Re: [exim] smtp_accept_max_per_connection

Top Page
Delete this message
Reply to this message
Author: Bill Hacker
Date:  
To: exim-users
Subject: Re: [exim] smtp_accept_max_per_connection
Philip Hazel wrote:

> On Tue, 29 Nov 2005, Dean Brooks wrote:
>
>
>>I too wish there were at least 100 or maybe even 1000 variables
>>available for use instead of just 10. In large configurations, its
>>easy to hit the limit of 10 very quickly.
>
>
> I'll take that as a "yes" vote for the implied question that I've
> recently been asked about increasing the number by making it
> configurable at build time.


Sound.

> The default could also be upped, of course.
>


Not *too* much, methinks. (disclaimer - old Forth coder on the loose...)

Given that there can be several hundred simultaneous connections,
- each allocated ten acl_c(n) variables,

and that each connection could process multiple messages,
- each allocated ten acl_(n) variables,

and that the curent variables are not simple booleans, but capable of
holding potentially lengthy strings.......

Might it not be more economical to add a set of variables of a different
type?

i.e.:

'Flag':  acl_cf 0 thru 9 and acl_mf 0 thru 9 = boolean 1/0, yes/no, 
true/false     - all we need for many tests.


'Code': acl_cc 0 thru 9 and acl_mc 0 thru 9 = one-byte to four-byte 
'raw' numeric value.      - likewise, can carry scores, etc.


Should need less space, be more efficient to test/compare for many of
the situations where we are now 'wasting' a flexible variable-length
general-purpose variable, and (often) more complex code to test it

- and the name carries a built-in reminder of what it is and can/cannot
be used for.

JMHO,

Bill Hacker