Has anyone see this error in haproxy logs?
Somewhere between 65k and 70k sessions, haproxy starts spewing the error and things start dying from that point. CPU(s) get pegged at 100%, healthchecks start failing, etc.
Some relevant information:
-- incoming bandwith is no more than 50Mbps at 65k sessions
Config details:
maxconn 400000 mode http balance roundrobin option httpchk HEAD / HTTP/1.0 option forwardfor except 127.0.0.1 option httpclose
sysctl stuff:
net.ipv4.ip_local_port_range = 1024 65000 net.ipv4.tcp_max_syn_backlog = 400000 net.ipv4.tcp_synack_retries = 3
net.core.somaxconn = 400000 net.core.netdev_max_backlog = 400000 net.ipv4.tcp_keepalive_time = 1800 net.ipv4.tcp_max_tw_buckets = 1440000
thanks in advance,
--simon
Received on 2009/05/15 16:09
This archive was generated by hypermail 2.2.0 : 2009/05/15 16:15 CEST