| T.R | Title | User | Personal
Name | Date | Lines |
|---|
| 426.1 | Not 100% sure it's releated, but... | LESLIE::LESLIE | Egad! Iconised! | Fri Mar 17 1989 00:59 | 7 |
| Kristy,
I'd cut your MODPARAMS.DAT down to the minimum, then run autogen
with feedback. You still seem to be specifying way more than I ever
would.
-Andy
|
| 426.2 | | COMCAD::KRISTY | Iconize me! | Fri Mar 17 1989 09:30 | 6 |
| Andy,
All the stuff that is in HUGS's MODPARAMS.DAT is there because it
is needed. The machine doesn't come up at all or stay up very
long without the parameters that are currently in there.
|
| 426.3 | Ooops.. Should have said "5-8 lines in modparams".. :-) | STAR::MFOLEY | Rebel without a Clue | Fri Mar 17 1989 10:56 | 16 |
|
Kristy,
Follow CW's rule, Anything more than 58 lines in Modparams and you're
either a fool or a genius. (goes something like that)
Especially get rid of the SWAPFILE=0 and the PAGEFILE=0. Let
AUTOGEN do it's thing! Do you use FEEDBACK?
If a new satellite doesn't come up and stay up and be semi-usable
for a day or two while feedback runs, then it's a bug and should
be QAR'd. Putting a ton of lines in MODPARAMS could potentially
hide the problem.
mike
|
| 426.4 | | COMCAD::KRISTY | Iconize me! | Fri Mar 17 1989 11:19 | 3 |
| It's a brand spanking new satellite and hasn't been up long enough
for Feedback to real useful.
|
| 426.5 | CW is right... | MELTIN::dick | Schoeller - Xperimenting with Xnotes | Fri Mar 17 1989 12:15 | 10 |
| CW really is right. Strip your modparams.dat to the minimum. Delete all
old feedback files (if there are any). Use sysgen to set the system to the
default values. Run autogen and reboot. After about 2-3 days of normal
use, autogen with feedback. Do this about once a week for the 1st month
and then at irregular intervals after that. You will get much better
behavior than if you try to do that by manually tweaking modparams.dat.
Dick
I learned my lesson on this issue the hard way.
|
| 426.6 | | COMCAD::KRISTY | Iconize me! | Fri Mar 17 1989 13:08 | 10 |
| The reason the parameters are in there is because I started out with
the default values and they weren't good enough. When I let AUTOGEN do
its thing, it makes my machine's performance even worse. AUTOGEN
doesn't take into account the LPs the machine is running.
Then again, the question I asked isn't whether my modparams.dat is too
long or not, my question was asking what might be causing the problem
of the DECwindows login page not showing up after completely logging
out.
|
| 426.7 | I have this problem as well | DRIFT::WOOD | Laughter - the best medicine | Fri Mar 17 1989 13:40 | 9 |
| Kristy is not the only person with this problem. At least 50% of the time (it
seems more like 90%), when I log out, I do not get the login form back and have
to reboot. And I think I remember someone else mentioning the problem in this
notes file.
I would be happy to research the issue, if I only knew what to look for...
John
|
| 426.8 | | LESLIE::LESLIE | Trainee Lemming | Fri Mar 17 1989 14:12 | 18 |
| 1) I realise that you didn't post your Modparams.dat here for the
purpose of monopolising the discussion. However, PLEASE take the time
to QAR your problems,this sounds worrying.
2) What veriosn of VMS are you running? What is the configuration of
the cluster?
3) What is the configuration of the workstation?
4) Take a look at what DEC$SERVER_0 is doing after you log out (on
another terminal, of course). WHat State is it in after you logout?
Also, do a SHOW MEM/POOL/FULL/ALL and INSTALL LIST /GLOBAL/SUMM. Post
the results here.
5) STOP/ID the decw$server_0 process having noted where it went, and
@SYS$MANAGER:DECW$STARTUP - this should give you back your login
screen.
|
| 426.9 | as you wish... :-) | COMCAD::KRISTY | Iconize me! | Fri Mar 17 1989 15:41 | 85 |
| re: .-1:
Question 2: We are running VMS V5.1 in a homogenous cluster.
Question 3: The machines in question (the ones that are experiencing this) are
all VS2000 with 6 mb and RD5x local disks.
Question 4: DECW$SERVER_0 was in HIB state; There was also a DECW$TE_0 process
that was in LEF state.
After logging out (but not killing either of the two processes listed above),
install list/global/summary gave me the following results:
284 Global sections used, 24144/75856 Global pages used/unused
Doing a show mem/pool/full/all produced the following:
System Memory Resources on 17-MAR-1989 15:13:13.82
Physical Memory Usage (pages): Total Free In Use Modified
Main Memory (6.00Mb) 12288 2881 9407 0
Slot Usage (slots): Total Free Resident Swapped
Process Entry Slots 32 14 15 3
Balance Set Slots 30 17 13 0
Small Packet (SRP) Lookaside List Packets Bytes Pages
Current Total Size 3712 356352 696
Initial Size (SRPCOUNT) 2300 220800 432
Maximum Size (SRPCOUNTV) 9200 883200 1725
Free Space 35 3360
Space in Use 3677 352992
Packet Size/Upper Bound (SRPSIZE) 96
Lower Bound on Allocation 32
I/O Request Packet (IRP) Lookaside List Packets Bytes Pages
Current Total Size 600 105600 207
Initial Size (IRPCOUNT) 600 105600 207
Maximum Size (IRPCOUNTV) 2400 422400 825
Free Space 72 12672
Space in Use 528 92928
Packet Size/Upper Bound (fixed) 176
Lower Bound on Allocation 97
Large Packet (LRP) Lookaside List Packets Bytes Pages
Current Total Size 150 247200 483
Initial Size (LRPCOUNT) 150 247200 483
Maximum Size (LRPCOUNTV) 600 988800 1932
Free Space 139 229072
Space in Use 11 18128
Packet Size/Upper Bound (LRPSIZE + 80) 1648
Lower Bound on Allocation 1088
Nonpaged Dynamic Memory
Current Size (bytes) 899584 Current Total Size (pages) 1757
Initial Size (NPAGEDYN) 899584 Initial Size (pages) 1757
Maximum Size (NPAGEVIR) 2699776 Maximum Size (pages) 5273
Free Space (bytes) 187920 Space in Use (bytes) 711664
Size of Largest Block 132064 Size of Smallest Block 16
Number of Free Blocks 86 Free Blocks LEQU 32 Bytes 8
Paged Dynamic Memory
Current Size (PAGEDYN) 999936 Current Total Size (pages) 1953
Free Space (bytes) 789888 Space in Use (bytes) 210048
Size of Largest Block 785392 Size of Smallest Block 16
Number of Free Blocks 23 Free Blocks LEQU 32 Bytes 14
DISK$HUGS_57287:[SYSEXE]SWAPFILE.SYS;1
Free Blocks 49320 Reservable Blocks 49320
Total Size (blocks) 49992 Paging File Number 1
Swap Usage (processes) 3 Paging Usage (processes) 0
This file is used exclusively for swapping.
DISK$HUGS_57287:[SYSEXE]PAGEFILE.SYS;1
Free Blocks 16771 Reservable Blocks -11902
Total Size (blocks) 19992 Paging File Number 3
Swap Usage (processes) 0 Paging Usage (processes) 17
This file can be used for either paging or swapping.
Of the physical pages in use, 5459 pages are permanently allocated to VMS.
After killing the two processes, I did another show mem/pool/full/all and
got a fair amount more free pages. I did an @decw$startup and my login
screen did indeed come back, but not for a good five minutes.
|
| 426.10 | It just happened again - some details... | DRIFT::WOOD | Laughter - the best medicine | Fri Mar 17 1989 15:46 | 165 |
| > 2) What veriosn of VMS are you running? What is the configuration of
> the cluster?
> 3) What is the configuration of the workstation?
The workstation is a standalone VS2000 with 14MB, 2RD54, a TK50, and 8-plane
color. It is running the SDC versions of VMS V5.1 and DECwindows.
> 4) Take a look at what DEC$SERVER_0 is doing after you log out (on
> another terminal, of course). WHat State is it in after you logout?
> Also, do a SHOW MEM/POOL/FULL/ALL and INSTALL LIST /GLOBAL/SUMM. Post
> the results here.
See below for the output from these commands, and my MODPARAMS.DAT.
> 5) STOP/ID the decw$server_0 process having noted where it went, and
> @SYS$MANAGER:DECW$STARTUP - this should give you back your login
> screen.
Thanks, this worked and is much better than rebooting. I knew there had to be a
better way.
Please note - I hadn't re-autogened for a while, but I hadn't changed much
either. I just completed that process and rebooted. I'll attach a reply after
logging out and back in a few times, but that won't be until early next week.
Thanks for the help.
John
=========
MODPARAMS.DAT:
VAXCLUSTER=0
ADD_SRPCOUNT = 1024
ADD_SRPCOUNTV = 2048
ADD_RESHASHTBL = 512
ADD_NPAGEDYN = 400000
MIN_MAXBUF = 2048
SCSNODE="DRIFT"
SCSSYSTEMID=2564
CLISYMTBL=500
LOCKIDTBL = 512
RESHASHTBL = 2048
LNMSHASHTBL = 512
ADD_GBLPAGES = 2500
ADD_GBLSECTIONS = 120
MIN_MAXPROCESSCNT = 40
MIN_LRPCOUNT = 50
MIN_LRPCOUNTV = 100
MIN_PQL_DBYTLM = 20000
MIN_VIRTUALPAGECNT = 40000
MIN_PAGEDYN = 1100000
MIN_NPAGEDYN = 1000000
MIN_GBLPAGFIL = 12200
=============
SHOW SYS after logging out and not getting a login screen:
VAX/VMS V5.1 on node DRIFT 17-MAR-1989 15:19:49.01 Uptime 0 22:12:46
Pid Process Name State Pri I/O CPU Page flts Ph.Mem
00000041 SWAPPER HIB 16 0 0 00:00:17.71 0 0
00000103 SYSTEM CUR 5 124 0 00:00:03.03 535 386
00000044 ERRFMT HIB 8 747 0 00:00:08.68 75 100
00000045 JOB_CONTROL HIB 9 10086 0 00:01:40.35 167 358
00000047 OPCOM HIB 8 144 0 00:00:03.64 251 144
00000049 NETACP HIB 10 1282 0 00:00:37.45 325 350
0000004A EVL HIB 6 54 0 00:00:08.64 21064 48 N
0000004B PCFS_SERVER HIB 5 5456 0 00:01:40.27 345 588
0000004C DECW$SERVER_0 HIB 8 6990 0 00:34:16.28 19359 3677
0000004D RDMS_MONITOR LEF 15 24 0 00:00:01.34 914 37
0000004E RPC$SWL HIB 9 26 0 00:00:00.42 122 173
0000004F REMACP HIB 9 24 0 00:00:00.27 77 46
00000051 IDE$LN03R HIB 4 179 0 00:00:07.26 383 235
00000052 IDE$POST HIB 4 74 0 00:00:04.64 353 289
00000053 SYMBIONT_0003 HIB 4 9 0 00:00:03.89 229 55
00000054 SYMBIONT_0004 HIB 4 20 0 00:00:00.88 230 42
00000055 SYMBIONT_0005 HIB 4 18 0 00:00:00.88 239 42
00000056 SYMBIONT_0006 HIB 4 18 0 00:00:00.84 239 42
00000057 SYMBIONT_0007 HIB 4 18 0 00:00:00.89 239 42
00000058 DNS$ADVER HIB 4 1081 0 00:00:18.09 402 350
00000059 DFS$COM_ACP HIB 10 13 0 00:00:00.36 97 164
0000005A DFS$00010001_1 HIB 9 22 0 00:00:00.87 445 519
0000005B NMAIL_0001 HIB 4 1577 0 00:00:15.21 2476 36
0000005C NMAIL_0002 HIB 4 41 0 00:00:00.95 245 36
00000069 DECW$TE_1 LEF 4 32227 0 00:12:57.80 4367 2641
==============
Output from a show mem/pool/full/all:
System Memory Resources on 17-MAR-1989 15:20:18.86
Physical Memory Usage (pages): Total Free In Use Modified
Main Memory (14.00Mb) 28672 13370 15280 22
Slot Usage (slots): Total Free Resident Swapped
Process Entry Slots 50 24 26 0
Balance Set Slots 45 21 24 0
Small Packet (SRP) Lookaside List Packets Bytes Pages
Current Total Size 1384 132864 260
Initial Size (SRPCOUNT) 1384 132864 260
Maximum Size (SRPCOUNTV) 5122 491712 961
Free Space 1010 96960
Space in Use 374 35904
Packet Size/Upper Bound (SRPSIZE) 96
Lower Bound on Allocation 32
I/O Request Packet (IRP) Lookaside List Packets Bytes Pages
Current Total Size 306 53856 106
Initial Size (IRPCOUNT) 306 53856 106
Maximum Size (IRPCOUNTV) 680 119680 234
Free Space 171 30096
Space in Use 135 23760
Packet Size/Upper Bound (fixed) 176
Lower Bound on Allocation 97
Large Packet (LRP) Lookaside List Packets Bytes Pages
Current Total Size 54 88992 174
Initial Size (LRPCOUNT) 50 82400 161
Maximum Size (LRPCOUNTV) 100 164800 322
Free Space 44 72512
Space in Use 10 16480
Packet Size/Upper Bound (LRPSIZE + 80) 1648
Lower Bound on Allocation 1088
Nonpaged Dynamic Memory
Current Size (bytes) 999936 Current Total Size (pages) 1953
Initial Size (NPAGEDYN) 999936 Initial Size (pages) 1953
Maximum Size (NPAGEVIR) 1542656 Maximum Size (pages) 3013
Free Space (bytes) 416384 Space in Use (bytes) 583552
Size of Largest Block 348400 Size of Smallest Block 16
Number of Free Blocks 71 Free Blocks LEQU 32 Bytes 12
Paged Dynamic Memory
Current Size (PAGEDYN) 1099776 Current Total Size (pages) 2148
Free Space (bytes) 893472 Space in Use (bytes) 206304
Size of Largest Block 891792 Size of Smallest Block 16
Number of Free Blocks 22 Free Blocks LEQU 32 Bytes 14
DISK$VMSRL5:[SYS0.SYSEXE]SWAPFILE.SYS
Free Blocks 2896 Reservable Blocks 2896
Total Size (blocks) 2896 Paging File Number 1
Swap Usage (processes) 0 Paging Usage (processes) 0
This file is used exclusively for swapping.
DISK$VMSRL5:[SYS0.SYSEXE]PAGEFILE.SYS
Free Blocks 29426 Reservable Blocks -17478
Total Size (blocks) 37400 Paging File Number 3
Swap Usage (processes) 0 Paging Usage (processes) 25
This file can be used for either paging or swapping.
Of the physical pages in use, 5063 pages are permanently allocated to VMS.
==============
Output from install list/global/summary:
Summary of Local Memory Global Sections
276 Global Sections Used, 30648/2552 Global Pages Used/Unused
|
| 426.11 | | LESLIE::LESLIE | I work here. I carry a badge. | Fri Mar 17 1989 17:19 | 65 |
| re: .9
>DISK$HUGS_57287:[SYSEXE]SWAPFILE.SYS;1
> Total Size (blocks) 49992 Paging File Number 1
>DISK$HUGS_57287:[SYSEXE]PAGEFILE.SYS;1
> Total Size (blocks) 19992 Paging File Number 3
Kristy, your swapfile is HUGE and unnecessarily so. I run with a 3400
block swapfile. Very happily indeed. Swapfiles that big were made
redundant by VMS V5. A relic of 4.7's Modparams.dat maybe?
Similarly your pagefile is TINY! Please try making it at least 40000
blocks. I use 50k.
>284 Global sections used, 24144/75856 Global pages used/unused
Mein Gott! 75856 pages unused? Ya don't need that! Really1
RE: .10
>DISK$VMSRL5:[SYS0.SYSEXE]PAGEFILE.SYS
> Total Size (blocks) 37400 Paging File Number 3
This seems a little low, too. But I don't really know how relevant this
is.
-------------
If this happens regularly, do a SHOW PROC/ID=mumble/CONT to see if the
server process is actually doing anything. note the pc etc and include
it in your qar. You are submitting a qar, right?
From the above you'll have gathered that I didn't see anything really
obvious beyond pagefile stuff. Which is probably irrelevant. As I have
nothing to do with DECwindows developers anyway, that probably isn't
too great a surprise, right?
Incidentally,I use this command file to autogen with feedback. It works
for me.
----------
$! requires privs of course
$! original author: cw hobbs (I think)
$SET DEFAULT SYS$SPECIFIC:<SYSEXE>
$COPY VAXVMSSYS.PAR BEFORE.PAR
$
$@SYS$UPDATE:AUTOGEN SAVPARAMS SETPARAMS FEEDBACK
$
$DEF/USER SYS$OUTPUT BEFORE.LIS
$MC SYSGEN
USE BEFORE.PAR
SHOW/ALL
SHOW/SPEC
$
$DEF/USER SYS$OUTPUT VAXVMSSYS.LIS
$MC SYSGEN
USE CURRENT
SHOW/ALL
SHOW/SPEC
$
$DIFF /PAR BEFORE.LIS VAXVMSSYS.LIS
$
$ if p1 .eqs. "REBOOT" then -
@SYS$SYSTEM:SHUTDOWN 0 SHUTDOWN YES NO LATER YES NONE
|
| 426.12 | AUTOGEN/FEEDBACK Testimonial | TOWNS::DOERING | Aquarius of Age ?? | Fri Mar 17 1989 23:01 | 26 |
|
After following this particular note, I tried AUTOGEN + FEEDBACK
using the command procedure provided.
I don't know if it was the phase of the moon or what, but I hadn't
seen the problem reported in .0 at all until this week, when our
VS3200 did it to me about 3-4 times. I just logged in remotely,
and stopped/restarted DECW; curious, but not overly concerned.
This afternoon, I ran AUTOGEN + FEEDBACK on it, rebooted, logged in,
did a few things for awhile (30-45 mins), and then ended the session.
The login screen, came back, it seemed, almost instantly. I haven't
looked very closely at the DIFFs produced, but there were something
like 30/50 of sections/differences found (or however DIFF reports it).
I then went and did it on 2-3 of our other nodes in our 16 node LAVC,
and saw similar results. My next quest will be to see what, if
anything, it will do for our 4MB VS2000's that we have in the cluster.
Thanks for the info, Andy
Randy
|
| 426.13 | What's with the TE | DECWIN::FISHER | Burns Fisher 381-1466, ZKO3-4/W23 | Mon Mar 20 1989 12:18 | 13 |
| I am suspicious given that both people who reported trouble had a terminal
emulator process sitting around. That should go away when you log out.
I'm not sure about the implication here, except possibly that something
is incomplete about your login.
To get going again even faster, btw, stop/id the terminal emulator process
and then try @DECW$STARTUP APPS.
That should be faster than a complete DECW$STARTUP which has to restart the
server as well.
Burns
|
| 426.14 | Autogen didn't help | DRIFT::WOOD | Laughter - the best medicine | Mon Mar 20 1989 12:19 | 19 |
| I just tried logging out after using the system all morning and the problem is
still there - the login screen did not appear after logout. This is after
doing an Autogen on Friday.
In this case, feedback didn't help much since I log out every night and ended up
rebooting every night when this hung. I'll try either pausing overnight, or
restarting DECwindows for a few days so feedback can have a chance to help.
FWIW, the applications that I'm running all the time include 4 DECterms created
via CHILD, VAXnotes (T2.0-23), Calendar, and Cardfiler. Some others are run
randomly, but not since the last reboot.
Now where did I see the info on how to submit a QAR...
John
P.S. If anyone in ZK would like to look at it when it is hung, send me mail and
I can try and arrange for a demonstration.
|
| 426.15 | Opened QAR #796 | DRIFT::WOOD | Laughter - the best medicine | Mon Mar 20 1989 13:32 | 10 |
| I did open a QAR on this problem, but I just referenced this note for the
details.
I am modifying my startup to temporarily eliminate the use of the CHILD utility
to try and remove some of the variables. I'll report back after a day or two
of logging out and back in. I don't really suspect CHILD, but it is the only
piece of unofficial software on the system (other than NOTES field test).
John
|
| 426.16 | | LESLIE::LESLIE | I work here. I carry a badge. | Mon Mar 20 1989 14:26 | 4 |
| I use Child to kick off three DECterms, with no adverse effects.
Andy
|
| 426.17 | CHILD may be exposing a window mgr problem | PRNSYS::LOMICKAJ | Jeff Lomicka | Tue Mar 21 1989 10:44 | 10 |
| I have noticed that if there are no terminal emulator windows open, that
the session manager (or whatever it is that kills processes when you log
out) won't stop the terminal controller. If the terminal processes die
fast enough, this may be related to the problem. I have only noticed it
when I remote DECterms up, and it didn't prevent the LOGIN display, all
it did was leave behind my custom screen background.
I'm looking into it today.
|
| 426.18 | set display/noperm? | SPIDER::HERRLICH | | Wed Mar 22 1989 13:30 | 12 |
| Do you do a set display/noperm or have an automatic procedure that does it
every so often?
I can't test it out at the moment because I'm running ULTRIX), but I do remember
a problem that if you set display/noperm on the display that the session manager
uses you will lose the display when you logout.
Makes some sense... at some point during the logout to login process nobody is
using the display, so being a nice temporary device it goes away.
- Alan
|
| 426.19 | maybe it's just something I'm doing wrong... | COMCAD::KRISTY | Iconize me! | Wed Mar 22 1989 13:46 | 20 |
| Here is part of the remote_startup.com
$ run/detached sys$system:loginout.exe -
/input=$decw:remote_child.com -
/error=$decw:remote_child.log -
/output=$decw:remote_child.log -
/proc="Hugs DECterm"-
/buffer_limit=30000/enqueue_limit=800/extent=10000 -
/io_buffered=100/io_direct=100/maximum_working_set=2048 -
/page_file=30000/authorize/file_limit=64/working_set=512
and this is remote_child.com
$ set display/create/node=HUGS
$ child/image=sys$system:loginout.exe/nopass/detach/shrink/x=50/y=125 -
/ititle="Fasted"/title="Fasted Hugs I" -
/setup=$decw:decw$child1.dat/page=30
$ exit
|
| 426.20 | Too often for me now | DAVE::MITTON | MS-DOS, Over 15 million sold | Tue Jun 13 1989 22:12 | 20 |
| Well, this is now happening to me almost every time.
Especially, if I take down all active processes, nice and neatly before
hitting the QUIT button in the Session Manager.
After it quits, No DEC logo appears and the screen remains in a random
solid color.
(interestingly this became more reproducible after I upped the
secondary PAGEFILE size on my GPX.)
My normal recovery now, is to SETHOST in from a PC, and rerun
DECW$STARTUP.COM
What should I be looking at to debug this?
What information would be interesting for me to provide?
SDC VMS V5.1 & DECwindows, VAXstation II/GPX
Dave.
|
| 426.21 | some more info | DAVE::MITTON | MS-DOS, Over 15 million sold | Tue Jun 13 1989 22:37 | 34 |
| I played with this again to check some hunches...
- It only fails if I close down all windows first
If I leave everything/anything up, it works fine
- The screen doesn't go random, it stays in my defined background
color. ie: it never gets cleared to black then grey halftone like it
normally would. Which confuses the color map when I come in over the
network and kick DECW$STARTUP to get it up again.
(hopefully that gives a clue whereabouts it died.... and
why does the screen normally do all those color changes anyways???)
The following is my most recent DECW$SM.LOG
-------------
%SET-I-NEWLIMS, new working set: Limit = 200 Quota = 200 Extent = 2048
%SET-W-NOTSET, error modifying DUA0:
-CLI-E-IVDEVTYPE, invalid device type - specify a mailbox device
%SET-E-NOTSET, error modifying process name
-SYSTEM-F-DUPLNAM, duplicate name
Indecision is the key to flexibility.
Identification of created subprocess is %Xde
Identification of created subprocess is %Xdf
%DCL-S-SPAWNED, process DECW$CALENDAR spawned
Session Message: Started DECterm controller
XIO: non-translatable vms error code: 0x18292, vms message:
%rms-e-fnf, file not found
%XLIB-F-IOERROR, xlib io error
MITTON logged out at 13-JUN-1989 22:10:27.38
|
| 426.22 | I have the same problem on V5.1 & V5.2-42W | KAFS30::BOIVIN | Moi, j'viens du nord! | Wed Jun 14 1989 08:18 | 44 |
| I have the same problem as .20 but I have narrowed it down to my
DECW$SYLOGIN.COM. In DECW$SYLOGIN I use CHILD to start up VWSLAT as a
subprocess:
$ CHILD -
/SHRINK -
/PAGE=24 -
/PROC="VWSlat" -
/ITITLE="VWSlat" -
/TITLE="VWSLAT control terminal" -
"run sys$system:vwslat"
Everything comes up fine but if I quit the session, the login screen never
comes back. *Occasionally* there's still a DECterm process alive after I've
quit. I have worked around the problem by starting an early DECterm subprocess
in DECW$SYLOGIN...
$! Create an Early DECterm so we can start VWSlat early.
$ spawn/nowait/input=nl:/proc=early_decterm run sys$system:decw$terminal
$!
$! Wait for the mailbox to show up
$!
$ node_name = f$getsyi( "nodename" )
$ waiting:
$ WAIT 0:0:05
$ mbx = f$logical( "DECW$DECTERM_MAILBOX_''node_name'::0.0")
$ IF mbx .NES. "" THEN -
IF f$getdvi( mbx, "exists") THEN GOTO decterm_ready
$ WAIT 0:0:05
$ GOTO waiting
$ decterm_ready:
$! now let's start up VWSLAT.
$ CHILD -
/SHRINK -
/PAGE=24 -
/PROC="VWSlat" -
/ITITLE="VWSlat" -
/TITLE="VWSLAT control terminal" -
"run sys$system:vwslat"
Hope this can shed some light?
Ed
|