fixup of config files for optimal performance

Scale-setup-fixups

1) sed -i 's/            net_obj = VirtualNetwork(net_name, project_obj, id_perms=id_perms)/&\n            net_obj.uuid = str(uuid.uuid4())/' /usr/lib/python2.7/dist-packages/vnc_openstack/neutron_plugin_db.py
    Repeat the same for SG neutron_to_vnc()
2) In contrail-api.conf add these to DEFAULT section
    subnet_allocator_write=cassandra
    subnet_allocator_read=cassandra
    list_optimization_enabled=True
3) Increase the FD limits in /etc/contrail/supervisor_openstack.conf (min_fds) and /etc/default/neutron-server (ulimit -n ),
    for rabbitmq-server: /etc/default/rabbitmq-server (ulimit -n)
4) other sysctl tweaks
    net.netfilter.nf_conntrack_max = 256000
    net.netfilter.nf_conntrack_tcp_timeout_time_wait = 30
    net.ipv4.tcp_syncookies = 1
    net.ipv4.tcp_tw_recycle = 1
    net.ipv4.tcp_tw_reuse = 1
    net.ipv4.tcp_fin_timeout = 30
    net.unix.max_dgram_qlen = 1000
    net.ipv4.tcp_keepalive_time = 5
    net.ipv4.tcp_keepalive_probes = 5
    net.ipv4.tcp_keepalive_intvl = 1
    
/usr/lib/python2.7/dist-packages/vnc_cfg_api_server/vnc_cfg_ifmap.py

    973 #            self.config_log(string_buf.getvalue(),
    974 #                level=SandeshLevel.SYS_ERR)
    975 #            trace_msg(trace, name='MessageBusNotifyTraceBuf',
    976 #                              sandesh=self._sandesh, error_msg=errmsg)