So guys.
I simply just want to do this simple thing: log all logs from pods within a GKE cluster into elastic search, so that i can analyse it in Kibana. we intend to build very advanced and specific, context based monitoring and management for our product, but first things first: all logs must be consolidated into some elasticsearch repo.
This is very similar to what happens by default when you enable logging when creating a cluster from google developer console, the only difference is that the logs are consolidated in Google Cloud Logging, which is NOT what we want.
A lot of documentation every where within Kubernetes.io and GitHub, but honestly i've become more confused that clear.
I'm tempted to just run `kubectl create -f ` on the 4 files i can see on https://github.com/kubernetes/kubernetes/tree/master/cluster/addons/fluentd-elasticsearch But it looks like a hack to me... how would i be sure the fluent-d logging agents would be automatically deployed on each container/pod with that approach? how would i be sure things will work just as fine as it is with the Google Cloud Logging setup that comes out of the box? what of when my cluster scales to more that one node? would things still work fine?
Thanks, any insight would be greatly appreciated!