On Fri, Sep 18, 2009 at 4:01 PM, Jonah Horowitz <[email protected]> wrote: > I fixed the nf_contrack problem with this (really just the first one, > but the others were good too). > > HAProxy sysctl changes > > For network tuning, add the following to /etc/sysctl.conf: > > net.ipv4.netfilter.ip_conntrack_max = 16777216 thanks, mine is actually in two othe places under proc. i'd checked it and thought 150000 was fine, but this is def a bigger number.
still see maxconn at 10000, i can't figure out what is holding this down so low. ( i understand that nf_conntrack wasn't going to fix it, just puzzled.) > net.ipv4.tcp_max_tw_buckets = 16777216 > > increase TCP max buffer size setable using setsockopt() > > net.core.rmem_max = 16777216 > net.core.wmem_max = 16777216 > > increase Linux autotuning TCP buffer limits min, default, and max number > of bytes to use set max to at least 4MB, or higher if you use very high > BDP paths > > net.ipv4.tcp_rmem = 4096 87380 16777216 > net.ipv4.tcp_wmem = 4096 65536 16777216 > > -jonah > > -----Original Message----- > From: David Birdsong [mailto:[email protected]] > Sent: Friday, September 18, 2009 3:06 PM > To: haproxy > Subject: artificial maxconn imposed > > I've set ulimit -n 20000 > > maxconn in defaults is 16384 and still somehow when i check the stats > page,maxconn is limited to 10000, sure enough requests start piling > up. > > any suggestions on where else to look? i'm sure it's an OS thing, so: > > Fedora 10 x86_64 16GB of RAM > > this command doesn't turn anything up > find /proc/sys/net/ipv4 -type f -exec cat {} \; | grep 10000 > > > (also dmesg shows nf_conntrack: table full, dropping packet.) which i > think is another problem. might be time to switch to a *BSD. > >
