Re: slapcat kills slapadd

[Quanah Gibson-Mount <quanah@zimbra.com>]

--On November 18, 2009 5:50:59 PM +0100 Peter Mogensen <apm@mutex.dk> wrote:
> Yes...
> Knowing that I would have expected slapcat to fail/crash.

I think you're missing the point.  *slapadd* with -q expects no other tools 
to be touching its environment.  It makes perfect sense to me that it 
crashed.

> But I wouldn't expect slapadd to crash to - being the first to open the
> BDB environment.
> If any slapadd can be crashed by any user/shell executing slapcat (with
> write perms to the dir), then it can be difficult to protect an long
> running (hours) slapadd from being crashed by accident.

I'd suggest then not allowing random users to slapcat your directory.  And 
if you have processes that run slapcat, then set that up in a shell script 
with protections, like you touch a file called /etc/noslapcat and if that 
exists, then your process will exit without running slapcat.


--Quanah

--

Quanah Gibson-Mount
Principal Software Engineer
Zimbra, Inc
--------------------
Zimbra ::  the leader in open source messaging and collaboration