Package: greylistd Version: 0.8.3 Single process is not able to handle high traffic.
When greylistd gets more queries (via socket) that it can handle (for example in exim.conf you put read timeout = 10s) it stops working at all. What happens is that it processes every request but it's not able to process it in 10 seconds so exim timeouts (and that happens for every request). Changing greylistd to fork where each child would process X number of queries would probably make things much better. I wonder if sqlite would be better/faster than using python disct for storing triplets? -- Arkadiusz MiĆkiewicz PLD/Linux Team arekm / maven.pl http://ftp.pld-linux.org/