Re: [Snort-inline-users] ALERTFLUSHSTREAM: adjusted base_seq.

[Will M. <wil...@gm...>]

Ok, so fatal error is due having DEBUG enabled along with s4inline.=20
When debug is enabled there is a check that is done in bounds.h that
if SafeMemcpy fails, it kills your snort process.

The SafeMemcpy fails during reassembly due to the fact that you are
recieving in-window out of sequence packets.

We use the sequence number in each packet to determine where in memory
to copy the packet payload into our larger reassembled buffer.   If
the sequence number is off by 10000 or something, the buffer gets
corrupted and when we reset the base_seq number everything goes south
from there.

It is very hard to implement stream reassembly when you cannot verify
the validity of a packet within an established connection.

This is the issue that Victor and I are trying to resolve right now.=20
We have only noticed it on bridged connections. When running in NAT
mode packets are put in there proper order before they reach the QUEUE
target.

If you are running stream4inline+bridgemode, I would suggest disabling
until we get the new stream4inline written.

Regards,

Will

On 7/27/05, Adrian Soogemackelyk <soo...@gm...> wrote:
> Bridge mode.
>=20
> On 7/27/05, Will Metcalf <wil...@gm...> wrote:
> > Are you running in bridge or NAT mode?
> >
> > Regards,
> >
> > Will
> >
> > On 7/27/05, Adrian Soogemackelyk <soo...@gm...> wrote:
> > > I have a snort-inline box with the Snort-inline-2.3.0-RC1.diff patche=
d
> > > to it.  Recently, snort-inline runs a while and then it dies without
> > > any real indication why except I see about 60 "ALERTFLUSHSTREAM:
> > > adjusted base_seq" messages on the screen.  I looked at the code and
> > > this seems to be a stream4inline feature.  My preprocessor line in my
> > > conf file looks like this:
> > >
> > > preprocessor stream4: timeout 30, memcap 67108864,
> > > disable_evation_alerts, stream4linline, truncate
> > > preprocessor stream4_reassemble: clientonly, noalerts, ports 21 23 25
> > > 53 80 143 110 111
> > >
> > > Any clues what might be causing snort to die like that?
> > >
> > > -Adrian
> > >
> > >
> > > -------------------------------------------------------
> > > SF.Net email is Sponsored by the Better Software Conference & EXPO Se=
ptember
> > > 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices
> > > Agile & Plan-Driven Development * Managing Projects & Teams * Testing=
 & QA
> > > Security * Process Improvement & Measurement * http://www.sqe.com/bsc=
e5sf
> > > _______________________________________________
> > > Snort-inline-users mailing list
> > > Sno...@li...
> > > https://lists.sourceforge.net/lists/listinfo/snort-inline-users
> > >
> >
>=20
>=20
> -------------------------------------------------------
> SF.Net email is Sponsored by the Better Software Conference & EXPO Septem=
ber
> 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices
> Agile & Plan-Driven Development * Managing Projects & Teams * Testing & Q=
A
> Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf
> _______________________________________________
> Snort-inline-users mailing list
> Sno...@li...
> https://lists.sourceforge.net/lists/listinfo/snort-inline-users
>