samba 2.2.8/unexpected.c VMS suppressing storing of mis-dire

[John E. Malmberg]

In article <CF0913E9C3D53D4E94DE47FDE25C633D0532F40B at hermes.cofiroute.com>,
COLLOT Jean-Yves <jean-yves.collot at cofiroute.fr> writes:
> Actually, there is a reason.
>
> John Malmberg wrote:
>> The module unexpected.c has a VMS specific change to suppress storing
>> of mis-directed packets in the designated TDB.
>
> I don't really know why, but I have on my site quite a big number of those
> packets supposed to be kept temporarily in UNEXPECTED.TDB
>
> Since Dave Jones provided a new set of TDB routines, based upon RMS indexed
> files, such an activity of insert/delete records in that file provoked,
> after a few days or a few weeks, very bad performances, and high I/O
> activity on that file, because it became badly fragmented.
>
> To fix this, I had 2 choices:
> 1. Run every few days a procedure that should stop samba, convert
> UNEXPECTED.TDB and restart samba.
> 2. suppress the store/delete of the packets in UNEXPECTED.TDB
>
> It may not be the best choice, but as far as nobody here is interested in
> those unexpected packets anyway, and stopping/restarting samba is quite
> perturbing for the users, I decided to stop storing that information.
>
> I think it would have been better to add some specific VMS Global Parameter
> for enabling or disabling this feature, but I guess that at that time I was
> in some hurry, so I just disabled the storage and forgot about it since.

It would seem that you made the best decision available for getting the
job done on VMS.

This is valuable information to know for when I start working on 3.x and 4.x.

It may be that a more efficient method is needed in redirecting those packets,
or even better finding out why they were misdirected.

And also you seem to have observed no real problem with just dropping the
misdirected packets, so I wonder what the actual impact is, or with the feature
enabled, how reproducable it is.

-John
wb8tyw at qsl.net
Personal Opinion Only