Ted Cooper wrote:
*snip*
I've had two single domain machines obliterated by sender callouts and a
joe job. Being a small operator on the end of a multi-million (billion?)
email callout bomb is a case of throwing up your arms in surrender and
simply shutting down the server until the attack subsides.
'Obliterated'?
You must have one Helluva good backbone for things to degenerate to that
sad sate of affairs.
;-)
Our previous VIA C6 1.5 GHz single-core hit saturation on the 100 Mbps
uplink long before they ran out of sufficient resources to carry on
regardless, plus respond to ssh so as to 'renice' Exim, ID the threat
source(s) while on the march, then LBL or FW them.
Our newer AMD-64 E350 1.6 GHz dual-cores won't even breath all that hard
at uplink saturation, so I don't [ have to ] worry about it at all in
anything close to 'real time'.
And I'm agnostic - at least as to what other folks choose to run.
Linux - IF that is what you run - should fare not one byte worse than a
*BSD on smtp load carrying ability, and you need at least a saturated
GigE to stress a *BSD box. If even then.... so...
Consider throttling your system. The I/O, not Exim. Less overhead.
You'll still have to deal with Joe jobs - but no longer on an emergency
basis.
Bill
--
韓家標
--
## List details at https://lists.exim.org/mailman/listinfo/exim-users
## Exim details at http://www.exim.org/
## Please use the Wiki with this list - http://wiki.exim.org/
