|
> > And another thing: If a real server crashes ldirectord needs a long time
> > to discover that this server isn't responding. I assume it is due to the
> > timeout of the read request. Maybe a ping prior to the read request
> > would solve this problem...
I've found that reducing the value of tcp_syn_retry helps with this
problem as well.
I personally have solved this problem thusly:
echo 1 > /proc/sys/net/ipv4/tcp_syn_retries
It default is 10. On my machine(s) this results in a reduction from 100
seconds to 10 for a totally down server.
You might even want to go as low as zero -- in my case, I'm using direct
routing, and on my LAN, I can reasonably expect that I should not have to
resend a syn (and certainly not more than once) to a properly functioning
server.
Your mileage may vary, of course, but it's something to consider.
Thanks,
Kyle Sparger - Senior System Administrator
Dialtone Internet - Extremely Fast Web Systems
(954) 581-0097 - Voice (954) 581-7629 - Fax
ksparger@xxxxxxxxxxxxxxxxxxxx
http://www.dialtoneinternet.net
|
