Latency when Pods communicate to coredns
search cancel

Latency when Pods communicate to coredns

book

Article ID: 336979

calendar_today

Updated On:

Products

VMware NSX

Issue/Introduction

This article can be used to find out the cause how latency happened

Symptoms:
Packets latency seen when pods communicate to coredns

Environment

VMware NSX-T Data Center 2.5.x
VMware NSX-T Data Center

Cause

1: Get Src Pod id
kubectl get pod

2: Get the tag (Privated vlan) that Pod is using
ovs-vsctl show

3: Get the switchport that Pod is using
net-stats -l

4: Do packet capture on this switchport
pktcap-uw --switchport --dir 2 -o pod.pcap

5: Example Packet analysis
The Src Pod did not send AAAA DNS query only A DNS query send the first time.
and after 5 seconds, it sent A and AAAA DNS query.

Resolution

Issue is not caused by nsx network

Workaround:
Add options in Pod resolv.conf
echo 'options single-request-reopen' >>/etc/resolv.conf

Additional Information

Impact/Risks:
5 Seconds latency when pods accessing coredns