# SxxxPxx 1123 UAT Testing

1. Gather node / pod sysctl config

sysctl net.netfilter.nf_conntrack_tcp_be_liberal

2. Scale App down to 1, modify the node sysctl setting

On node
sysctl -w net.netfilter.nf_conntrack_tcp_be_liberal=1

Restart pod, check sysctl config inside pod again.

## check config inside pod

oc debug node/<node_name>
NAME=<pod-name>
NAMESPACE=<pod-namespace>
pod_id=$(chroot /host crictl pods --namespace ${NAMESPACE} --name ${NAME} -q)
ns_path="/host/$(chroot /host bash -c "crictl inspectp $pod_id | jq '.info.runtimeSpec.linux.namespaces[]|select(.type==\"network\").path' -r")"
nsenter $nsenter_parameters -- sysctl net.netfilter.nf_conntrack_tcp_be_liberal

## Check tcp timeout

sh-5.1# cat /proc/sys/net/ipv4/tcp_keepalive_time
7200


net.ipv4.tcp_keepalive_time = 7200
net.ipv4.tcp_keepalive_probes = 9
net.ipv4.tcp_keepalive_intvl = 75

sysctl net.ipv4.tcp_keepalive_time
sysctl net.ipv4.tcp_keepalive_probes
sysctl net.ipv4.tcp_keepalive_intvl

Results in the following behavior:

Trigger a keepalive probe every 7200 seconds (2 hours) after the last received packet
最後一次收到TCP封包後隔7200秒傳送tcp keepalive probe

If the connection is idle and the remote host does not reply, probe the remote host every 75 seconds
如果remote host無回應，且連線處於閒置狀態,每隔75秒傳送一次probe

Close the connection after 9 failed probes
9次probe失敗後關閉連線

https://access.redhat.com/solutions/19029