Performance issues with intertwining Kubernetes networks
I am creating a Kubernetes cluster (v1.6.1) on AWS with one master and two slaves, then I start a mysql instance with helm and deploy a simple Django web app that queries the last five rows from the database and displays it. For my web -service I specify "type: LoadBalancer", which creates the ELB on AWS.
If I use weave networks and scale my web application to at least two replicas, then I start to experience inconsistent response times - most of the time this is reasonable (eg 0.1-0.2 sec), but 20-40 % requests take much longer (3-5 s, sometimes even more than 15 s). However, if I switch to a "flannel" network, everything works quickly, even with 20-30 replicas of the web application. All machines have enough resources, so there won't be a problem.
I tried debugging to find out what is causing the delay, and the best explanation I have is that AWS ELB does not work well with intertwine. Has anyone faced similar problems? What could be the problem? Please let me know if I have to provide relevant information.
PS I am new to using Kubernetes.
source to share
No one has answered this question yet
Check out similar questions: