Skip to content

Latest commit

 

History

History
197 lines (163 loc) · 7.37 KB

getting_started_hetzner.md

File metadata and controls

197 lines (163 loc) · 7.37 KB
Raw

Getting started on Hetzner

clone and init submodules

git clone https://github.com/cornelius-keller/ansible-coroeos-kubernetes.gi
cd ansible-coroeos-kubernetes/
git submodule init
Submodule 'library' (https://github.com/cornelius-keller/ansible-ovh) registered for path 'library'
Submodule 'roles/ansible-coreos-bootstrap' (https://github.com/sigma/ansible-coreos-bootstrap.git) registered for path 'roles/ansible-coreos-bootstrap'
git submodule init
Submodule 'library' (https://github.com/cornelius-keller/ansible-ovh) registered for path 'library'
Submodule 'roles/ansible-coreos-bootstrap' (https://github.com/sigma/ansible-coreos-bootstrap.git) registered for path 'roles/ansible-coreos-bootstrap'
git submodule update
Cloning into 'library'...
remote: Counting objects: 18, done.
remote: Total 18 (delta 0), reused 0 (delta 0), pack-reused 18
Unpacking objects: 100% (18/18), done.
Checking connectivity... done.
Submodule path 'library': checked out 'f87a217cef869a7c2ce69178441cb02299b321b2'
Cloning into 'roles/ansible-coreos-bootstrap'...
remote: Counting objects: 314, done.
remote: Total 314 (delta 0), reused 0 (delta 0), pack-reused 313
Receiving objects: 100% (314/314), 1.01 MiB | 1.03 MiB/s, done.
Resolving deltas: 100% (131/131), done.
Checking connectivity... done.
Submodule path 'roles/ansible-coreos-bootstrap': checked out '8cd508c21868babdf32a7e0ec078df77cdf49611'

upload ýour public key to Hetzner

  • Log into the the Hetzner robot.

  • Navigate to the server list in the left menu. hetzner server list

  • chose key management hetzner key management

  • chooose new key, give it a name, and copy the output of

    cat ~/.ssh/id_rsa.pub

into the form and save it. hetzner new key

  • copy the id of the key for use in the inventory. hetzer key list

create an inventory

  mkdir ../ansible-coreos-inventory 
  cp inventory-hetzner.ini.sample the ../ansible-coreos-inventory/inventory-hetzner.ini

  • change the ips of your servers to match your server ips.

  • copy the id of your public key you uploaded to hetzner into the inventory value rescue_authorized_key

  • create a robot account and copy the username and password into the inventory. hetzner robot account

    For Examble:

    hetzner_webservice_username=#ws+7FPjagF7
hetzner_webservice_password=<your password>

  • create a uuid and key for ceph

# uuidgen
3867334c-2a6d-4468-b970-c878f0d36fee
# python ceph-key.py
AQCBK2FYAAAAABAAxcEwUSSPc7Zt1VJ9fjYH8A==
  • add to the ceph fsid and key to the inventory
ceph_fsid=3867334c-2a6d-4468-b970-c878f0d36fee
ceph_key=AQCBK2FYAAAAABAAxcEwUSSPc7Zt1VJ9fjYH8A==

  • add the public keys to the inventory that should have accces to the cluster:

    `mkdir ../ansible-coreos-inventory/group_vars`

  • add the public keys that should have access to to the cluster to the file: ../ansible-coreos-inventory/group_vars../ansible-coreos-inventory/group_vars/all.yml For example:

coreos_public_keys:
- "ssh-rsa AAAAB3NzaC1yc2EAAAADAQABAAABAQCqj5FfbBok7q8qtHqj+g5aut4eT6y78QLtC7igcGriZnbEHX9aUfpOgue+Bq1GcIXp9uSOzp+R5OrP0FzPTK9nDgm7R1wnp1zTFOF8LorcLH7ii/9p793O/bvcGNc6OaPGwzIA0naI9pumyIiArEbsnUJlWqGWchHmAm+3McN3QVxTQY6/+aW1Dt5dnC5rbvgB1lfOOhxfr19ED1zV1qgeFKkHptJ1llIkyNLyXNSiMYNuuC2pFn5F3w+Nfe+hRq8gRsJRnDuwcVibNcfR3egZ8sYyHhCWdR0HM1ZLfzW/ens37rahNEkCJrrZUYGbZ3sSDjkWoEOFk/rcdge+detV jck@jck"

Your inventory shuld now look like this:

 [etcd-node]
 176.9.123.237 coreos_hostname=core01
 176.9.122.117 coreos_hostname=core02
 176.9.118.230 coreos_hostname=core03
 [etcd-proxy]


 [coreos:children]
 etcd-node
 etcd-proxy

 [kubernetes-master]
 176.9.123.237
 176.9.122.117
 [kubernetes-node]
 176.9.118.230
 [kubernetes:children]
 kubernetes-master
 kubernetes-node
 [ceph-mon]
 176.9.123.237
 176.9.122.117
 176.9.118.230
 [ceph-osd]
 176.9.123.237
 176.9.122.117
 176.9.118.230

 [all:vars]
 #kube_master_dns_name=kubemaster.bettertest.de
 hyperkube_aci=quay.io/cornelius/hyperkube
 hyperkube_aci_tag=v1.4.4_coreos.0
 kube_master_ip=176.9.123.237
 kube_cluster_name=privat
 kubernetes_version=1.4.4
 k8s_service_ip_range=10.100.0.0/16
 k8s_service_ip=10.100.0.1
 k8s_dns_service_ip=10.100.0.10
 k8s_dns_domain=cluster.local
 hetzner_webservice_username=#ws+7FPjagF7
 hetzner_webservice_password=justforthedemo
 baremetal_provider=hetzner
 kubectl_checksum=1ffbcfdec9961093b5fb1eec0b65f37af71a825dafff733de90d6ed6db647729
 ansible_python_interpreter=/home/core/pypy/bin/python
 ceph_fsid=3867334c-2a6d-4468-b970-c878f0d36fee
 ceph_key=AQCBK2FYAAAAABAAxcEwUSSPc7Zt1VJ9fjYH8A==
 ceph_osd_dir=/home/core/data/ceph/osd
 ceph_osd_type=osd_disk
 ceph_osd_device=/dev/sdb
 rescue_authorized_key=21:74:2f:a4:08:8d:7c:83:bc:b4:3a:8a:5a:bd:3e:5f

bootsrap your coreos cluster

# ansible-playbook -i ../ansible-coreos-inventory/inventory-hetzner.ini bootstrap_coreos.yml

You should see an output like this:

asciicast

Check if the bootstrap was successfull:

# kubectl get nodes
NAME            STATUS    AGE
176.9.118.230   Ready     9m
176.9.122.117   Ready     9m
176.9.123.237   Ready     10m

To ceck if the ceph cluster is healthy you can log into one of the nodes and issue ceph -s

ssh core@176.9.118.230
Last login: Mon Dec 26 17:07:36 UTC 2016 from 84.191.220.116 on pts/0
CoreOS stable (1185.5.0)
Update Strategy: No Reboots
Failed Units: 1
locksmithd.service
core@core03 ~ $ ceph -s
cluster 3867334c-2a6d-4468-b970-c878f0d36fee
 health HEALTH_WARN
        80 pgs degraded
        80 pgs stuck unclean
        80 pgs undersized
        recovery 40/60 objects degraded (66.667%)
 monmap e1: 3 mons at {ceph-mon-j8h1q=10.1.43.6:6789/0,ceph-mon-pbv8y=10.1.3.3:6789/0,ceph-mon-uvril=10.1.56.3:6789/0}
        election epoch 6, quorum 0,1,2 ceph-mon-pbv8y,ceph-mon-j8h1q,ceph-mon-uvril
  fsmap e5: 1/1/1 up {0=mds-ceph-mds-2484866350-hu4o7=up:active}
 osdmap e15: 2 osds: 2 up, 2 in
        flags sortbitwise
  pgmap v32: 80 pgs, 3 pools, 2068 bytes data, 20 objects
        78568 kB used, 5576 GB / 5576 GB avail
        40/60 objects degraded (66.667%)
              80 active+undersized+degraded

If the cluster is unhealthy try again after a few minutes unil you see someting like this:

core@core03 ~ $ ceph -s
cluster 3867334c-2a6d-4468-b970-c878f0d36fee
health HEALTH_OK
monmap e1: 3 mons at {ceph-mon-j8h1q=10.1.43.6:6789/0,ceph-mon-pbv8y=10.1.3.3:6789/0,ceph-mon-uvril=10.1.56.3:6789/0}
   election epoch 6, quorum 0,1,2 ceph-mon-pbv8y,ceph-mon-j8h1q,ceph-mon-uvril
 fsmap e5: 1/1/1 up {0=mds-ceph-mds-2484866350-hu4o7=up:active}
osdmap e20: 3 osds: 3 up, 3 in
   flags sortbitwise
pgmap v50: 80 pgs, 3 pools, 2068 bytes data, 20 objects
   116 MB used, 8364 GB / 8364 GB avail
         80 active+clean

Cluster check asciicast:

asciicast