Re: [exim] Thanks, Re: Limiting incoming connections on a …

Página superior
Eliminar este mensaje
Responder a este mensaje
Autor: W B Hacker
Fecha:  
A: exim users
Asunto: Re: [exim] Thanks, Re: Limiting incoming connections on a per-domain basis
Steve Sobol wrote:

>
>
> --On Sunday, September 24, 2006 1:34 PM +0800 W B Hacker
> <wbh@???> wrote:
>
>
>>> The only real downside to this solution was having to list the entire
>>> /27 worth of IP addresses one at a time, but I can live with that
>>> shortcoming. On the other hand, if anyone has figured out how to use
>>> CIDR notation in MySQL hostlist queries, let me know and I'll send you
>>> a virtual beer!
>>
>>
>> Should be intuitive if you are using the 'network' data types (INET,
>> CIDR,
>> MACADDR) for storage and retreival.
>
>
> I don't think those types are available in MySQL. But I haven't looked,
> and I'm using an older version of the software anyhow.
>


Age has its virtues - pre-DB2, pre-SQL PostgreSQL's age, not mine ..

;-)

I did look - after posting - and the 'network' data types are NOT listed for MySQL.

OTOH, 'license wars' aside, there is no longer much downside to MySQl -> PG
migration, and 'transactional' awareness, better ACID complaince, etc. come
along w/o need of kludogng in an InnoDb back-end, etc.

On modern hardware, either MySQL or PG will be (more than) 'fast enough', and PG
takes MySQL tables perhaps a bit more readily than the reverse, plus being more
'standards' compliant w/r SQL syntax.

- the special data types and other extensions, of course, being among the
exceptions, but useful, nonetheless.

Except for a few well-documented syntax differences, seldom used in Exim SQL
setups, one doesn't even have to change Exim's lookup calls from pg_sql to
my_sql or the reverse.

Another way - as we used to have to do in RIM/R:Base, was to 'emulate' a missing
data type by subterfuge.

In this case, one could perhaps store the IP *not* in dotted notation, but in
binary or hex (an easy external conversion) in two fields of a record, an upper
bound and a lower bound.

Where the upper and lower bounds were identical, a single IP would be matched,
where there was a range between them, a range.

But my roots are showing...

'forth definitions binary...'

;-)

Bill