Strange

[Russ Allbery]

Robert J Adams <radams at siscom.net> writes:

> I noticed that our Diablo feeder box is taking in around 50gig/day.. but
> it's feed to our inn box is about half of that (gig/day wise).. we are
> getting a lot of rejects...

None of the things you list below should account for anywhere close to
that volume.  They're cases where INN is more strict about what sorts of
garbage it's willing to accept than Diablo is.

> Jan 10 20:32:07.695 - news-xfer.siscom.net
> <uhve4.15598$eQ2.507557 at news1.rdc2.on.home.com> 437 Line 1111 includes null
> character

> Anyway to tell it store that anyway?

No, nor do you want to.  INN badly mangles articles with embedded nuls.
It may be worth at some point teaching it how to handle them correctly,
but we started rejecting them up front specifically because it wasn't
doing the right thing with them later.

> And about 2500 of these:

> Jan 10 10:57:45.109 ? news-xfer.siscom.net <5378.3YNS4D$aC7 at bbs.tku.edu.tw>
> 437 article includes CR without LF(1)

I'm somewhat surprised INN is actually rejecting these, although that
still isn't very many.

> Jan 10 07:16:36.165 - news-xfer.siscom.net <8EB745918Coplccom at 24.2.9.59> 437
> Whitespace in "Newsgroups" header -- "alt.binaries.sounds.mp3,
> alt.binaries.sounds.mp3.dance"

That's a blatant violation of RFC 1036, although it will become legal with
the next revision of the news standards so at some point INN will be
updated to accept them.  Again, though, there shouldn't be that many of
these, as a lot of servers will reject them.

-- 
Russ Allbery (rra at stanford.edu)         <URL:http://www.eyrie.org/~eagle/>