[279] in Project_DB
Re: Needless disruption of Project DB last week by O&S.
daemon@ATHENA.MIT.EDU (Peter B. Kelley)
Mon Jan 12 16:54:29 1998
Date: Mon, 12 Jan 1998 16:53:54 -0500
To: Bill Cattey <wdc@MIT.EDU>
From: "Peter B. Kelley" <kelley@MIT.EDU>
Cc: rar@MIT.EDU, rferrara@MIT.EDU, o&s-facmgt@MIT.EDU, project-db@MIT.EDU
It was the '/usr/local/bin' directory and not '/usr/local' that we had a
problem with earlier and only on one machine. The error occurrred just with
the installation of a new release of Oracle.
--Pete
At 03:38 PM 1/12/98 -0500, Bill Cattey wrote:
>Excerpts from mail: 12-Jan-98 Re: Needless disruption of .. "Peter B.
>Kelley"@MIT.ED (2061*)
>
>
>> Bill,
>
>> I just double-checked all scripts that get run. In no case does any script
>> change the access on the '/usr/local' directory. We've run into this
>> problem before on '/usr/local/bin', and we explicitly check for '755' on
>> that directory just to make sure that access is set correctly.
>
>> --Pete
>
>
>Thank you for the quick response to my concern.
>
>How is it that you've run into a problem before of /usr/local having
>incorrect modes? This never happens with any of the E40 servers. Might
>there be a problem with personnel changing things and not keeping track
>of what they do? Or perhaps a script that changes the modes in a
>non-obvious way? These sorts of mysteries are disasters waiting to
>happen.
>
>One of the ways in which we try to head off such mysteries is to do
>audits of changes to scripts. Maybe sometime we should get your
>developers together with some folks from over here and do a code walk
>through or audit of the scripts you install on machines you take care of.
>
>It might be a fun bit of cross-polination of culture. And the code
>would probably benefit from the additional scrutiny.
>
>-wdc
>
