Install the cloud dispatcher

The cloud dispatch service is an experimental service for running containers on cloud VMs. It eliminates the need for SLURM, Node Manager, and SLURM dispatcher. It works with Microsoft Azure and Amazon EC2; future versions will also support Google Compute Engine.

The cloud dispatch service can run on any node that can connect to the Arvados API service, the cloud provider’s API, and the SSH service on cloud VMs. It is not resource-intensive, so you can run it on the API server node.

Only one dispatch process should be running at a time. If you are migrating a system that currently runs crunch-dispatch-slurm, it is safest to remove the crunch-dispatch-slurm service entirely before installing arvados-dispatch-cloud.

~$ sudo systemctl --now disable crunch-dispatch-slurm
~$ sudo apt-get remove crunch-dispatch-slurm

Create a dispatcher token

If you haven’t already done so, create an Arvados superuser token to use as SystemRootToken in your cluster config file.

On the API server, use the following commands:

~$ cd /var/www/arvados-api/current
$ sudo -u webserver-user RAILS_ENV=production bundle exec script/create_superuser_token.rb
zzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzz

Create a private key

Generate an SSH private key with no passphrase. Save it in the cluster configuration file (see PrivateKey in the example below).

~$ ssh-keygen -N '' -f ~/.ssh/id_dispatcher
Generating public/private rsa key pair.
Your identification has been saved in /home/user/.ssh/id_dispatcher.
Your public key has been saved in /home/user/.ssh/id_dispatcher.pub.
The key fingerprint is:
[...]
~$ cat ~/.ssh/id_dispatcher
-----BEGIN RSA PRIVATE KEY-----
MIIEpQIBAAKCAQEAqXoCzcOBkFQ7w4dvXf9B++1ctgZRqEbgRYL3SstuMV4oawks
ttUuxJycDdsPmeYcHsKo8vsEZpN6iYsX6ZZzhkO5nEayUTU8sBjmg1ZCTo4QqKXr
...
oFyAjVoexx0RBcH6BveTfQtJKbktP1qBO4mXo2dP0cacuZEtlAqW9Eb06Pvaw/D9
foktmqOY8MyctzFgXBpGTxPliGjqo8OkrOyQP2g+FL7v+Km31Xs61P8=
-----END RSA PRIVATE KEY-----

You can delete the key files after you have copied the private key to your configuration file.

~$ rm ~/.ssh/id_dispatcher ~/.ssh/id_dispatcher.pub

Configure the dispatcher

Add or update the following portions of your cluster configuration file, /etc/arvados/config.yml. Refer to config.defaults.yml for information about additional configuration options.

Clusters:
  uuid_prefix:
    ManagementToken: xyzzy
    SystemRootToken: zzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzz
    Services:
      Controller:
        ExternalURL: "https://uuid_prefix.arvadosapi.com"
      DispatchCloud:
        InternalURLs:
          "http://localhost:9006": {}
    Containers:
      CloudVMs:
        # BootProbeCommand is a shell command that succeeds when an instance is ready for service
        BootProbeCommand: "sudo systemctl status docker"

        # --- driver-specific configuration goes here --- see Amazon and Azure examples below ---

      DispatchPrivateKey: |
        -----BEGIN RSA PRIVATE KEY-----
        MIIEpQIBAAKCAQEAqXoCzcOBkFQ7w4dvXf9B++1ctgZRqEbgRYL3SstuMV4oawks
        ttUuxJycDdsPmeYcHsKo8vsEZpN6iYsX6ZZzhkO5nEayUTU8sBjmg1ZCTo4QqKXr
        FJ+amZ7oYMDof6QEdwl6KNDfIddL+NfBCLQTVInOAaNss7GRrxLTuTV7HcRaIUUI
        jYg0Ibg8ZZTzQxCvFXXnjseTgmOcTv7CuuGdt91OVdoq8czG/w8TwOhymEb7mQlt
        lXuucwQvYgfoUgcnTgpJr7j+hafp75g2wlPozp8gJ6WQ2yBWcfqL2aw7m7Ll88Nd
        [...]
        oFyAjVoexx0RBcH6BveTfQtJKbktP1qBO4mXo2dP0cacuZEtlAqW9Eb06Pvaw/D9
        foktmqOY8MyctzFgXBpGTxPliGjqo8OkrOyQP2g+FL7v+Km31Xs61P8=
        -----END RSA PRIVATE KEY-----
    InstanceTypes:
      x1md:
        ProviderType: x1.medium
        VCPUs: 8
        RAM: 64GiB
        IncludedScratch: 64GB
        Price: 0.62
      x1lg:
        ProviderType: x1.large
        VCPUs: 16
        RAM: 128GiB
        IncludedScratch: 128GB
        Price: 1.23

Minimal configuration example for Amazon EC2:

Clusters:
  uuid_prefix:
    Containers:
      CloudVMs:
        ImageID: ami-01234567890abcdef
        Driver: ec2
        DriverParameters:
          AccessKeyID: EALMF21BJC7MKNF9FVVR
          SecretAccessKey: yKJAPmoCQOMtYWzEUQ1tKTyrocTcbH60CRvGP3pM
          SecurityGroupIDs:
          - sg-0123abcd
          SubnetID: subnet-0123abcd
          Region: us-east-1
          EBSVolumeType: gp2
          AdminUsername: debian

Minimal configuration example for Azure:

Clusters:
  uuid_prefix:
    Containers:
      CloudVMs:
        ImageID: "https://zzzzzzzz.blob.core.windows.net/system/Microsoft.Compute/Images/images/zzzzz-compute-osDisk.55555555-5555-5555-5555-555555555555.vhd"
        Driver: azure
        DriverParameters:
          SubscriptionID: XXXXXXXX-XXXX-XXXX-XXXX-XXXXXXXXXXXX
          ClientID: XXXXXXXX-XXXX-XXXX-XXXX-XXXXXXXXXXXX
          ClientSecret: 2WyXt0XFbEtutnf2hp528t6Wk9S5bOHWkRaaWwavKQo=
          TenantID: XXXXXXXX-XXXX-XXXX-XXXX-XXXXXXXXXXXX
          CloudEnvironment: AzurePublicCloud
          ResourceGroup: zzzzz
          Location: centralus
          Network: zzzzz
          Subnet: zzzzz-subnet-private
          StorageAccount: example
          BlobContainer: vhds
          DeleteDanglingResourcesAfter: 20s
          AdminUsername: arvados

Test your configuration

First, add the appropriate package repository for your distribution.

Next, install the arvados-server package.

On Red Hat-based systems:

~$ sudo yum install arvados-server

On Debian-based systems:

~$ sudo apt-get install arvados-server

Run the cloudtest tool to verify that your configuration works. This creates a new cloud VM, confirms that it boots correctly and accepts your configured SSH private key, and shuts it down.

~$ arvados-server cloudtest && echo "OK!"

Refer to the cloudtest tool documentation for more information.

Install the dispatcher

On Red Hat-based systems:

~$ sudo yum install arvados-dispatch-cloud
~$ sudo systemctl enable arvados-dispatch-cloud

On Debian-based systems:

~$ sudo apt-get install arvados-dispatch-cloud

Note:

The arvados-dispatch-cloud package includes configuration files for systemd. If you’re using a different init system, configure a service to start and stop an arvados-dispatch-cloud process as desired.

Verify the dispatcher is running

Use your ManagementToken to test the dispatcher’s metrics endpoint.

~$ token="xyzzy"
~$ curl -H "Authorization: Bearer $token" http://localhost:9006/metrics
# HELP arvados_dispatchcloud_containers_running Number of containers reported running by cloud VMs.
# TYPE arvados_dispatchcloud_containers_running gauge
arvados_dispatchcloud_containers_running 0
[...]

Previous: Sample compute node ping script

The content of this documentation is licensed under the Creative Commons Attribution-Share Alike 3.0 United States licence.
Code samples in this documentation are licensed under the Apache License, Version 2.0.