[Mimedefang] Graphdefang 0.9 missing events?

[John Kirkland]

As far as an easy test goes...

try running 'file' against your SummaryDB.  Mine says:

SummaryDB.db:  Berkeley DB (Hash, version 8, native byte-order)

If you are having a problem with this... your's will say version 2 or
something like that.

Don't call me an expert on this, but I'm doing some pretty weird stuff
with storing nested hashes, and I suspect that the earlier versions of
DB_File don't support it.  You can try 2 things... just upgrade your
DB_File from the BerkeleyDB distribution.  If that doesn't do it, then do
the whole thing.

-John

On Sun, 2 Nov 2003, Mark wrote:

> > I'm suspecting one of the following:
> >
> > 1. You are having DBFile issues and only seeing the items from the most
> > recent time you ran graphdefang and not the ones in the SummaryDB file.
>
> That may be it. In fact, today I noticed that when I feed graphdefang a new
> /var/log/maillog, it seems to have completely forgotten about everything
> before! I was gonna write about that as soon as I got the missing events
> solved. :)
>
> Though I must say, that I have never had a problem with DB_File before. And
> I use it extensively to interface with a DRAC's "hash" database.
>
> > This is common on BSD systems. There's info in my faq about this at
> > www.bl.org/~jpk/graphdefang
>
> I saw that. But I do not have those palette errors described there.
>
> I could update my BerkeleyDB; but that would be a last resort measure; as
> this is integral to the system, and also used by sendmail et al. And I do
> not potentially want to break such a vital part of the system.
>
> Is there not an easy test I can do to test the functionality of the
> SummaryDB? I seems created alright. It is about 5MB (ca. two weeks worth of
> data).
>
> At any rate, I appreciate your help,
>
> - Mark
>