system/3808: select() does not timeout correctly

[rogerw@pccorp.com]

>Number:         3808
>Category:       system
>Synopsis:       select() does not timeout correctly
>Confidential:   yes
>Severity:       serious
>Priority:       low
>Responsible:    bugs
>State:          open
>Quarter:        
>Keywords:       
>Date-Required:
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Thu Jun 03 20:30:02 GMT 2004
>Closed-Date:
>Last-Modified:
>Originator:     Roger Waterhouse
>Release:        3.4
>Organization:
net
>Environment:
	
	System      : OpenBSD 3.4
	Architecture: OpenBSD.i386
	Machine     : i386
>Description:
	I am using the postfix MTA (compiled form source and it is timeing out
        too quickly when a reverse DNS lookup fails. It timesout in approx 13
        seconds instead of the specified 300 seconds. I posted this on the 
        postfix list and the author stated "Postfix implements read/write 
        timeouts with select(). You may want to capture a system call trace to
        see why select() returns in less than 300 seconds."  
>How-To-Repeat:
	This is very difficult to reproduce as the only time it shows up is when
        a MTA that has a broken Reverse DNS (looping or some other problem)
        attempts to connect. This is pretty rare (5 or 6 times in the last
        three months on a system that accepts approx 7k messages/day).
        System was built from cvs 3.4 about 3 months ago and hasn't been updated        since, postfix version is 2.0.18. 
>Fix:
	Get everybody on the internet to fix their DNS. Yeah right <grin>. 
        Actually I can get away with putting entries in /etc/hosts. This is
        a bit reactive, but works.


>Release-Note:
>Audit-Trail:
>Unformatted: