T.R | Title | User | Personal Name | Date | Lines |
---|
560.1 | Verify CDD$DEFAULT is valid | BUFFER::VICKERS | Perfect is the enemy of good | Thu Apr 23 1992 19:04 | 12 |
| I have seen similar results when the DATATRIEVE default library
setting, CDD$DEFAULT, is not set to an existing node in the library.
You might want to verify any other DATATRIEVE specific logicals as well
as the ALL-IN-1 interface to DATATRIEVE does seem a bit less forgiving
than is the DCL interface during initialization.
Good luck,
don
P.S. Don't forget to be careful with keeping ALL-IN-1 (and DATATRIEVE)
in all caps so jerks like me don't complain about the trademark. :')
|
560.2 | Security alert | IOSG::TALLETT | Just one more fix, then we can ship... | Thu Apr 23 1992 20:32 | 11 |
|
Installing Datatrieve with SYSPRV sounds like a security hole
to me. Can't I just knock up a record layout for, say, SYSUAF.DAT
and gaily write to the UAF file?
Seems like you should change the protection on the CDD file, but
then again, I don't know if that is a problem (but it doesn't
sound as big a problem as a rampant DTR image with SYSPRV).
Regards,
Paul
|
560.3 | Logicals seem to be ok | KERNEL::WILESL | Louise Wiles, TP/IM Support | Fri Apr 24 1992 13:59 | 14 |
| Re: .1
Thanks Don,
I checked CDD$DEFAULT, it's set to CDD$TOP, from ALL-IN-1 I spawned out
to DCL level, from here I checked with sho log DTR* & CDD*. Is this the
best way to get this info?
Thanks,
Louise.
PS Sorry about the ALL-IN-1 caps mistake, but I actually entered the
note prematurely in my haste ;-)
|
560.4 | Use SET WATCH FILE to check for protected files | BUFFER::VICKERS | Perfect is the enemy of good | Fri Apr 24 1992 23:58 | 19 |
| Louise,
I assume that you're still getting the 'no privilege' message.
The next step that I would take would to use
$ SET WATCH FILE/CLASS=MAJOR
to get an idea of which files are being accessed the point of the
error.
It may be that there are some other dictionary files which are in
CDD$TOP or something like that.
Showing the logicals is the only way I know to verify the settings so
you seem to be set in that regard.
Hang in there,
don
|
560.5 | A problem with system startup | KERNEL::WILESL | Louise Wiles, TP/IM Support | Mon Apr 27 1992 10:33 | 15 |
| Thanks for all the replies.
I've had a call from them to tell me that they had got a new system
startup which was starting the layered products up in batch rather than
running them sequentially. When they reverted to their original
sequential startup, things were all ok.
I've yet to have a look at their new 'improved' startup, but my feeling
is that if things are being started off in batch then one product may
not have completed its startup before another startup begins.
Thanks anyway,
Louise.
|