Super High Latency From Host to Host
-
We are seeing crazy high latency (25-45MS) pinging from our PFSENSE gateway to any other host (and vice versa) on the same layer-2 domain (same switch). The machine has 16 cores (AMD) 32 GB ram and a dual 10gb NIC. The total traffic going through the NIC is less than 1 gigabit/s. We have NAT enabled and have about 1500 users going through the box. Between any other two hosts on the same switch the ping is < 0.9ms
Any suggestions?
Some information from the box:
State Table Total Rate
current entries 220113
searches 26480915673 182833.8/s
inserts 454504314 3138.1/s
removals 454284201 3136.5/s
ounters
match 602771956 4160.0/s
bad-offset 0 0.0/s
fragment 1535799 10.6/s
short 4776 0.0/s
normalize 21462 0.1/s
memory 80657911 556.7/s
bad-timestamp 0 0.0/s
congestion 0 0.0/s
ip-option 27088 0.2/s
proto-cksum 68710 0.5/s
state-mismatch 1629077 11.2/s
state-insert 378 0.0/s
state-limit 0 0.0/s
src-limit 0 0.0/s
synproxy 0 0.0/s
divert 0 0.0/sload averages: 0.55, 0.64, 0.54 up 1+16:23:01 10:52:54
238 processes: 18 running, 147 sleeping, 73 waitingMem: 383M Active, 29M Inact, 807M Wired, 132K Cache, 66M Buf, 30G Free
Swap: 32G Total, 32G Free -
Same behaviour on both LAN and WAN? What if you ping the WAN IP from the LAN interface on the firewall? And is it the same for a client on the LAN pinging IPs on the WAN subnet?
-
We are seeing crazy high latency (25-45MS) pinging from our PFSENSE gateway to any other host (and vice versa) on the same layer-2 domain (same switch). The machine has 16 cores (AMD) 32 GB ram and a dual 10gb NIC. The total traffic going through the NIC is less than 1 gigabit/s. We have NAT enabled and have about 1500 users going through the box. Between any other two hosts on the same switch the ping is < 0.9ms
Any suggestions?
Did your particular hardware configuration ever work well in the past ?