[97] in bugtraq
Re: adjunct
daemon@ATHENA.MIT.EDU (Brad Powell)
Mon Oct 24 18:19:30 1994
Date: Mon, 24 Oct 94 11:43:11 PDT
From: Brad.Powell@EBay.Sun.COM (Brad Powell)
To: bugtraq@fc.net, hobbit@bronze.lcs.mit.edu
>From bugtraq-owner@fc.net Thu Oct 20 21:20:05 1994
>Date: Fri, 21 Oct 1994 00:02:18 -0400
>From: hobbit@bronze.lcs.mit.edu (*Hobbit*)
>Question about this stupidity:
>
> AUpwdauthd:##nobody:65533:65533::/:/bin/false
>
>Why did Sun do this? Back whenever, the adjunct mechanism worked just
>fine without having to have AUxxx entries in the files. What's the point?
>It really hosed me the day I went to "clean up" a 4.1.3 box and didn't
>yet know that these were necessary.
>
>_H*
>
It was a compatibility issue. See the README for the patch 100564-(07)
Since it was a "Jumbo" patch, and we didn't want one patch to be in
conflict with another (C2 origional and BSM)
===========================================================================
* The following pseudo-users must be added to /etc/passwd and *
* /etc/security/passwd.adjunct before changing any binaries *
* This is so the auditing of the rpc.pwdauthd and rpc.yppasswd can occur *
* These additions do not need to be done on NIS client machines since
* they will pick these changes up from the NIS master.
* *
* /etc/passwd additions: *
AUpwdauthd:##AUpwdauthd:10:10:::/bin/false
AUyppasswdd:##AUyppasswdd:11:10:::/bin/false *
*/etc/security/passwd.adjunct additions: *
AUpwdauthd:*::::: *
AUyppasswdd:*::::: *
===========================================================================
