AlexisDucastel

RKE2 commands

Install

curl -sL https://get.rke2.io | sh
systemctl daemon-reload
systemctl start rke2-server

AlexisDucastel

SELECT 
  ROUND(base.mem_MB, 3) as base_mem_MB,
  ROUND(per_conn.mem_MB, 3) as per_conn_mem_MB,
  @@GLOBAL.max_connections as max_conn,
  ROUND(@@GLOBAL.max_connections * per_conn.mem_MB, 3) as max_conn_MB,
  ROUND(@@GLOBAL.max_connections * per_conn.mem_MB + base.mem_MB, 3) as total_MB
  FROM
  ( SELECT (@@GLOBAL.key_buffer_size + @@GLOBAL.query_cache_size + @@GLOBAL.tmp_table_size + @@GLOBAL.innodb_buffer_pool_size + @@GLOBAL.innodb_log_buffer_size)/1024/1024 as mem_MB) as base,
  ( SELECT (@@GLOBAL.sort_buffer_size + @@GLOBAL.read_buffer_size + @@GLOBAL.read_rnd_buffer_size + @@GLOBAL.join_buffer_size + @@GLOBAL.thread_stack + @@GLOBAL.binlog_cache_size) /1024/1024 as mem_MB) as per_conn \G

AlexisDucastel

kubectl get x -o yaml | yq e '.items[]|splitDoc' - | awk '/^  name:/{file=$NF".yaml"} !/^--/{s=s $0"\n"} /^---/{print s > (file); close(file); s=""}'

AlexisDucastel

#!/bin/bash

function pod_wait_for_status {
	while true
	do
		STATUS=$(kubectl get pod $1 2>/dev/null|tail -n 1| awk '{print $3}')
		debug 2 "Pod $1 is in state $STATUS (waiting for $2)"
		[ "$STATUS" = "$2" ] && break
		sleep 1
	done

AlexisDucastel

Docker(-compose) with shorewall

The shorewall documentation explains in http://shorewall.org/Docker.html how to configure shorewall for use with docker. The problem with the configuration is that it only allows connections from the host to the main bridge docker0. Connections to other networks on dynamically created bridges, with names starting by default with br-, is blocked. Instead of the recommended contents of /etc/shorewall/interfaces, use wild-card interface names as follows:

#ZONE	INTERFACE	OPTIONS
#dock	docker0		bridge     # disabled default recommendation
dock 	docker0		physical=docker+,routeback=1
dock 	br		physical=br-+,routeback=1

AlexisDucastel

Recover cluster.rkestate file from controlplane node

RKE

Run on controlplane node, uses any found hyperkube image

docker run --rm --net=host -v $(docker inspect kubelet --format '{{ range .Mounts }}{{ if eq .Destination "/etc/kubernetes" }}{{ .Source }}{{ end }}{{ end }}')/ssl:/etc/kubernetes/ssl:ro --entrypoint bash $(docker inspect $(docker images -q --filter=label=org.label-schema.vcs-url=https://github.com/rancher/hyperkube.git) --format='{{index .RepoTags 0}}' | tail -1) -c 'kubectl --kubeconfig /etc/kubernetes/ssl/kubecfg-kube-node.yaml -n kube-system get configmap full-cluster-state -o json | jq -r .data.\"full-cluster-state\" | jq -r .' > cluster.rkestate

AlexisDucastel

PLEG tester

A few commands to run to test what triggers PLEG.

Docker response time

When using Docker, all container statuses are compared and it needs to happen within 3 minutes. Else the following log will be shown:

AlexisDucastel

kind: PersistentVolumeClaim
apiVersion: v1
metadata:
  name: rancher-server
spec:
  accessModes:
  - ReadWriteOnce
  resources:
    requests:
      storage: 10Gi

AlexisDucastel

cat /proc/net/tcp| cut -c -36|grep '0A$'| cut -c 16-19| while read line; do echo $((16#$line)); done

AlexisDucastel

#!/bin/bash
COUNT=$1
tail -n$COUNT /home/ubuntu/metrics.log | \
	awk '{M+=$1;U+=$2;N+=$3;S+=$4;I+=$5;T+=$6} END {printf "%d\t=%.2f+%.2f+%.2f+%.2f+%.2f\n",M/NR,U/NR,N/NR,S/NR,I/NR,T/NR}'