<div dir="ltr"><div class="gmail_extra"><br><div class="gmail_quote">On Fri, Mar 14, 2014 at 10:31 AM, thorkill <span dir="ltr"><<a href="mailto:trackid-ohcee4ve@grid.einherjar.de" target="_blank">trackid-ohcee4ve@grid.einherjar.de</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex"><div id=":1f4" class="" style="overflow:hidden">I my case it's "master" on FreeBSD 9.2 and a "slave" on FreeBSD 9.1. It<br>
seems that in some time around net.inet.tcp.keepidle (sysctl -a to<br>
check) when keepalive from OS kicks in the transsmission will be droped<br>
depending on your keepalive configuration. You should also see error<br>
messages in postgresql logs like:<br>
<br>
getsockopt(TCP_KEEPCNT) failed: Protocol not available<br>
<div class=""><br>
> Any ideas? This is so confusing because it is such an odd time interval<br>
> before it locks up. What's magical about 58 minutes?<br>
<br>
</div>My wild guess is:<br>
<br>
<a href="http://lists.freebsd.org/pipermail/freebsd-stable/2013-November/075781.html" target="_blank">http://lists.freebsd.org/pipermail/freebsd-stable/2013-November/075781.html</a><div style="width:16px;height:16px;display:inline-block">
</div><br>
<br>
<br></div></blockquote></div><br>Wow. These symptoms line up *exactly*.</div><div class="gmail_extra"><br></div><div class="gmail_extra">I notice the getsockopt error is logged a handful of times when slon starts and connects.</div>
<div class="gmail_extra"><br></div><div class="gmail_extra">I applied the patch from the above message thread and rebooted the kernel. When slon started, no getsockopt errors were reported. I'll let you know in an hour what happens.</div>
<div class="gmail_extra"><br></div><div class="gmail_extra">By the way, my net.inet.tcp.keepidle is 7200000, which should be 2 hours.</div></div>