Sounds like recompiling with a larger number of file descriptors and then running with a bigger limit/ulimit might do the trick. In any case, the number of concurrent policyd file handles shouldn't be less than the number of postfix instances on any given machine.
On Fri, 7 Sep 2007, Artem Bokhan wrote: > >> Frankly, I think you'd have to have a pretty hefty amount of processing >> to run that many email connections plus policyd plus whatever else might >> be running on your mail server (amavisd, spamassassin, etc.) to sustain >> that kind of >> load. > I have two MXes with 4 cores of Xeon 5130 (2.00GHz) each with 8Gb of RAM. > Quite enough when all software is writen with C, so no bottlenecks like > with spamassassin. > The system is 50-70% iddleing, no disks bottlenecks. > >> Currently, we have two mail gateways and run spamassassin on a >> separate cluster of machines, and will >> likely expand to additional servers. The added advantage is that with >> enough servers, you can take one down for maintenance or one can fail >> without generally having a bottleneck in the throughput. >> ------------------------------------------------------------------------- This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now >> http://get.splunk.com/ _______________________________________________ policyd-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/policyd-users
