Instructions:
In order to successfully run the Vagrantfile your laptop will need the following:
- 4GiB of RAM
- 8GiB of RAM will give better performance
- Change the
memorysetting in theVagrantfile
- 20GiB of free disk space
- VirtualBox 5.1.28 Download
- Vagrant 2.0.0 Download
- Operating Systems
- Mac OSx 10.12.6
- Windows 10
- Windows 7
- Ubuntu 16.04 & 16.10
Once you have VirtualBox and Vagrant installed, clone this repo and use the Vagrantfile to launch your instance of IBM Cloud Private.
$ git clone https://github.com/IBM/deploy-ibm-cloud-private.git
$ cd deploy-ibm-cloud-privateinstall: vagrant up
stop: vagrant halt
start: vagrant up
uninstall: vagrant destroy
login to master node: vagrant ssh
suspend: vagrant suspend
resume: vagrant resume
The included Vagrantfile defines a single Ubuntu 16.04 VM. On that VM we install LXD - a pure-container hypervisor that runs unmodified Linux guest operating systems with VM-style operations at incredible speed and density. Utilizing LXD we can create a multi-node cluster all running on a single VM and all accessible from the terminal on your laptop. This makes for a great platform for development, test, learning excersizes, and demos.
Once the Vagrantfile has installed and configured LXD and installed all of the IBM Cloud Private community edition prereqs on the VM and all LXD containers defined within, it will begin the process of installing IBM Cloud Private community edition. This process should take about 20-30 mins to complete. Once done you will be able to access the IBM Cloud Private community edition web console here: https://192.168.27.100:8443 (assuming you did not have to change the base_segment value in the Vagrantfile see Conflicting Network Segments section below for details)
This Vagrantfile will stand up a single VirtualBox VM and deploy the master and proxy nodes onto it. It will configure and start two additional lxc containers within the VirtualBox VM and install the worker nodes onto them (cfc-worker1 & cfc-worker2). The cpu and memory allocated in the Vagrantfile will be shared by the VM and the lxc containers running the worker nodes within it. You can check the status of the lxc instances by ssh'ing into the VM using vagrant ssh and then running the command lxc list:
| NAME | STATE | IPV4 | IPV6 | TYPE | SNAPSHOTS |
|---|---|---|---|---|---|
| cfc-worker1 | RUNNING | 192.168.27.101 (eth0) 172.17.0.1 (docker0) 10.1.213.128 (tunl0) | PERSISTENT | 0 | |
| cfc-worker2 | RUNNING | 192.168.27.102 (eth0) 172.17.0.1 (docker0) 10.1.54.64 (tunl0) | PERSISTENT | 0 |
If you see the addresses 192.168.56.101 or 192.168.56.102 for either cfc-worker1 or cfc-worker2 that means there was a conflicting network segment for the 192.168.27.x network on your system. You will need to change the base_segment value in the Vagrantfile to a value that will not overlap any existing segments on your machine. See the comments in the Vagrantfile for examples
Because we are using lxc containers to run the cfc-worker1 and cfc-worker2 nodes you will see the IBM Cloud Private community edition Dashboard report 3x as much memory availble in your cluster than you allocated via the memory setting in the Vagrantfile. Don't worry, this is normal. LXD is sharing the total memory available to the VirtualBox VM with each lxc instance so the amount of memory you see being reported in the IBM Cloud Private community edition console is a result of each node reporting that it has the amount of memory allocated to the VirtualBox host in the memory setting. The best way to think about how lxc instances share host resources is to think of lxc instances as being applications that run on your laptop where each lxc instance is an application and the host where the lxc instances are running is the laptop. Each lxc instance can request memory or cpu from the host as needed and return those resources when they are no longer needed just like running multiple applications on your laptop does.
The first thing you need to do is to clone this repo down and set up to use softlayer cli and ansible:
$ git clone https://github.com/IBM/deploy-ibm-cloud-private.git
$ cd deploy-ibm-cloud-private
$ sudo pip install -r requirements.txtNext you need to prepare the softlayer CLI with your credentials:
$ slcli config setup
Username: XXXXXX
API Key or Password: XXXXXXXXXXXXXXXXXXXXX
Endpoint (public|private|custom) [public]:
Timeout [0]:
:..............:..................................................................:
: name : value :
:..............:..................................................................:
: Username : XXXXXX :
: API Key : XXXXXXXXXXXXXXXXXXXXX :
: Endpoint URL : https://api.softlayer.com/xmlrpc/v3.1/ :
: Timeout : not set :
:..............:..................................................................:
Are you sure you want to write settings to "/home/XXXX/.softlayer"? [Y/n]: YCreate a SSH key to use:
$ ssh-keygen -f cluster/ssh_key -P ""
Generating public/private rsa key pair.
Overwrite (y/n)? y
Your identification has been saved in cluster/ssh_key
Your public key has been saved in cluster/ssh_key.pub
$ slcli sshkey add -f cluster/ssh_key.pub icp-key
$ slcli sshkey list
:........:....................:.................................................:.......:
: id : label : fingerprint : notes :
:........:....................:.................................................:.......:
: 926125 : icp-key : 2c:ba:f4:8d:0c:4f:35:3d:fb:98:0b:30:b3:2e:a4:09 : - :
:........:....................:.................................................:.......:Set the
sl_ssh_keyvariable inclusters/config.yamlwith the id from the output of theslcli sshkey listcommand.
Pick a datacenter and a VLAN to deploy to:
slcli vlan list
:.........:........:......:..........:............:..........:.................:............:
: id : number : name : firewall : datacenter : hardware : virtual_servers : public_ips :
:.........:........:......:..........:............:..........:.................:............:
: 2073387 : 866 : - : No : dal09 : 0 : 1 : 61 :
: 2073385 : 812 : - : No : dal09 : 0 : 1 : 13 :
Set the
sl_datacenterandsl_vlanvariables inclusters/config.yamlusing the datacenter name (ex dal09) and vlan id (ex 2073387). Unless you're very familiar with your softlayer account you may need to use the softlayer web portal to pick a VLAN to use. If you skip this step SL will pick a random vlan and you may not be able to communicate on the backend network.
Using ansible playbooks we've prepared you can create the VMs necessary to deploy ICP:
While the playbook runs fast, it can take some time for the systems to actually come online and be ready.
$ ansible-playbook playbooks/create_sl_vms.yml
PLAY [create servers] ****************************************************************************************************************************************************************
TASK [Include cluster vars] **********************************************************************************************************************************************************
ok: [localhost]
TASK [create master] *****************************************************************************************************************************************************************
changed: [localhost] => (item=icp-master01)
TASK [create workers] ****************************************************************************************************************************************************************
changed: [localhost] => (item=icp-worker01)
changed: [localhost] => (item=icp-worker02)
PLAY RECAP ***************************************************************************************************************************************************************************
localhost : ok=3 changed=2 unreachable=0 failed=0 Once the systems are online you should be able to prepare them for the ICP install using the following Ansible playbook. This playbook will clean up some Softlayer quirks, install Docker and slipstream the softlayer dynamic inventory dependencies into the ICP installer.
$ ssh-add cluster/ssh_key
$ ansible-playbook -i hosts prepare_sl_vms.yml -u root
PLAY [ensure connectivity to all nodes] **********************************************************************************************************************************************
TASK [check if python is installed] **************************************************************************************************************************************************
ok: [169.46.198.209]
ok: [169.46.198.195]
ok: [169.46.198.216]
...
...
TASK [Run the following command to deploy IBM Cloud Private] *************************************************************************************************************************
ok: [169.46.198.212] => {
"msg": "ssh root@169.46.198.212 docker run -e SL_USERNAME=<SL_USERNAME> -e SL_API_KEY=<SL_API_KEY> -e LICENSE=accept --net=host --rm -t -v /root/cluster:/installer/cluster icp-on-sl install"
}
PLAY RECAP ***************************************************************************************************************************************************************************
169.46.198.195 : ok=8 changed=6 unreachable=0 failed=0
169.46.198.209 : ok=8 changed=6 unreachable=0 failed=0
169.46.198.216 : ok=8 changed=6 unreachable=0 failed=0 Once that playbook has finished running we can use our modified deployer to deploy ICP (the final line in the previous command will give you the command syntax, you just need to plug in your SL credentials)
$ cat ~/.softlayer
[softlayer]
username = XXXXX
api_key = YYYY
$ ssh root@169.46.198.216 docker run -e SL_USERNAME=XXXXX -e SL_API_KEY=YYYY -e LICENSE=accept --net=host \
--rm -t -v /root/cluster:/installer/cluster icp-on-sl install
...
...
PLAY RECAP *********************************************************************
169.46.198.XXX : ok=44 changed=22 unreachable=0 failed=0
169.46.198.YYY : ok=69 changed=39 unreachable=0 failed=0
169.46.198.ZZZ : ok=45 changed=22 unreachable=0 failed=0
POST DEPLOY MESSAGE ************************************************************
UI URL is https://169.46.198.XXX:8443 , default username/password is admin/admin
Access the URL using the username, password provided in last few lines of the ICP deployment
Click on admin on the top right hand corner of the screen to bring up a menu and select "Configure Client".
Copy and Paste the provided commands into a shell:
kubectl config set-cluster cfc --server=https://169.46.198.216:8001 --insecure-skip-tls-verify=true
kubectl config set-context cfc --cluster=cfc
kubectl config set-credentials user --token=eyJhbGciOiJSUzI1NiIsImtpZCI6IjY5NjI2ZDJkNjM2NjYzMmQ3MzY1NzI3NjY5NjM2NTJkNmI2NTc5NjkiLCJ0eXAiOiJKV1QifQ.eyJhdWQiOiJjZmMtc2VydmljZSIsImV4cCI6MTUwNTc5MTMwOCwiaWF0IjoxNTA1NzQ4MTA4LCJpc3MiOiJodHRwczovL21hc3Rlci5jZmM6ODQ0My9hY3MvYXBpL3YxL2F1dGgiLCJwcm9qZWN0cyI6WyJkZWZhdWx0Il0sInN1YiI6ImFkbWluIn0.wnjScyVuUUJyqt7AW4J6PrZ135dp8nWQTUmAqcjja15vaY4GQsu2lpZTn6AhmzQNWYfaYcLwc1ApcRBAB3h9oQlJOpSMI96U9XDYcMi1_i1AX8zN1EAFnklURK6TKPMy2OfNx8KOoWHTtRqBm-NDcla25aCpIrdHi7P9OPB2dcrHv7cYLBNwB6zWnzkM1EnRXYQIXDtKs1iX1K-A5Ph0Si3LIUg3LOjNML3Yn2D7vCWdItaGs86EE-2R2VVkYsLO19G09KwcLnhf5CmxxTjPDp2dOQjfwIFWbTmVQCFORtqj2Gt3X2EQFBwSru-e9M-fYcUiv6bqpd7WLufo-7q3bg
kubectl config set-context cfc --user=user --namespace=default
kubectl config use-context cfcCheck that you can run some basic commands against the cluster:
$ Client Version: version.Info{Major:"1", Minor:"6", GitVersion:"v1.6.4", GitCommit:"d6f433224538d4f9ca2f7ae19b252e6fcb66a3ae", GitTreeState:"clean", BuildDate:"2017-05-19T18:44:27Z", GoVersion:"go1.7.5", Compiler:"gc", Platform:"linux/amd64"}
Server Version: version.Info{Major:"1", Minor:"5", GitVersion:"v1.5.2", GitCommit:"08e099554f3c31f6e6f07b448ab3ed78d0520507", GitTreeState:"clean", BuildDate:"2017-02-05T08:03:16Z", GoVersion:"go1.7.5", Compiler:"gc", Platform:"linux/amd64"}
$ kubectl get nodes
NAME STATUS AGE VERSION
169.46.198.195 Ready 34m v1.5.2
169.46.198.209 Ready 34m v1.5.2
169.46.198.216 Ready,SchedulingDisabled 36m v1.5.2From here you should be able to interact with ICP via either the Web UI or the kubectl command.