Cory Smelosky <b4 at gewt.net> writes:
On Mon, 7 Oct 2013, Cory Smelosky wrote:
Let's assume you have no need for your prior accounting information...
$ SET ACCOUNTING/NEW_FILE/ENABLE=IMAGE
Execute the UUCP command(s) which fail.
$ ACCOUNTING/SINCE=TODAY
Post that output and let's see if there's anything revealing in there.
Also, you can $ SET ACCOUNTING/DISABLE=IMAGE because /ENABLE=IMAGE can
become a resource hog.
Doing so now.
I disabled TERM/INQ on login in SYLOGIN and the abort went away
(ACCOUNTING didn't have anything useful to say). I know I can disable it
per-line/per-use but this was the quickest test.
Now we're getting somewhere.
Try putting:
$ IF "''F$mode()'".EQS."NETWORK" THEN $ EXIT
at the beginning of your SYLOGIN.COM and LOGIN.COM files.
You can not SET TERM/INQUIRE if there's no terminal. ;)
--
VAXman- A Bored Certified VMS Kernel Mode Hacker VAXman(at)TMESIS(dot)ORG
Well I speak to machines with the voice of humanity.
On Mon, Oct 7, 2013 at 11:31 PM, Cory Smelosky <b4 at gewt.net> wrote:
On Mon, 7 Oct 2013, Cory Smelosky wrote:
Let's assume you have no need for your prior accounting information...
$ SET ACCOUNTING/NEW_FILE/ENABLE=IMAGE
Execute the UUCP command(s) which fail.
$ ACCOUNTING/SINCE=TODAY
Post that output and let's see if there's anything revealing in there.
Also, you can $ SET ACCOUNTING/DISABLE=IMAGE because /ENABLE=IMAGE can
become a resource hog.
Doing so now.
I disabled TERM/INQ on login in SYLOGIN and the abort went away (ACCOUNTING didn't have anything useful to say). I know I can disable it per-line/per-use but this was the quickest test.
I've seen this screw lots of stuff. I have stuff in my personal LOGIN.COM that configures the terminal they way I like it. I just wrap it in the following:
$ interactive = (f$mode() .eqs. "INTERACTIVE")
$ network = (f$mode() .eqs. "NETWORK")
$ batch = (f$mode() .eqs. "BATCH")
...
$ if (interactive)
$ then
$ set terminal ...
$ endif
Regards, Tim.
On Mon, 7 Oct 2013, Brian Schenkenberger, VAXman- wrote:
Cory Smelosky <b4 at gewt.net> writes:
On Mon, 7 Oct 2013, Brian Schenkenberger, VAXman- wrote:
Cory Smelosky <b4 at gewt.net> writes:
On Mon, 7 Oct 2013, Jordi Guillaumes i Pons wrote:
Any chance you have got two different DCLTABLES in SYS$COMMON and SYS$SPECIFIC?
I'd answer if I could find where in sys$common and sys$specific they
are...
DCLTABLES is an executable. It's a mapped image section. It's name is
DCLTABLES.EXE and it should be in the common root of SYS$LIBRARY. Thus,
you should be looking in SYS$COMMON:[SYSLIB]. If you have DCLTABLES.EXE
in SYS$SPECIFIC:[SYSLIB], strange things (ie. unpredictable results) can
occur when products add their command definitions to DCLTABLES.EXE.
Looks like I don't have a DCLTABLES.EXE in SYS$SPECIFIC.
The question then begs to be asked, why do you think your command
definitions are changing or being deleted.
Well, after the install, or random subsequent logins CC, COBOL, FORTRAN, or PASCAL all were valid commands. Now after subsequent reboots/logins they are not.
--
Cory Smelosky
http://gewt.net Personal stuff
http://gimme-sympathy.org Projects
Cory Smelosky <b4 at gewt.net> writes:
On Mon, 7 Oct 2013, Brian Schenkenberger, VAXman- wrote:
Cory Smelosky <b4 at gewt.net> writes:
On Mon, 7 Oct 2013, Jordi Guillaumes i Pons wrote:
Any chance you have got two different DCLTABLES in SYS$COMMON and SYS$SPECIFIC?
I'd answer if I could find where in sys$common and sys$specific they
are...
DCLTABLES is an executable. It's a mapped image section. It's name is
DCLTABLES.EXE and it should be in the common root of SYS$LIBRARY. Thus,
you should be looking in SYS$COMMON:[SYSLIB]. If you have DCLTABLES.EXE
in SYS$SPECIFIC:[SYSLIB], strange things (ie. unpredictable results) can
occur when products add their command definitions to DCLTABLES.EXE.
Looks like I don't have a DCLTABLES.EXE in SYS$SPECIFIC.
The question then begs to be asked, why do you think your command
definitions are changing or being deleted.
--
VAXman- A Bored Certified VMS Kernel Mode Hacker VAXman(at)TMESIS(dot)ORG
Well I speak to machines with the voice of humanity.
On Mon, 7 Oct 2013, Cory Smelosky wrote:
Let's assume you have no need for your prior accounting information...
$ SET ACCOUNTING/NEW_FILE/ENABLE=IMAGE
Execute the UUCP command(s) which fail.
$ ACCOUNTING/SINCE=TODAY
Post that output and let's see if there's anything revealing in there.
Also, you can $ SET ACCOUNTING/DISABLE=IMAGE because /ENABLE=IMAGE can
become a resource hog.
Doing so now.
I disabled TERM/INQ on login in SYLOGIN and the abort went away (ACCOUNTING didn't have anything useful to say). I know I can disable it per-line/per-use but this was the quickest test.
--
Cory Smelosky
http://gewt.net Personal stuff
http://gimme-sympathy.org Projects
On Mon, 7 Oct 2013, Brian Schenkenberger, VAXman- wrote:
Cory Smelosky <b4 at gewt.net> writes:
On Mon, 7 Oct 2013, Jordi Guillaumes i Pons wrote:
Any chance you have got two different DCLTABLES in SYS$COMMON and SYS$SPECIFIC?
I'd answer if I could find where in sys$common and sys$specific they
are...
DCLTABLES is an executable. It's a mapped image section. It's name is
DCLTABLES.EXE and it should be in the common root of SYS$LIBRARY. Thus,
you should be looking in SYS$COMMON:[SYSLIB]. If you have DCLTABLES.EXE
in SYS$SPECIFIC:[SYSLIB], strange things (ie. unpredictable results) can
occur when products add their command definitions to DCLTABLES.EXE.
Looks like I don't have a DCLTABLES.EXE in SYS$SPECIFIC.
--
Cory Smelosky
http://gewt.net Personal stuff
http://gimme-sympathy.org Projects
Cory Smelosky <b4 at gewt.net> writes:
On Mon, 7 Oct 2013, Jordi Guillaumes i Pons wrote:
Any chance you have got two different DCLTABLES in SYS$COMMON and SYS$SPECIFIC?
I'd answer if I could find where in sys$common and sys$specific they
are...
DCLTABLES is an executable. It's a mapped image section. It's name is
DCLTABLES.EXE and it should be in the common root of SYS$LIBRARY. Thus,
you should be looking in SYS$COMMON:[SYSLIB]. If you have DCLTABLES.EXE
in SYS$SPECIFIC:[SYSLIB], strange things (ie. unpredictable results) can
occur when products add their command definitions to DCLTABLES.EXE.
--
VAXman- A Bored Certified VMS Kernel Mode Hacker VAXman(at)TMESIS(dot)ORG
Well I speak to machines with the voice of humanity.
On Mon, 7 Oct 2013, Brian Schenkenberger, VAXman- wrote:
Cory Smelosky <b4 at gewt.net> writes:
On Mon, 7 Oct 2013, Cory Smelosky wrote:
Hello,
After the reinstall to 7.3 I've made progress! UUCP no longer gives quota
exceeded errors...it just gives a message about it having aborted.
Nothing else is in the debug logs.
Enable image accounting and see what is left in there after UUCP aborts.
Doing so now. Should have an answer in 9 minutes.
CC, Pascal, COBOL, and FORTRAN also
seem to be in DCLTABLES for awhile...but after I log out they're no longer
listed. WTF?
How are you "listing" CC, Pascal and Fortran verbs in DCLTABLES? Where
is DCLTABLES.EXE on your system? It should reside in SYS$COMMON[SYSLIB].
I may be mistaken as to if it's getting shoved in to DCLTABLES...as running the startup script isn't defining any symbols for CC or similar, I assume they were being tossed in to DCLTABLES.
I've run their requisite startup scripts! How the hell did
I break DCLTABLES?
What makes you think that you did?
The error seems impossible to happen spontaneously.
(I've increased the page table limits in case they
were getting full. I'm also getting no debug info related to this,
either)
At least the quota issues disappeared. ;)
Let's assume you have no need for your prior accounting information...
$ SET ACCOUNTING/NEW_FILE/ENABLE=IMAGE
Execute the UUCP command(s) which fail.
$ ACCOUNTING/SINCE=TODAY
Post that output and let's see if there's anything revealing in there.
Also, you can $ SET ACCOUNTING/DISABLE=IMAGE because /ENABLE=IMAGE can
become a resource hog.
Doing so now.
--
Cory Smelosky
http://gewt.net Personal stuff
http://gimme-sympathy.org Projects
On 07/10/2013 14:41, Brian Schenkenberger, VAXman- wrote:
Mark Wickens <mark at wickensonline.co.uk> writes:
On 07/10/2013 14:04, Daniel Soderstrom wrote:
Only has 13 views so I was thinking you might not have seen this video.
http://m.youtube.com/watch?v=UfBMyvIBYAw
There are so few videos. Have any of you seen the doco that was released by PBS in the US? Its not available from anywhere except the PBS online strore and they wont ship overseas.
Regards,
Daniel.
Wow, thanks for sharing that, what an amazing data centre.
It'll make the folks on this list drool for sure!
I saw your comment there too. I finally got to see what you look like.
Yeah, and unlike most peoples online photos, that one is fairly recent! It's actually a passport photo. No smiling allowed. I do smile occasionally, when my extended family isn't making my life hell ;)
I was thinking a page with photos of hecnetters might be a good idea. I could easily be persuaded otherwise however!
Mark.
--
http://www.wickensonline.co.ukhttp://hecnet.euhttp://declegacy.org.ukhttp://retrochallenge.nethttps://twitter.com/#!/%40urbancamo
On Mon, 7 Oct 2013, Jordi Guillaumes i Pons wrote:
Any chance you have got two different DCLTABLES in SYS$COMMON and SYS$SPECIFIC?
I'd answer if I could find where in sys$common and sys$specific they are...
Jordi Guillaumes i Pons
Barcelona - Catalunya - Europa
El 07/10/2013, a les 5:15, Cory Smelosky <b4 at gewt.net> va escriure:
On Mon, 7 Oct 2013, Cory Smelosky wrote:
Hello,
After the reinstall to 7.3 I've made progress! UUCP no longer gives quota exceeded errors...it just gives a message about it having aborted. Nothing else is in the debug logs. CC, Pascal, COBOL, and FORTRAN also seem to be in DCLTABLES for awhile...but after I log out they're no longer listed. WTF? I've run their requisite startup scripts! How the hell did I break DCLTABLES? (I've increased the page table limits in case they were getting full. I'm also getting no debug info related to this, either)
At least the quota issues disappeared. ;)
--
Cory Smelosky
http://gewt.net Personal stuff
http://gimme-sympathy.org Projects
--
Cory Smelosky
http://gewt.net Personal stuff
http://gimme-sympathy.org Projects
