Whitelisted

[Ken Anderson]

Furnish, Trever G wrote:

>>-----Original Message-----
>>From: Ken Anderson [mailto:ka at PACIFIC.NET]
>>Sent: Monday, July 28, 2003 10:18 AM
>>To: MAILSCANNER at JISCMAIL.AC.UK
>>Subject: Re: Whitelisted
>>
>>2. Mount /var/spool/MailScanner/incoming as tmpfs
>
>
> Huh?  I am probably just at a disadvantage here because redhat doesn't ship
> a man page for mount that documents the tmpfs option, but I thought the
> incoming spool had to be on the same mount point as the outgoing spool or MS
> would gag.  Doesn't that mean that we would instead want to mount
> /var/spool/MailScanner, not only the incoming dir?  And besides, if tmpfs is
> good enough for incoming, why wouldn't we also want to use it for outgoing?

/var/spool/MailScanner/incoming is the only safe thing to mount in
memory. If the machine dies you don't lose anything. Julian explained
why this is the case.. I can't recall exactly how mailscanner handles
this though, sorry.

in /etc/fstab:

none /var/spool/MailScanner/incoming/ tmpfs defaults 0 0

>
>>4. Sendmail's MaxRecipients = x is also a major factor, since each
>>message is cloned x times. I believe the default is 100 or
>>128... That's
>>probably too high if you are splitting all incoming mail into 1 recip
>>per message!
>

In sendmail.cf MaxRecipients is the maximum number of recipients for a
message that sendmail will accept. Using queue groups, we break messages
up _after_ we accept them. If you set MaxRecipients=1, you'll be denying
all mail entering your system with more than one recip, you won't be
splitting it up - Senders will see "Too many recipients" errors. :-(

Ken


> Lost me here too - sorry if I'm showing extreme ignorance.  I'm getting
> ready to implement this, trying to understand thoroughly first.  I thought
> MaxRecipients was getting set to 1, meaning "any message with more than 1
> recipient, split into separate messages with at most 1 recipient."  So if I
> have maxrecipients set to 1 and a msg comes in with ten recipients, I've
> cloned it 9 times, correct, not 1 time (maxrecipient being 1)?
>
> -t.
>
>