[mod.computers.vax] UIC corruption ?

A105@UWOCC1.BITNET (Brent Sterner) (02/04/86)

   I've got a strange problem.  I have an account with uic=[1,4] on
which I run daily midnight jobs (1 a day) and also am doing development
on.  Today I decided to PURGE test.com and it failed.  Investigation
led me to notice that 4 of the files had a "strange" uic.  The uic was
valid, but was not [1,4].  Further, the most recent midnight.log files
also had a strange uic, but not the same as the other strange uic.
   This is an 8600 in a cluster with an hsc (nothing else) and 3 ra81
disks.  I called TSC.  After a long walk-through (I am a beginner) it
was recommended that I guarantee future "corruption" not be simply due
to another uic having access to the directory.  So I set the directory
protection for the directory root to (o,g,w); only the system can play
there now.  (DEC: the TSC call number is C860204367)
   Further, I did a brief listing of the UAF file, and searched it for
non-normal entries.  Nothing unexpected showed up.
   As a beginner, a lot goes by this net without it registering.  But
I think I saw a similar complaint recently.  Did I?  Anyone seen or heard
of this before?  I'm new to VMS, but I've been around other systems long
enough to know that the LAST thing I need is a corrupted file system.
   The false UICs have been corrected to prevent future confusion;  I
don't know any way to verify the rest of my structure.


Brent Sterner
Computing & Communications Services
Natural Sciences Building
The University of Western Ontario
London, Ontario, Canada
N6A 5B7
Telephone (519)679-2151  (secretary)
                   2167  (direct)