Skip to content

Latest commit

 

History

History
374 lines (260 loc) · 16.7 KB

README.md

File metadata and controls

374 lines (260 loc) · 16.7 KB

Cloudify Environment Setup

NOTE: This setup will create a Cloudify Manager and then configure it.

The setup will not configure an existing Cloudify Manager installation.

Note: Without bootstrap, deployment should take 5 minutes. With bootstrap, up to 40 minutes.

To ask a question or report an issue, please use github issues or visit the Cloudify users groups.

Alternative: Deploy your Cloudify Manager using an IaaS Variant.

Purpose

This blueprint sets up a reference environment for executing the Cloudify Examples. When you execute this blueprint, you will create all the networks, routes, security groups, NICs, VMs, plugins, and secrets that you will need to run the other Cloudify Examples.

Pre-requisites

Infrastructure

When you execute the blueprint, you will provision the following resources in your cloud:

  • Common Infrastructure

    • manager_key: An SSH key pair used to connect to the Cloudify Manager.
    • agent_key: An SSH key pair used by the Cloudify Manager to connect other VMs that it creates.
    • ManagerSetup: see ManagerSetup.

  • AWS Infrastructure:

    • subnet0_nic_eip0
    • elasticip_nat_gateway
    • vpc
    • internet_gateway
    • subnet0
    • subnet1
    • nat_gateway
    • cloudify_security_group
    • subnet0_nic0
    • route_table0
    • route_table1
    • cloudify_host_cloud_config
    • cloudify_host

  • Azure Infrastructure:

    • resource_group
    • storage_account
    • availability_set
    • subnet0_nic_ip
    • virtual_network
    • security_group
    • subnet0
    • subnet1
    • subnet0_nic_ip_configuration
    • subnet1_nic_ip_configuration
    • subnet0_nic
    • subnet1_nic
    • cloudify_host

  • Openstack Infrastructure:

    • external_network
    • network0_subnet_port_fip
    • network0
    • network1
    • network0_router
    • network0_subnet
    • network1_subnet
    • cloudify_security_group
    • network0_subnet_port
    • network1_subnet_port
    • cloudify_host_cloud_config
    • cloudify_host

  • GCP Infrastructure

    • cfys0ip0 - Static External IP
    • cfynetwork
    • cfynetwork_subnet0
    • cfynetwork_subnet1
    • cfycloudify_security_group
    • cfycloudify_security_group1
    • controller_key
    • cfycloudify_host - Cloudify Manager VM

Secrets

  • Common Secrets:
    • agent_key_private
    • agent_key_public

Bootstrap

  • AWS Secrets:

    • vpc_id: This is the ID of the vpc. The same vpc that your manager is attached to.
    • private_subnet_id: This is the ID of a subnet that does not have inbound internet access on the vpc. Outbound internet access is required to download the requirements. It must be on the same vpc designated by VPC_ID.
    • public_subnet_id: This is the ID of a subnet that does have internet access (inbound and outbound). It must be on the same vpc designated by VPC_ID.
    • availability_zone: The availability zone that you want your instances created in. This must be the same as your public_subnet_id and private_subnet_id.
    • ec2_region_endpoint: The AWS region endpint, such as ec2.us-east-1.amazonaws.com.
    • ec2_region_name: The AWS region name, such as ec2_region_name.
    • aws_secret_access_key: Your AWS Secret Access Key. See here for more info. This may not be provided as an environment variable. The string must be set as a secret.
    • aws_access_key_id: Your AWS Access Key ID. See here for more info. This may not be provided as an environment variable. The string must be set as a secret.

  • Azure Secrets:

    • location: See here.
    • mgr_virtual_network_name: This is the name of the virtual network that your manager is attached to.
    • mgr_subnet_name: The is the name of the subnet that your manager is attached to.
    • mgr_resource_group_name: This the resource group that your manager is sitting in.
    • client_secret: Your Azure Service Principal password. See here.
    • client_id: Your Azure Service Principal appId. See here.
    • tenant_id: Your Azure Service Principal tenant. See here.
    • subscription_id: Your Azure subscription ID.

  • Openstack Secrets:

    • external_network_name: This is the network on your Openstack that represents the internet gateway network.
    • public_network_name: An openstack network. (Inbound is expected, outbound is required.)
    • public_subnet_name: A subnet on the public network.
    • private_network_name: An openstack network. (Inbound is not expected, outbound is required.)
    • private_subnet_name: A subnet on the network. (Inbound is not expected, outbound is required.)
    • router_name: This is a router that is attached to your Subnets designated in the secrets public_subnet_name and private_subnet_name.
    • region: Your Keystone V2 region.
    • keystone_url: Your Keystone V2 auth URL.
    • keystone_tenant_name: Your Keystone V2 tenant name.
    • keystone_password: Your Keystone V2 password.
    • keystone_username:Your Keystone V2 username.

  • GCP Secrets

    • client_x509_cert_url
    • client_email
    • client_id
    • project_id
    • private_key_id
    • private_key
    • management_network_name
    • zone
    • region
    • management_subnetwork_name
    • private_subnetwork_name
    • ubuntu_trusty_image
    • centos_core_image
    • small_instance_type

Preparation

You must choose one of two installation paths. These are:

  1. Bootstrap Cloudify. This is available regardless of cloud provider/hosting provider/IaaS. Slower.
  2. Use a pre-baked Cloudify image. This is currently only available in Openstack and AWS. Faster.

Bootstrap

To execute bootstrap, add "bootstrap: True" as a single line to your "*.yaml" file. This is the default already in the sample input files.

Using a pre-baked image.

You will find a list of pre-bootstrapped images on Cloudify's Downloads page.

  • AWS: AMIs are listed in the "aws-blueprint.yaml" under the cloudify_ami input. You may also find links here. Also, change the "bootstrap: True" to "False" in your inputs file.
  • Openstack: Follow these instructions to upload the Openstack QCOW image to Openstack. You will also need to find the correct values for cloudify_image, centos_core_image, ubuntu_trusty_image, small_image_flavor, large_image_flavor. Ask your Openstack Admin for more info on these.
  • Azure & GCP: There is not currently a pre-bootstrapped image for Azure and GCP, so bootstrap is the only option.

Environment Installation Steps

Once you have decided on a Manager installation method, you can proceed on the path of execution:

  1. Install Cloudify CLI. Make sure that your CLI is using a local profile. (You must have executed cfy profiles use local in your shell.

  2. Download and extract this blueprint archive (link) to your current working directory. You will create an inputs yaml file. Examples are provided in the inputs directory already if you prefer.

  3. To install your environment's infrastructure, execute one of the example commands below, inserting your account credentials in the *.yaml file located in the inputs directory for your IaaS.

Note: This command should be run from the same directory in which you extracted the blueprint in the previous step. If it stops after installing plugins (some users have noted this issue), run the command again without '--install-plugins' to complete the setup.

For AWS run:

$ cfy install cloudify-environment-setup-latest/aws-blueprint.yaml -i cloudify-environment-setup-latest/inputs/aws.yaml --install-plugins --task-retries=30 --task-retry-interval=5

For Azure run:

$ cfy install cloudify-environment-setup-latest/azure-blueprint.yaml -i cloudify-environment-setup-latest/inputs/azure.yaml --install-plugins --task-retries=30 --task-retry-interval=5

For Openstack run:

$ cfy install cloudify-environment-setup-latest/openstack-blueprint.yaml -i cloudify-environment-setup-latest/inputs/openstack.yaml --install-plugins --task-retries=30 --task-retry-interval=5

For GCP run:

$ cfy install cloudify-environment-setup-latest/gcp-blueprint.yaml -i cloudify-environment-setup-latest/inputs/gcp.yaml --install-plugins --task-retries=30 --task-retry-interval=5

Manager Setup Steps

Manager Setup is the last phase of execution. A file called ./instructions.txt is created in your cwd. It documents follow-up steps to complete Manager configuration.

These are the Manager setup steps:

  1. Run this command to get the path to the instructions file:
$ cfy deployments outputs -b cloudify-environment-setup-latest/

The output should look similar to:

{
  "1-Instructions": "/path/to/cwd/instructions.txt",
  "2-Demo": "cfy install https://github.com/cloudify-examples/nodecellar-auto-scale-auto-heal-blueprint/archive/4.1.zip -b demo -n aws-blueprint.yaml"
}

IMPORTANT

If you chose the bootstrap: true option before execution, you will need to follow all of the steps documented in the ./instructions.txt file.

If you chose the bootstrap: false option before execution, you will only need to execute step 4.

  1. SSH into your Manager VM.
ssh -i ~/.ssh/cfy-manager-key-os [email protected]
  1. Install the Cloudify RPM on the Manager VM.
sudo rpm -i http://repository.cloudifysource.org/cloudify/4.1.0/ga-release/cloudify-enterprise-cli-4.1.rpm
  1. Bootstrap the Manager:
cfy bootstrap /opt/cfy/cloudify-manager-blueprints/simple-manager-blueprint.yaml -i public_ip=xxx.xxx.xxx.xxx -i private_ip=192.168.120.5 -i ssh_user=centos -i ssh_key_filename=/home/centos/.ssh/key.pem -i agents_user=ubuntu -i ignore_bootstrap_validations=false -i admin_username=admin -i admin_password=admin
  1. Exit your Manager VM and initialize your CLI profile.
cfy profiles use -s centos -k ~/.ssh/cfy-manager-key-os -u admin -p admin -t default_tenant xxx.xxx.xxx.xxx
  1. Create Secrets.
cfy secrets create ......
  1. Upload Plugins
cfy plugins upload ......

Your manager is now ready. Proceed to the example blueprints!

Try the Nodecellar Auto-scale Auto-heal.

Environment Teardown Steps

When you are ready to uninstall your environment, run:

$ cfy profiles use local
$ cfy uninstall --allow-custom-parameters -p ignore_failure=true --task-retries=30 --task-retry-interval=5 -b cloudify-environment-setup-latest

Troubleshooting

502 Bad Gateway

  • If cfy profiles use ... fails with this output, the service has started in error state. Try restarting the VM.
<head><title>502 Bad Gateway</title></head>
<body bgcolor="white">
<center><h1>502 Bad Gateway</h1></center>
<hr><center>nginx/1.8.0</center>
</body>
</html>

[Errno 61] Connection refused

  • If cfy profiles use ... fails with the following output, then the server is refusing your connection because of too many requests. Most likely this is an issue with your network. Expect this issue to come up a lot until your network service improves.
Attempting to connect...
HTTPConnectionPool(host='**.***.***.***', port=80): Max retries exceeded with url: /api/v3/provider/context (Caused by NewConnectionError('<urllib3.connection.HTTPConnection object at 0x10d9d0590>: Failed to establish a new connection: [Errno 61] Connection refused',))

Connection aborted, BadStatusLine

  • If cfy profiles use ... fails with the following output, check the internet connection.
Attempting to connect...
Can't use manager 34.226.3.116. ('Connection aborted.', BadStatusLine("''",))

InfluxDB wont start

  • If cfy bootstrap... will not progress beyond influxdb, there is an issue with accessing the InfluxDB service on that port. This can be either an issue with a security group rule, or the IP may not be the primary interface.
2017-07-19 06:03:35.359  CFY <manager> [influxdb_yrqzak.create] Task started 'fabric_plugin.tasks.run_script'
[10.239.1.41] out: /tmp2017-07-19 06:03:36.959  LOG <manager> [influxdb_yrqzak.create] INFO: Installing InfluxDB...
2017-07-19 06:03:37.048  LOG <manager> [influxdb_yrqzak.create] INFO: Checking whether SELinux in enforced...
2017-07-19 06:03:37.425  LOG <manager> [influxdb_yrqzak.create] INFO: Downloading resource influxdb_NOTICE.txt to /opt/cloudify/influxdb/resources/influxdb_NOTICE.txt
2017-07-19 06:03:39.387  LOG <manager> [influxdb_yrqzak.create] INFO: Checking whether /opt/cloudify/influxdb/resources/influxdb-0.8.8-1.x86_64.rpm is already installed...
2017-07-19 06:03:39.923  LOG <manager> [influxdb_yrqzak.create] INFO: yum installing /opt/cloudify/influxdb/resources/influxdb-0.8.8-1.x86_64.rpm...
2017-07-19 06:03:41.116  LOG <manager> [influxdb_yrqzak.create] INFO: Deploying InfluxDB configuration...
2017-07-19 06:03:41.200  LOG <manager> [influxdb_yrqzak.create] INFO: Deploying blueprint resource components/influxdb/config/config.toml to /opt/influxdb/shared/config.toml
2017-07-19 06:03:42.565  LOG <manager> [influxdb_yrqzak.create] INFO: Deploying blueprint resource components/influxdb/config/cloudify-influxdb to /etc/sysconfig/cloudify-influxdb
2017-07-19 06:03:43.561  LOG <manager> [influxdb_yrqzak.create] INFO: Deploying blueprint resource components/influxdb/config/cloudify-influxdb.service to /usr/lib/systemd/system/cloudify-influxdb.service
2017-07-19 06:03:45.063  LOG <manager> [influxdb_yrqzak.create] INFO: Deploying blueprint resource components/influxdb/config/logrotate to /etc/logrotate.d/influxdb
2017-07-19 06:03:46.568  LOG <manager> [influxdb_yrqzak.create] INFO: Waiting for 192.168.121.5:8086 to become available...
2017-07-19 06:05:53.842  LOG <manager> [influxdb_yrqzak.create] INFO: 192.168.121.5:8086 is not available yet, retrying... (1/24)
[10.239.1.41] out:
[10.239.1.41] out:
[10.239.1.41] out:

Invalid Block Device Mapping

  • If cfy install... results in the below error, update the cloudify_host_block_device_mapping section to '/dev/sda1' in your input file.
<Response><Errors><Error><Code>InvalidBlockDeviceMapping</Code><Message>Invalid device name /dev/sda</Message></Error></Errors><RequestID>c1419ad6-5a27-457d-a8b7-d70c40ac8093</RequestID></Response>

Instance type Availability Zone Mismatch

  • If cfy install ... aws-blueprint.yaml execution fails with this error code, AWS is currently restricting deployments in your availability zone. Trying changing the availability zone to one of those suggested in the error message.
<Response><Errors><Error><Code>Unsupported</Code><Message>Your requested instance type (t1.micro) is not supported in your requested Availability Zone (us-east-1e). Please retry your request by not specifying an Availability Zone or choosing us-east-1a, us-east-1b, us-east-1d, us-east-1c.</Message></Error></Errors><RequestID>c00ced89-5e1d-4bcc-aac1-a8eff6e19935</RequestID></Response>