[Search for users] [Overall Top Noters] [List of all Conferences] [Download this site]

Conference csc32::consolemanager

Title:POLYCENTER Console Manager
Notice:Kits, Scans, Docs on CSC32:: as PCM$KITS:,PCM$DOCS:, PCM$SCANS:
Moderator:CSC32::BUTTERWORTH
Created:Thu Aug 06 1992
Last Modified:Fri Jun 06 1997
Last Successful Update:Fri Jun 06 1997
Number of topics:1541
Total number of notes:6564

690.0. "/dev/null permissions" by MUNDIS::MBACHMANN (MAIK BACHMANN @MFR) Mon Apr 24 1995 14:15

Hi there,



DCR150
OSF3.0B on Sable machine


When starting console by /sbin/init.d/consoled start
permissions of /dev/null are changed so that world cannot write onto
it any longer.
This is done by one of the background processes ( consoled or ENS ).

Workaround:

   vi /sbin/init.d/consoled

   goto end

   sleep 5
   chmod 666 /dev/null



Anyone seen this problem - an " official " patch or ....


Thanks  -  Maik
T.RTitleUserPersonal
Name
DateLines
690.1Have seen "very secure" /dev/null too !EEMELI::SYVANENTero Syv�nen MCS @FNO 879-4567Thu May 11 1995 22:1114

	>>Anyone seen this problem - an " official " patch or ....

        A customer of mine was also hit by this bug. I enabled audit
        subsystem to log fchmod() and chmod() calls and it really turned
        out that consoled was calling fchmod() and changed protection of
        /dev/null to 754 !

        I'm also *very* interested of an official patch.

        Regards

        		Tero
690.2PBUNIX::pgbAnd through the square window...Fri May 12 1995 10:5512
Hi,

  Nowhere in the code do we do a chmod to 754 of any files, we do however chmod
  to 777, 774 and 644 in various places. All our chmods are to do with log and
  archive files (socket files too on OSF/1).

  Do you happen to have /dev/null specified as a log or archive directory path for
  any of your systems?

Cheers,
Phil

690.3See it here too...KAOOSC::boivinMoi, j'viens du nord!Fri May 12 1995 13:339
Good day,

I see the problem here too. I initially blamed DECinspect but the
problem is still occuring after changing Inspect's check of /dev/null. 

I'm also interested in finding out who the culprit is (PCM or other)
and getting a patch...

Ed
690.4no redirect to /dev/nullMUNDIS::MBACHMANNMAIK BACHMANN @MFRTue May 16 1995 08:2637
Hi Phil,



sorry, but I do not redirect logfile to /dev/null - even this
should not touch permissions on /dev/null.


This is very urgent - since we go into acceptance in 3 weeks
and cannot live with this behaviour. We think about cancelling PCM
in our product if we do not get a patch for it.
And it is definitely PCM who does changing the permissions.

Try this:

    while true
    do
       ls -alt /dev/null
       sleep 2
    done


    In the meantime call /sbin/init.d/consoled start
    be sure the permissions of /dev/null are correct before starting.



If you can help , please do it - because it will cost us a lot of points
in aceptance if we can't get rid of this.
We will QAR this as well.

Feel free to contact me directly:
unix mail : bachmann�epa001.enet.dec.com
            


Thanks for your help - Maik
690.5YOSSAM::SIMONSimon Jackson @reo 830 X3879Tue May 16 1995 13:286
Hi,
    reproduced it!!!

I am working on a fix at the moment, will keep you updated.

Cheers Simon...
690.6YOSSAM::SIMONSimon Jackson @reo 830 X3879Tue May 16 1995 15:555
Maik,
     the problem is fixed in the latest V1.6 OSF kit. It has just been put on
the net see note 2.LAST. Please try it out and confirm your problem is fixed.

Cheers Simon...
690.7Thats real good news - thanksMUNDIS::MBACHMANNMAIK BACHMANN @MFRWed May 17 1995 07:389
Thanks a lot for the quick solution,

I will try it here in Munich.

When is the next "official" release planned ( layered products CDROM )?????



Thanks - Maik
690.8ASAPZEDAR::simonSimon Jackson 830 x3879Wed May 17 1995 10:379
Maik,
     because this is really a maintainence release (plus a few new
bits) we are planning to submit when the major problems which the
kit is meant to fix are validated and the bugs reported are fixed. 

We hope this will be in 2-3 weeks. Obviously then is is down to
when the CDROMS are generated for the platforms.

Cheers Simon...