . . . . . . . . . . . . . . . 8
2.14 BASIC and Autopatch . . . . . . . . . . . . . . . 9
3.0 Less Critical Problems . . . . . . . . . . . . . . . 9
3.1 Bundled software . . . . . . . . . . . . . . . . 10
3.2 New Microcode and One Word Global Byte Pointers 10
3.3 Password Encryption . . . . . . . . . . . . . . 10
3.4 Job Numbers . . . . . . . . . . . . . . . . . . 10
3.5 Swapping Space . . . . . . . . . . . . . . . . . 11
3.6 New RSX20F . . . . . . . . . . . . . . . . . . . 11
3.7 SYSDPY . . . . . . . . . . . . . . . . . . . . . 11
3.8 MASSBUS Device Designations . . . . . . . . . . 11
3.9 Non-Wheel Access to Bootable Packs . . . . . . . 12
3.10 Reconstruction of Index-Table during Startup . . 12
3.11 PDVOP% JSYS call on .PONAM function code may
fail. . . . . . . . . . . . . . . . . . . . . . 12
3.12 PS: Structure Name and CFS . . . . . . . . . . . 13
3.13 Alternate Password Validation Algorithms . . . . 13
3.14 MMAILR and the use of POBOX: . . . . . . . . . . 14
3.15 MAILER . . . . . . . . . . . . . . . . . . . . . 15
3.16 Node name/number database size . . . . . . . . . 15
4.0 DUMPER Beware File . . . . . . . . . . . . . . . . 15
5.0 Beware Entries for the EXEC . . . . . . . . . . . 15
6.0 Beware Entries for GALAXY V5. . . . . . . . . . . 16
6.1 Dismounting disks under CFS . . . . . . . . . . 16
6.2 MOUNTR.CMD No Longer Used . . . . . . . . . . . 16
6.3 DEVICE-STATUS.BIN now on PS:[SYSTEM] . . . . . . 17
TOPS-20 Beware File, V6.1, Field Test Tape 6 Page 3
19 Jul 85
6.4 Two Structures of the Same Name . . . . . . . . 17
6.5 DECnet Node Online/Offline Messages . . . . . . 17
6.6 MOUNTR Compatibility with V5.1 . . . . . . . . . 18
7.0 Beware Entries for DECnet support . . . . . . . . 18
8.0 Beware Entries for TCP/IP support . . . . . . . . 18
8.1 Internet CI Support . . . . . . . . . . . . . . 18
8.2 Trailer Encapsulation Detection . . . . . . . . 18
9.0 Directory of front-end file system . . . . . . . . 18
TOPS-20 Beware File, V6.1, Field Test Tape 6 Page 4
Function of this Document 19 Jul 85
1.0 Function of this Document
This document supplements the TOPS-20 manuals by pointing out hints
and potential problems discovered at the last minute.
|
| Items which have been added since the Beware File for Field Test Tape
| #5 are marked with change bars in the left margin.
2.0 Critical Cautions and Corrections (Read these carefully)
2.1 LINK V6 Required
The TOPS-20 V6.1 monitor can only be built with LINK V6 or later.
LINK V6 is included on the TOPS-20 V6.1 distribution tape in this
field test kit. LINK V6 will run under TOPS-20 V6.0 and TOPS-20 V5.1
with Autopatch tape 8 installed. LINK V6 requires the SPLFK% JSYS and
thus will not run under earlier TOPS-20 monitors.
When linking TOPS-20 V6.1 monitors, it is normal to get LNKFTH
warnings. This is due to the way code is moved to section 6. See
Appendix B of TOPS20.DOC for more information.
2.2 Patches to the FT6 Monitor
Note that you may have to insert the following patches to any monitor
that you plan to run. The monitors on the field test tapes will not
have these patches installed.
In general, it is wise not to patch the monitor unless you are having
or anticipate having the particular problem addressed by a given
patch.
In the following patches, the user's input is shown in lower case, the
system's output in upper case. The $ symbols represent the escape
key.
2.3 LATSRV patch for stopping RESCHK BUGHLTs
A problem in the monitor causes TOPS-20 to crash with a RESCHK BUGHLT
when the command "ZERO COUNTERS" is given to LCP. Insert the
following patch if you have that problem.
TOPS-20 Beware File, V6.1, Field Test Tape 6 Page 5
Critical Cautions and Corrections (Read these carefully) 19 Jul 85
$get system:monitr.exe
$ddt
DDT
lazco+12/ EXTEND T1,LABFCK#+43 $<
FFF/ sosle t1 $1>
FFF+1/ 0 EXTEND T1,LABFCK#+42
FFF+2/ 0 JUMPA T1,LAZCO#+13
LAZCO#+12/ EXTEND T1,LABFCK#+42 JUMPA FFF1
^Z
$save system:monitr.exe
2.4 Patches to BASIC running on TOPS-20 V6.1 Field Test Tape 6
When running BASIC on TOPS-20 V6.1 Field Test Tape 6 with microcode
versions 350 and later, "?Illegal Memory Read" errors can occur
because of erroneous one word global byte pointers. The following
patches are corrections to the patch shipped with TOPS-20 V6.0.
!For BASIC-PLUS-2/20 versions prior to Autopatch Edit 335
@ENABLE
$GET SYS:BASIC.EXE
@DDT
DDT
SCCAVE+122/ DPB R1,JSYS4 SETZM JSYS4
SCCAVE+123/ DPB R1,JSYS3 SETZM JSYS3
SCCAVE+147/ DPB R1,JSYS4 SETZM JSYS4
SCCAVE+150/ DPB R1,JSYS3 SETZM JSYS3
DOOLD+52/ DPB J,JSYS4 SETZM JSYS4
DOOLD+53/ DPB J,JSYS3 SETZM JSYS3
DOOLD+161/ DPB J,JSYS4 SETZM JSYS4
DOOLD+162/ DPB J,JSYS3 SETZM JSYS3
^Z
@SAVE SYS:BASIC.EXE 0 120
BASIC.EXE.2 Saved
!For BASIC-PLUS-2/20 versions after Autopatch Edit 335
@ENABLE
$GET SYS:BASIC.EXE
@DDT
DDT
TOPS-20 Beware File, V6.1, Field Test Tape 6 Page 6
Critical Cautions and Corrections (Read these carefully) 19 Jul 85
SCCAVE+122/ DPB R1,JSYS4 SETZM JSYS4
SCCAVE+123/ DPB R1,JSYS3 SETZM JSYS3
SCCAVE+147/ DPB R1,JSYS4 SETZM JSYS4
SCCAVE+150/ DPB R1,JSYS3 SETZM JSYS3
^Z
@SAVE SYS:BASIC.EXE 0 120
BASIC.EXE.2 Saved
2.5 RP20 Disks
2.5.1 Serial Numbers
It is no longer necessary to use CHECKD for assigning serial numbers
to RP20 disks; in fact, the CHECKD command has been removed. The
monitor now creates drive serial numbers for RP20s. The DSN will be
8000+(unit number), decimal of course.
2.5.2 DX20 Microcode
Field Test Tape 5 included a new version of the DX20 microcode for
RP20 disks. Version 3(16) fixes a number of problems with error
recovery and bad block handling. It is important that the 8000
controller be in 370 mode for this new version. Please contact
field-service and have them verify that this condition is met if you
haven't already done so. (Reference 8000-2 MRM page U-210.)
2.6 Executive DDT and MDDT now run in their own section.
In the TOPS-20 V6.0 monitor, users executing JSYS's in MDDT used a
global stack pointer had a chance of crashing the system, if the JSYS
wasn't prepared for it.
Under V6.1 this is no longer a problem.
|
|
|
| 2.7 MASSBUS Disk - both ports to same KL
|
| There are problems with dual-porting a MASSBUS disk to 2 channels on
| the same KL. We recommend that you don't do this at this time;
| HOMGON BUGHLTs may occur. We are looking into the problem now.
TOPS-20 Beware File, V6.1, Field Test Tape 6 Page 7
Critical Cautions and Corrections (Read these carefully) 19 Jul 85
2.8 CI Disk Configurations
The following configurations are illegal (not supported by TOPS-20)
and TOPS-20 makes no predictions as to what will happen in such
siutations.
1. a MASSBUS disk dual-ported between a 6.1 system and a 5.1
system
2. a MASSBUS disk dual-ported between two 6.1 systems which are
on different CIs
2.9 Breaking Away from the Cluster
The OPR command SET PORT CI UNAVAILABLE is used if it is desireable to
break a system away from the cluster to run diagnostics on the CI.
OPR will direct that certain actions be taken so all multi-access
disks will be properly dismounted.
When the diagnostic procedures have been completed, the OPR command
SET PORT CI AVAILABLE must be issued. Upon issuing this command, the
user will be prompted with "Is this the only TOPS-20 system currently
running on the CI?". If the response is "No", the user will be
notified that timesharing must be rebooted before the system is
allowed to rejoin the cluster. If the response is "Yes", the CI
microcode will be reloaded and started.
2.10 %Drive forced offline because a running system hasn't joined
cluster
If you ever see this message, most likely there is another 20 which is
answering REQUEST-IDs but there is no CFS connection to it. Unless it
is obvious that the system is hung while booting, it's CI port is
still running but TOPS-20 is not. If it is hung, get a crash dump;
if it is down, PAR>HALT to stop the CI port.
2.11 MSCP Server
TOPS-20's MSCP server is not intended to be a general-purpose facility
around which one designs applications or standard procedures which
require long-term accessing of a disk. Rather, it is intended to be
used for quick access to disks which you otherwise can't access due to
your hardware configuration.
TOPS-20 Beware File, V6.1, Field Test Tape 6 Page 8
Critical Cautions and Corrections (Read these carefully) 19 Jul 85
If you choose to allow access to disks via the MSCP server you must
realize you are going to encounter some performance and operational
characteristics which may be less than desireable. That choice, of
course, is up to you; is the increase in accessibility to certain
disks worth the 1) potential decrease in performance AND 2)
operational confusion which may result? Here are 2 examples of the
price you pay:
1. If you run DUMPER to save files requiring access through an
MSCP server, you will most likely find that the DUMPER job
itself runs quite well but you are guaranteed the system
whose MSCP server is being used will pay a performance price.
2. If you shutdown a system whose MSCP server is being used, you
may hang jobs on other systems which require your MSCP server
for disk access if you do not follow the messages which OPR
outputs concerning structures which should be dismounted.
When a system shuts down timesharing, it notifies the other
20s in the cluster and they tell OPRs about any structures
which are mounted via the MSCP server on the system which is
going down.
|
|
| NOTE
|
| The MSCP server performance impact has
| been significantly reduced since FT5.
|
|
2.12 Power Fail Restart
| At this time the power fail restart code does not work for the CI or
| the NI. This will be fixed before 6.1 goes to the SDC.
2.13 Running V5.1 on a machine with a CI20 or NIA20 Installed
In order to run TOPS-20 V5.1 on a system with a CI20 or an NIA20
installed, you must patch the monitor, so that when it scans for RH20s
at system startup, it ignores the four higher-numbered RH20 slots.
This patch changes the value tested for ending the scan from 57400,,0
(code for RH20 slot 7) to 55400,,0 (code for RH20 slot 3).
;Get the V5.1 monitor
$get monitr.exe
$ddt
DDT
rh2nxt+4/ CAMLE Q1,LPHYH2#+13 $q[ 57400,,0 55400,,0
TOPS-20 Beware File, V6.1, Field Test Tape 6 Page 9
Critical Cautions and Corrections (Read these carefully) 19 Jul 85
^Z
$save monitr.exe
2.14 BASIC and Autopatch
Symptom:
With microcode versions 350 and later, OLD file.ext
gives ?Illegal Memory Read.
Diagnosis:
In module KOLD, routine DOOLD, after the file is read
in, it is closed with the CLOSF% JSYS. To zero ACs 2-4, the
following code sequence was being used.
JSYS2 _ .JSYS3 _ .JSYS4 _ 0;
This generated code to do DPB instructions with bogus
byte pointers. The old microcode did not trap these.
Cure:
Don't dot JSYS3 and JSYS4 so that they will be zeroed as intended. This is
reflected in Autopatch Edit 335 (DDT patch follows):
$get sys:basic.exe
@ddt
DDT
doold+52/ DPB J,JSYS4 setzm jsys4
doold+53/ DPB J,JSYS3 setzm jsys3
doold+161/ DPB J,JSYS4 setzm jsys4
doold+162/ DPB J,JSYS3 setzm jsys3
<ctrl-z>
@save sys:basic.exe 0 120
BASIC.EXE.2 Saved
3.0 Less Critical Problems
TOPS-20 Beware File, V6.1, Field Test Tape 6 Page 10
Less Critical Problems 19 Jul 85
3.1 Bundled software
The following bundled software shipped with TOPS-20 V6.1, has not had
significant maintenance updating since the previous TOPS-20 release.
It is quite possible that your system has a version of one or more of
the following programs at higher revision levels than those included
on this tape. Please review the list and compare them to those on
your system before superseding them.
1. MAKLIB
2. SCAN
3. WILD
3.2 New Microcode and One Word Global Byte Pointers
The KL microcode shipped with TOPS-20 versions 6.0 and 6.1 allows One
Word Global Byte Pointers in section zero. User software which has
byte pointers with P fields greater than 44 in section zero may now
fail. Under previous versions of the microcode, P fields greater than
44 were documented as giving undefined results.
One Word Global Byte Pointers are discussed in the Processor Reference
Manual.
KL microcode version 407 is shipped with TOPS-20 V6.1. KL microcode
version 350 was shipped with TOPS-20 V6.0.
3.3 Password Encryption
In general do not start using password encryption under TOPS-20 v6.1
until you are past the point at which it will be likely that you would
have to back off to V5.1 level monitors. Passwords encrypted under
V6.0 or V6.1 will have to reset under V5.1.
The NFT and FAL on this field test tape have been modified to work
with password encryption.
3.4 Job Numbers
With TOPS-20 V6.0 the monitor now considers job numbers to be a
CFS-configuration wide resource, and in particular allocates blocks of
job numbers to the systems in a CFS configuration. This may result in
a particular system assigning job numbers up to 512, though no one
system will support more than 128 jobs. Any programs using/getting
job numbers from TOPS-20 will thus now have to be able to deal with
TOPS-20 Beware File, V6.1, Field Test Tape 6 Page 11
Less Critical Problems 19 Jul 85
the entire number range from 0 to 512 and be able to deal with a
sparse allocation of such numbers on a particular system.
3.5 Swapping Space
With V6.0 and V6.1 we have greatly increased the use of extended
addressing, This may require some sites to increase the amount of
swapping space over the standard build amount, depending upon the
extended addressing usage at the site.
3.6 New RSX20F
This kit contains a new version of RSX20F. This version must be used
in conjunction with TOPS-20 V6.0 and V6.1.
This new version is not compatible with V5.1 (and previous) monitors,
so any backoff procedures must take this into account.
3.7 SYSDPY
The tool SYSDPY has had significant changes made to it. It is a
requirement that the SYSDPY that is shipped with TOPS-20 V6.1 be run
with that version of the monitor, and no other. Similarly the
versions of SYSDPY that were shipped with TOPS-20 V5.1 and V6.0 will
run with ONLY with their respective versions of the TOPS-20 monitor.
Running an incorrect version of SYSDPY may CRASH the system.
In the DECnet displays, the number of messages sent and received were
displayed in octal in previous versions. Now SYSDPY displays these
numbers in decimal.
3.8 MASSBUS Device Designations
Several utilities are being changed to reference disks and tapes using
decimal numbers rather than octal. This is a result of a change made
as part of V6.0, initiated because:
1. all CI designators (node numbers and HSC-disk units numbers)
are being displayed in decimal
2. VMS displays all such designators in decimal.
TOPS-20 Beware File, V6.1, Field Test Tape 6 Page 12
Less Critical Problems 19 Jul 85
Given the CI convention, we were presented with the prospect of having
some types of disks with decimal unit numbers and some with octal. It
was felt that it was best to have all such numbers in the same base.
Thus it was concluded that all displays that are oriented towards
operators or regular users should use decimal base for
unit/controller/channel number/etc. This includes Galaxy, utilities
(UNITS/CHANS/...), CHECKD, Monitor startup code, SYSDPY.
Displays at lower levels (not in general user oriented) will continue
to show octal format, in particular, BUGxxx additional data fields and
DDT.
Sites with RP20's will be the most affected by this change.
3.9 Non-Wheel Access to Bootable Packs
Recent edits to MEXEC and FILINI cause jobs to be started with no
special capabilities. This means that ALL bootable packs must allow
non-wheel access to the file EXEC.EXE and the directory it resides in.
If you get the NO EXEC message when logging on you can go into the MX
at the CTY and type
MX>reset
MX>get ps:<new-system>exec.exe
MX>start
(This will work since MX at the CTY gives you wheel). Use that EXEC
to change the protection of DEFAULT-EXEC: (normally SYSTEM:EXEC.EXE)
to 771212 and the protection of SYSTEM: to 774040. This should do
the trick!
3.10 Reconstruction of Index-Table during Startup
When the system is first booted on a pre-6.0 PS:, the system will take
some time to reconstruct the index-table so that it will be able to
include the new PPN support. A message to this effect will occur
shortly after the [PS MOUNTED] message. Do not be alarmed at the
duration of this operation. The PPN rebuild will also cause a CHECKD
pass to be run at the end of this first boot.
3.11 PDVOP% JSYS call on .PONAM function code may fail.
The PDV definition of the .PVNAM symbol (along with other such symbols
previously described as addresses) are now defined to be indirect
words. Hence, the .PONAM function call of PDVOP% will not return the
name string if the caller is in a nonzero section and the program is
using PDV's where the .PVNAM symbol is using the old address format.
Under the old format, an address of the form 0,,addr used in a nonzero
TOPS-20 Beware File, V6.1, Field Test Tape 6 Page 13
Less Critical Problems 19 Jul 85
section would cause a section relative reference to the PDV name
string. This change will impact you only in the rare event you are
making use of this function code of the PDVOP% JSYS and that the
program is in a nonzero section.
3.12 PS: Structure Name and CFS
CFS requires that all structures on systems on the same CI have unique
names. In particular, no more than a SINGLE system can have its
public structure name called PS:.
This functionality was available in V5.1. However, it was never
stated that it was of any advantage to call the PS: disks something
other than PS:. CFS now requires that this be done for at least one
of the systems in a dual system CI configuration.
We highly recommend that policies be established that cause all public
structures to have names other than PS:.
3.13 Alternate Password Validation Algorithms
Release 6 allows customers to select one of two methods for handling
password validation errors.
1. The first method, and the one enabled by default, is exactly
that present in previous versions of TOPS-20. That is, each
time an invalid password is entered, the process is blocked
for three seconds and then a JSYS error is produced.
2. The second method, new to release 6, incorporates some
techniques used by TOPS-20 customers. This technique does
not have a delay when a password is entered incorrectly, but
rather the JSYS error is produced immediately. However,
after a predefined number of incorrect guesses in a
predefined time, all subsequent passwords presented to
TOPS-20, be they valid or invalid, will be rejected. Also,
once this threshold is reached, TOPS-20 issues a BUGINF so
that the operator and the administrator will know that there
may be an intruder on the system.
The specifics of the algorithm are as follows:
* The choice of algorithm is made by the resident word
PSPNTP. If this word is zero the old algorithm is used;
if it is non-zero, the new algorithm is used.
* The time interval for considering failed passwords is
MINTVL and is set to three minutes. It may be changed by
defining a value in PARAMS.
TOPS-20 Beware File, V6.1, Field Test Tape 6 Page 14
Less Critical Problems 19 Jul 85
* The maximum number of incorrect passwords allowed in
MINTVL is MXFLCT and this is set to five in the standard
monitor. It may be changed by defining a new value in
PARAMS.
* Finally, the value MXLOGO, set to ten in the monitor, is
the maximum number of failed or rejected passwords
allowed before the job is logged out. This may be set to
infinity if this "service" is not desired. Setting this
to infinity means that password "guessers" will never
know for certain when the monitor is unconditionally
rejecting passwords. If one wishes "intruders" logged
out instead of allowed to present passwords to be
rejected, MXLOGO should be set to the same value as
MXFLCT.
Whereas setting MXLOGO to infinity may be desirable,
setting MINTVL to infinity may unduly penalize poor
typists (a common problem). MINTVL should be selected so
as to maximize the likelihood of trapping intruders and
minimizing the likelihood of confounding poor typists.
The preset value of three minutes seems a good
compromise.
* The BUGINF produced in the second method is called BREAKI
and has four optional data items: the TTY number, the
logged in user number, structure number and directory
number. The second item, the user number, will be zero
for a job that has not yet logged in.
3.14 MMAILR and the use of POBOX:
The MM and MMAILR mail system (on the tools tape) has not been
modified to take advantage of the POBOX: logical name option of the
mail system. MM and MMAILR will continue to look for mail on PS: and
thus may not work in an environment defining POBOX: to be a structure
other than PS:.
Also the newest MM (from Stanford) will not assemble with the
universals and MACRO supplied with TOPS-20 release 6.1. The MM
supplied on the V6.1 tools tape is not the newest MM (from Stanford).
The MM supplied on the V6.1 tools tape however will assemble with the
release 6.1 universals and MACRO.
TOPS-20 Beware File, V6.1, Field Test Tape 6 Page 15
Less Critical Problems 19 Jul 85
3.15 MAILER
MAILER has been changed to use the logical name POBOX: for mail
delivery. In addition, code that was once in MAILER to create the TO:
and CC: fields has been moved to MAIL.EXE. This should allow both MS
and MAIL to use the same local MAILER. However, MS must be patched to
send its IPCF packet to MAILER and not MAILEX. If MS is patched in
such a way, there is no need to use MAILEX at all. Otherwise, MAILEX
should continue to be used for local mail sent by MS users.
3.16 Node name/number database size
The DECnet node name/number database in FT5 has a maximum size.
Depending on how the node names hash into the database, the maximum
number of nodes is between 4546 and 5000 nodes. The DEC internal
network has grown to more than 4800 nodes, and the node database may
therefore not be able to accommodate all nodes. DEC internal sites
may want to change the NRNODS parameter in SCLINK.MAC to a larger
value. You then need to recompile SCLINK and relink the monitor.
In FT Tape #6, the node database will be allowed to grow dynamically,
and the NRNODS parameter will be removed.
|
|
|
| 4.0 DUMPER Beware File
|
|
| 1. There have been substantial changes to DUMPER consisting of
| bug fixes and new features to support release 6. It is very
| important to read and understand the contents of the
| DUMPER.BWR beware file which is included in this area.
|
5.0 Beware Entries for the EXEC
1. Setting Terminal Type
Setting terminal type to a video type terminal no longer
makes PAUSE END-OF-PAGE the default, as it was in 5.1
2. New EXEC Commands and Features
Please refer to the TOPS-20 Commands Reference Manual for
details of new EXEC commands. The EXEC.DOC file, in the
documentation saveset on the Distribution Tape, summarizes
the changed and new commands.
TOPS-20 Beware File, V6.1, Field Test Tape 6 Page 16
Beware Entries for the EXEC 19 Jul 85
3. PCL EXEC
The 6.1 PCL Exec is now using a later version of PCL than the
6.0 Exec. Therefore if a user has a PCL environment file, it
must be re-PRESERVED before PCL will function correctly.
4. MIC and Autopatch
MIC and PCL will not be distributed on Autopatch tapes;
however, the customer will be able to build a new MIC/PCL
EXEC from the EXEC .REL files on Autopatch tapes and the MIC
and PCL .REL files on the Tools tape.
5. New Abbreviation for LOGOUT
In the version of the EXEC shipping with TOPS-20 V6.0, LOG is
taken as the abbreviation for LOGOUT (assuming that you are
LOGged in ) and will LOG you OUT without a warning.
6. LOGOUT "n" and <ctrl-E>CEASE now require confirmation.
7. By default /FAST logins are allowed, if a site wishes to
restrict this, it must use a <ctrl-E>SET NO
FAST-LOGINS-ALLOWED or SETSPD's DISABLE FAST-LOGIN-OPTION.
6.0 Beware Entries for GALAXY V5.
6.1 Dismounting disks under CFS
Dismounting disks under a CFS configuration requires additional steps
to the normal procedure. See Operators Guide for details. Failure to
following correct procedure could result in users/systems becoming
hung.
6.2 MOUNTR.CMD No Longer Used
MOUNTR.CMD is no longer examined by MOUNTR. All disk/structure
information should be entered once as SET STRUCTURE commands in OPR.
The settings will remain in effect as long as the DEVICE-STATUS.BIN
file remains intact. Further information can be found in the
documentation.
TOPS-20 Beware File, V6.1, Field Test Tape 6 Page 17
Beware Entries for GALAXY V5. 19 Jul 85
6.3 DEVICE-STATUS.BIN now on PS:[SYSTEM]
In 6-series monitors before FT5, MOUNTR looked for the
DEVICE-STATUS.BIN file in SYSTEM: (the logical definition of
SYSTEM:). In a CFS environment, SYSTEM: could be defined with
directories that are common to all the systems in the cluster so that
system files can be shared within the cluster. However,
DEVICE-STATUS.BIN is not a shareable system file.
For this reason we have changed MOUNTR to locate the DEVICE-STATUS.BIN
file in PS:[SYSTEM]. In doing so, a new DEVICE.STATUS.BIN in
PS:[SYSTEM] is created when the new MOUNTR runs for the very first
time. Since this is a new DEVICE-STATUS.BIN the structures'
attributes will be the default cases. This means that any non-default
attributes must be reinstalled with the OPRSET STRUCTURE xx: ATTRIBUTE
command. It would be convenient to create a .CMD file to contain the
OPR SET commands and TAKE this file in OPR mode. For future use, this
file can be used to restore lost structure attributes if
DEVICE-STATUS.BIN is lost. Also, we have increased the
DEVICE-STATUS.BIN file to two pages, therefore the checksum of the
file will not correspond to any previous versions of
DEVICE-STATUS.BIN.
6.4 Two Structures of the Same Name
Two structures of the same name cannot be mounted; MOUNTR will reject
such a mount request as being ambiguous. The correct sequence to
examine two structures each with the same name is to:
o Mount the first structure with the second structure offline.
o Place the second structure online.
o Mount the second structure with an alias.
6.5 DECnet Node Online/Offline Messages
QUASAR no longer informs OPR if DECnet nodes go online or offline.
The OPR>SHOW STATUS NETWORK command shows only the status of the local
CI nodes and non-local nodes specified in the OPR>START OBJECT
command. The DECnet status is now reported to OPR by DECnet Event and
Link messages. DECnet Events can be filtered with the NCP>SET LOGGING
command, DECnet Link messages can be disabled with the OPR>DISABLE
OUTPUT DECNET-LINK command.
TOPS-20 Beware File, V6.1, Field Test Tape 6 Page 18
Beware Entries for GALAXY V5. 19 Jul 85
6.6 MOUNTR Compatibility with V5.1
If you want to run the MOUNTR from the V6.1 field test tape under a
5.1 system, you will have to make the following patch to MOUNTR:
$get mountr
$ddt
DDT
ststr1+10/ MSTR jfcl
^z
$save
7.0 Beware Entries for DECnet support
See DECNET.BWR on the DECnet tape.
8.0 Beware Entries for TCP/IP support
8.1 Internet CI Support
Release 6.1 supports the transmission of IP datagrams on both the NI
and CI. The low order byte of the internet address for the CI
interface must be the same as the CI node number of the system. If
the low order byte and the CI node number disagree a CIPBAD BUGINF
will result and the internet CI interface will be disabled.
8.2 Trailer Encapsulation Detection
TOPS-20 does not support trailer encapsulated datagrams on the
Ethernet (NI). Many Unix and Unix-based systems used trailed
encapsulated datagrams. If a trailer encapsulated datagram is
detected a IPTENC BUGINF will result. The trailer encapsulation
detection software uses three portals on the NI. This feature can be
disabled by patching the word TRLENA to zero.
9.0 Directory of front-end file system
The KL Model B Installation Guide mentions that the TOPS-20 Beware
File lists the directory of the front-end file system after the
RSX-20F Floppies have been copied to the front-end disk. The
directory follows:
DB0:[5,5]
TOPS-20 Beware File, V6.1, Field Test Tape 6 Page 19
Directory of front-end file system 19 Jul 85
F11ACP.TSK;1521 77. C 13-NOV-84 11:42
TKTN.TSK;1521 6. C 13-NOV-85 11:42
MOU.TSK;1521 5. C 13-NOV-84 11:42
KLX.MCB;407 55. 12-JUL-85 14:39
HALT.CMD;1 1. 13-NOV-84 12:16
BOOT.EXB;11311 46. 12-JUL-85 14:41
BF16N1.A11;1 1. 13-NOV-84 11:43
BF64N1.A11;1 1. 13-NOV-84 11:43
SB0.CMD;1 2. 13-NOV-84 11:43
SB1.CMD;1 2. 13-NOV-84 11:43
CLOCK.CMD;1 1. 13-NOV-84 11:43
EBUS.CMD;1 1. 13-NOV-84 11:43
DEX.CMD;1 1. 13-NOV-84 11:43
KPALV.CMD;1 1. 13-NOV-84 11:43
TIMEO.CMD;1 1. 13-NOV-84 11:43
FMPAR.CMD;1 1. 13-NOV-84 11:43
CRAM.CMD;1 1. 13-NOV-84 11:43
DRAM.CMD;1 1. 13-NOV-84 11:43
MTBOOT.EXB;11311 40. 12-JUL-85 14:43
LOOP.CMD;1 1. 13-NOV-84 11:43
PARSER.TSK;705 71. C 13-NOV-84 11:43
KLDISC.TSK;1521 5. C 13-NOV-84 11:44
KLRING.TSK;1521 6. C 13-NOV-84 11:44
LOGXFR.TSK;1521 10. C 13-NOV-84 11:44
MIDNIT.TSK;1521 4. C 13-NOV-84 11:44
SETSPD.TSK;1521 5. C 13-NOV-84 11:44
KLI.TSK;1521 73. C 13-NOV-84 11:44
T20ACP.TSK;1521 8. C 13-NOV-84 11:44
BOO.TSK;1521 19. C 13-NOV-84 11:44
COP.TSK;1521 8. C 13-NOV-84 11:44
DMO.TSK;1521 5. C 13-NOV-84 11:44
INI.TSK;1521 23. C 13-NOV-84 11:45
PIP.TSK;1521 56. C 13-NOV-84 11:45
RED.TSK;1521 6. C 13-NOV-84 11:45
SAV.TSK;1521 23. C 13-NOV-84 11:45
UFD.TSK;1521 9. C 13-NOV-84 11:45
ZAP.TSK;1521 38. C 13-NOV-84 11:45
RSX20F.SYS;1521 59. C 13-NOV-84 11:45
RSX20F.MAP;1521 154. 13-NOV-84 14:40
RP2DBT.EXB;11311 69. 12-JUL-85 14:42
RP2MBT.EXB;11311 60. 12-JUL-85 14:43
TOTAL OF 956./956. BLOCKS IN 41 FILES