docs: How to upgrade bootstrap kubelet?

Signed-off-by: Suraj Deshmukh <suraj@kinvolk.io>

docs/how-to-guides/upgrade-bootstrap-kubelet.md

@@ -0,0 +1,44 @@
+# How to upgrade bootstrap kubelet
+
+## Introduction
+
+Bootstrap kubelet is the kubelet running on the node as a systemd service. Unlike most components on the cluster, lokoctl cannot update it. So there are some manual steps involved in updating the bootstrap kubelet.
+
+## Steps
+
+Apply the following steps to all the nodes one at a time.
+
+### Drain the node
+
+```
+kubectl drain <node name>
+```
+
+### Find out the IP
+
+Find the IP of the node by visiting the Cloud provider dashboard.
+
+```
+ssh core@<IP Address>
+```
+
+### On the node
+
+Run the following commands. **NOTE**: Export the correct and latest Kubernetes version, before proceeding to other commands.
+
+```
+export latest_kube=<latest kubernetes version e.g. v1.18.0>
+sudo sed -i "s|$(grep -i kubelet_image_tag /etc/kubernetes/kubelet.env)|KUBELET_IMAGE_TAG=${latest_kube}|g" /etc/kubernetes/kubelet.env
+sudo systemctl restart kubelet
+sudo journalctl -fu kubelet
+```
+
+Look at the logs carefully, if Kubelet fails to restart and instructs to do something like deleting a file, then delete it and restart the node.
+
+```
+sudo reboot
+```
+
+## Caveats
+
+- If a node which is running storage workload like Rook Ceph, then verify that the Ceph cluster is in **`HEALTH_OK`** state. If it is in **any other state, do not proceed with upgrades**. When the cluster is in `HEALTH_OK` state a few minutes of downtime of OSDs will not be a problem for Ceph cluster, that happens during node reboot.