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

Conference bulova::decw_jan-89_to_nov-90

Title:DECWINDOWS 26-JAN-89 to 29-NOV-90
Notice:See 1639.0 for VMS V5.3 kit; 2043.0 for 5.4 IFT kit
Moderator:STAR::VATNE
Created:Mon Oct 30 1989
Last Modified:Mon Dec 31 1990
Last Successful Update:Fri Jun 06 1997
Number of topics:3726
Total number of notes:19516

1444.0. "Can not log back after PAUSE" by CSC32::R_SOMBERG () Thu Sep 14 1989 18:48

    'Strange' problem with V5.2/DW.
    It is in LAVc, was running fine with V5.1B, they upgrated to V5.2, no
    errors during the installation including the DW software.
    No error during boot.
    Can login with no problem.
    Does PAUSE from the session mamanger and can not login back - using the
    correct password.
    Can login from the net using the same password (which was ABCDEFG).
    WE tured on security alarm and restarted the server and did two tests:
    
    1 - PAUSE and typed a bad password - we recieved a security alarm
    message. - we typed again something in the password prompt this time
    (the second) NOTHING at all.
    
    	(RESTART the server)
    
    2 - PAUSE and type the correct password, security alarm message
    indicate that DECW$SESSION successfuly accessed SYSUAF.DAT with SYSPRV
    but the paused screen stayed on.
    
    Looking around on his system every thing looks 'normal' the
    DECW$SESSION checksum's correctly.
    The above we checked on 3100 with 16 mb - but also the other two nodes
    on the system can not pause their session an log back.
    
    ??
    
    	Thanks, Reuven CSC/CS.

T.RTitleUserPersonal
Name
DateLines
1444.1CSC32::R_SOMBERGThu Sep 14 1989 18:525
    opsss.. forgot:
    With test #2 after the first securiy alarm message ,we not getting any
    other message doesn't matter what we are typing on the password prompt
    as with test #1.

1444.2Break-in evasion?BOMBE::MOOREBaN CaSe_sEnSiTiVe iDeNtIfIeRs!Thu Sep 14 1989 19:103
    Sounds like break-in evasion.  Check for intrusion records against
    WSAn: devices.

1444.3Try checking GBLPAGESKRIS10::COXKristen Cox - Dallas ACT Sys MgrFri Sep 15 1989 14:035
This happened to me as well.  I did not have enough GBLPAGES.  I set MIN_GBLPAGES
to 45K and it corrected the problem on all of my  nodes that had it.

KLC

1444.4same results with 45,000CSC32::R_SOMBERGFri Sep 15 1989 15:012
    Tried changing GBLPAGES to 45,000 - same results.

1444.5Another shot in the dark?KRIS10::COXKristen Cox - Dallas ACT Sys MgrFri Sep 15 1989 16:3911
I'm a new DECwindows user myself, but seem to have had my share of problems.
Two other things that have caused me similar problems are:

(1) My LOGIN.COM had some things that DECwindows didn't like.  In fact, I think
it was my DECW$LOGIN.COM.  Anyways, if you rename both files you can easily
tell.

(2) Check any .LOG files in your directory for possible errors...

I can't think of anything else...

1444.6CSC32::R_SOMBERGFri Sep 15 1989 17:275
    With out tests, we renamed : sylogin.com & login.com - no
    decw$login.com and the decw$sylogin.com is the default.
    
    	Reuven.

1444.7Sounds like something in SMHYDRA::COARNobody move or I'll shoot my foot!Mon Sep 18 1989 11:559
From .1, it sounds as though the session manager has vaildated the password, but
hasn't proceeded to the point where it tears down the pause window.  Not getting
any further alarms sounds like the SM isn't processing the subsequent input.

From another process, what does SHOW SYSTEM, ANALYZE /SYSTEM, and SHOW PROCESS
/CONTINUOUS tell you about the session manager process?

#ken	:-)}

1444.8The solution.CSC32::R_SOMBERGFri Sep 22 1989 18:5820
    This is a short summary of what we found:
    
    - We steped through the PAUSE code of the SM at the point where
      it tries to validate the password by reading the UAF. At this point
      it fails with RMS-E-RSZ.
    - If failed the PAUSE code did not triggered any error message to the 
      user, simply returned back.
    - We had only one security alarm message because it was using $GETUAI
      with a special context parameter (using the privous context)
    - In V5.2 there was a change in the $GETUAI system service to check if
      record read from the uaf is at least 644 - if not RSZ error.
    
    - THE USER USED SYSUAF.DAT FILE LEFT FROM V3 OF VMS - HE NEVER DID THE 
      CONVERT IN V4.
    
    - We converted his uaf and it is working fine.
    
    
    	Reuven.