youtube image
From YouTube: Episode 24: etcd, kpng, nginx project updates

Description

Lets review a few of our favorite upstream networking projects this week. As always we'll be able to answer antrea questions and topics along the way as they come up!

There was a recent etcd bug https://github.com/etcd-io/etcd/pull/13854 which brought down high density clusters and caused folks like vmware and red hat to suggest folks not to upgrade to kubernetes 1.22. We'll look at how etcds embedded stress perf-check tool along with etcd endpoint status --cluster can be used to see the raft indices of nodes and confirm wether any of your nodes has hopelessly fallen behind the leader node. Also, ricardo will show us a new way to look at the problem of securing L7 loadbalancers (even though they need some high privileges on the networking side), using hand crafterd linux jails.

And of course, another KPNG project update, this time, about how to get started on the windows kernelspace proxy.