Linkerd can be ran in High Availability or HA mode.
This feature is experimental and it’s only available in the edge release.
Here’s a short description of what
--ha does to the
memoryrequests to the linkerd control plane components.
Because it’s the control plane that requires the
ha config, you’ll need to use the
install command with the
linkerd install --ha | kubectl apply -f
You can also override the amount of controller replicas that you wish to run by passing in the
linkerd install --ha --controller-replicas=2 | kubectl apply -f