---
layout: default
navsection: installguide
-title: Prerequisites
+title: Planning and prerequisites
...
{% comment %}
Copyright (C) The Arvados Authors. All rights reserved.
SPDX-License-Identifier: CC-BY-SA-3.0
{% endcomment %}
-h2. Supported Cloud and HPC platforms
+Before attempting installation, you should begin by reviewing supported platforms, choosing backends for identity, storage, and scheduling, and decide how you will distribute Arvados services onto machines. You should also choose an Arvados Cluster ID, choose your hostnames, and aquire TLS certificates. It may be helpful to make notes as you go along using one of these worksheets: "New cluster checklist for AWS":new_cluster_checklist_AWS.xlsx - "New cluster checklist for Azure":new_cluster_checklist_Azure.xlsx - "New cluster checklist for on premises Slurm":new_cluster_checklist_slurm.xlsx
-Arvados can run in a variety of configurations. For compute scheduling, Arvados supports HPC clusters using @slurm@, and supports elastic cloud computing on AWS, Google and Azure. For storage, Arvados can store blocks on regular file systems such as ext4 or xfs, on network file systems such as GPFS, or object storage such as Azure blob storage, Amazon S3, and other object storage that supports the S3 API including Google Cloud Storage and Ceph.
+The installation guide describes how to set up a basic standalone Arvados instance. Additional configuration for features including "federation,":{{site.baseurl}}/admin/federation.html "collection versioning,":{{site.baseurl}}/admin/collection-versioning.html "managed properties,":{{site.baseurl}}/admin/collection-managed-properties.html and "storage classes":{{site.baseurl}}/admin/collection-managed-properties.html are described in the "Admin guide.":{{site.baseurl}}/admin
-h2. Hardware (or virtual machines)
+The Arvados storage subsystem is called "keep". The compute subsystem is called "crunch".
-This guide assumes you have seven systems available in the same network subnet:
+# "Supported GNU/Linux distributions":#supportedlinux
+# "Choosing which components to install":#components
+# "Identity provider":#identity
+# "Storage backend (Keep)":#storage
+# "Container compute scheduler (Crunch)":#scheduler
+# "Hardware or virtual machines":#machines
+# "Arvados Cluster ID":#clusterid
+# "DNS and TLS":#dnstls
-<div class="offset1">
-table(table table-bordered table-condensed).
-|_. Function|_. Number of nodes|
-|Arvados API, Crunch dispatcher, Git, Websockets and Workbench|1|
-|Arvados Compute node|1|
-|Arvados Keepproxy and Keep-web server|1|
-|Arvados Keepstore servers|2|
-|Arvados Shell server|1|
-|Arvados SSO server|1|
-</div>
-The number of Keepstore, shell and compute nodes listed above is a minimum. In a real production installation, you will likely run many more of each of those types of nodes. In such a scenario, you would probably also want to dedicate a node to the Workbench server and Crunch dispatcher, respectively. For performance reasons, you may want to run the database server on a separate node as well.
+h2(#supportedlinux). Supported GNU/Linux distributions
+
+{% include 'supportedlinux' %}
+
+h2(#components). Choosing which components to install
-h2. Supported GNU/Linux distributions
+Arvados consists of many components, some of which may be omitted (at the cost of reduced functionality.) It may also be helpful to review the "Arvados Architecture":{{site.baseurl}}/architecture to understand how these components interact.
table(table table-bordered table-condensed).
-|_. Distribution|_. State|_. Last supported version|
-|CentOS 7|Supported|Latest|
-|Debian 8 ("jessie")|Supported|Latest|
-|Debian 9 ("stretch")|Supported|Latest|
-|Ubuntu 14.04 ("trusty")|Supported|Latest|
-|Ubuntu 16.04 ("xenial")|Supported|Latest|
-|Ubuntu 18.04 ("bionic")|Supported|Latest|
-|Ubuntu 12.04 ("precise")|EOL|8ed7b6dd5d4df93a3f37096afe6d6f81c2a7ef6e (2017-05-03)|
-|Debian 7 ("wheezy")|EOL|997479d1408139e96ecdb42a60b4f727f814f6c9 (2016-12-28)|
-|CentOS 6 |EOL|997479d1408139e96ecdb42a60b4f727f814f6c9 (2016-12-28)|
+|\3=. *Core*|
+|"PostgreSQL database":install-postgresql.html |Stores data for the API server.|Required.|
+|"API server + Controller":install-api-server.html |Core Arvados logic for managing users, groups, collections, containers, and enforcing permissions.|Required.|
+|\3=. *Keep (storage)*|
+|"Keepstore":install-keepstore.html |Stores content-addressed blocks in a variety of backends (local filesystem, cloud object storage).|Required.|
+|"Keepproxy":install-keepproxy.html |Gateway service to access keep servers from external networks.|Required to be able to use arv-put, arv-get, or arv-mount outside the private Arvados network.|
+|"Keep-web":install-keep-web.html |Gateway service providing read/write HTTP and WebDAV support on top of Keep.|Required to access files from Workbench.|
+|"Keep-balance":install-keep-balance.html |Storage cluster maintenance daemon responsible for moving blocks to their optimal server location, adjusting block replication levels, and trashing unreferenced blocks.|Required to free deleted data from underlying storage, and to ensure proper replication and block distribution (including support for storage classes).|
+|\3=. *User interface*|
+|"Workbench2":install-workbench2-app.html |Primary graphical user interface for working with file collections and running containers.|Optional. Depends on API server, keep-web, websockets server.|
+|\3=. *Additional services*|
+|"Websockets server":install-ws.html |Event distribution server.|Required to view streaming container logs in Workbench.|
+|"Shell server":install-shell-server.html |Grant Arvados users access to Unix shell accounts on dedicated shell nodes.|Optional.|
+|\3=. *Crunch (running containers)*|
+|"arvados-dispatch-cloud":crunch2-cloud/install-dispatch-cloud.html |Run analysis workflows on cloud by allocating and freeing cloud VM instances on demand.|Optional|
+|"crunch-dispatch-slurm":crunch2-slurm/install-dispatch.html |Run analysis workflows distributed across a Slurm cluster.|Optional|
+|"crunch-dispatch-lsf":crunch2-lsf/install-dispatch.html |Run analysis workflows distributed across an LSF cluster.|Optional|
-h2(#repos). Arvados package repositories
+h2(#identity). Identity provider
-On any host where you install Arvados software, you'll need to set up an Arvados package repository. They're available for several popular distributions.
+Choose which backend you will use to authenticate users.
-h3. CentOS
+* Google login to authenticate users with a Google account.
+* OpenID Connect (OIDC) if you have Single-Sign-On (SSO) service that supports the OpenID Connect standard.
+* LDAP login to authenticate users by username/password using the LDAP protocol, supported by many services such as OpenLDAP and Active Directory.
+* PAM login to authenticate users by username/password according to the PAM configuration on the controller node.
-Packages are available for CentOS 7. To install them with yum, save this configuration block in @/etc/yum.repos.d/arvados.repo@:
+h2(#postgresql). PostgreSQL
-<notextile>
-<pre><code>[arvados]
-name=Arvados
-baseurl=http://rpm.arvados.org/CentOS/$releasever/os/$basearch/
-gpgcheck=1
-gpgkey=http://rpm.arvados.org/CentOS/RPM-GPG-KEY-curoverse
-</code></pre>
-</notextile>
+Arvados works well with a standalone PostgreSQL installation. When deploying on AWS, Aurora RDS also works but Aurora Serverless is not recommended.
-{% include 'install_redhat_key' %}
+h2(#storage). Storage backend
-h3. Debian and Ubuntu
+Choose which backend you will use for storing and retrieving content-addressed Keep blocks.
-Packages are available for Debian 8 ("jessie"), Debian 9 ("stretch"), Ubuntu 14.04 ("trusty"), Ubuntu 16.04 ("xenial") and Ubuntu 18.04 ("bionic").
+* File systems storage, such as ext4 or xfs, or network file systems such as GPFS or Lustre
+* Amazon S3, or other object storage that supports the S3 API including Google Cloud Storage and Ceph.
+* Azure blob storage
-First, register the Curoverse signing key in apt's database:
+You should also determine the desired replication factor for your data. A replication factor of 1 means only a single copy of a given data block is kept. With a conventional file system backend and a replication factor of 1, a hard drive failure is likely to lose data. For this reason the default replication factor is 2 (two copies are kept).
-{% include 'install_debian_key' %}
+A backend may have its own replication factor (such as durability guarantees of cloud buckets) and Arvados will take this into account when writing a new data block.
-Configure apt to retrieve packages from the Arvados package repository. This command depends on your OS vendor and version:
+h2(#scheduler). Container compute scheduler
-table(table table-bordered table-condensed).
-|_. OS version|_. Command|
-|Debian 8 ("jessie")|<notextile><code><span class="userinput">echo "deb http://apt.arvados.org/ jessie main" | sudo tee /etc/apt/sources.list.d/arvados.list</span></code></notextile>|
-|Debian 9 ("stretch")|<notextile><code><span class="userinput">echo "deb http://apt.arvados.org/ stretch main" | sudo tee /etc/apt/sources.list.d/arvados.list</span></code></notextile>|
-|Ubuntu 14.04 ("trusty")[1]|<notextile><code><span class="userinput">echo "deb http://apt.arvados.org/ trusty main" | sudo tee /etc/apt/sources.list.d/arvados.list</span></code></notextile>|
-|Ubuntu 16.04 ("xenial")[1]|<notextile><code><span class="userinput">echo "deb http://apt.arvados.org/ xenial main" | sudo tee /etc/apt/sources.list.d/arvados.list</span></code></notextile>|
-|Ubuntu 18.04 ("bionic")[1]|<notextile><code><span class="userinput">echo "deb http://apt.arvados.org/ bionic main" | sudo tee /etc/apt/sources.list.d/arvados.list</span></code></notextile>|
+Choose which backend you will use to schedule computation.
-{% include 'notebox_begin' %}
+* On AWS EC2 and Azure, you probably want to use @arvados-dispatch-cloud@ to manage the full lifecycle of cloud compute nodes: starting up nodes sized to the container request, executing containers on those nodes, and shutting nodes down when no longer needed.
+* For on-premises HPC clusters using "slurm":https://slurm.schedmd.com/ use @crunch-dispatch-slurm@ to execute containers with slurm job submissions.
+* For on-premises HPC clusters using "LSF":https://www.ibm.com/products/hpc-workload-management/ use @crunch-dispatch-lsf@ to execute containers with slurm job submissions.
+* For single node demos, use @crunch-dispatch-local@ to execute containers directly.
+
+h2(#machines). Hardware (or virtual machines)
+
+Choose how to allocate Arvados services to machines. We recommend that each machine start with a clean installation of a supported GNU/Linux distribution.
-fn1. Arvados packages for Ubuntu may depend on third-party packages in Ubuntu's "universe" repository. If you're installing on Ubuntu, make sure you have the universe sources uncommented in @/etc/apt/sources.list@.
+For a production installation, this is a reasonable starting point:
+<div class="offset1">
+table(table table-bordered table-condensed).
+|_. Function|_. Number of nodes|_. Recommended specs|
+|PostgreSQL database, Arvados API server, Arvados controller, Websockets, Container dispatcher|1|16+ GiB RAM, 4+ cores, fast disk for database|
+|Workbench, Keepproxy, Keep-web, Keep-balance|1|8 GiB RAM, 2+ cores|
+|Keepstore servers ^1^|2+|4 GiB RAM|
+|Compute worker nodes ^1^|0+ |Depends on workload; scaled dynamically in the cloud|
+|User shell nodes ^2^|0+|Depends on workload|
+</div>
+
+^1^ Should be scaled up as needed
+^2^ Refers to shell nodes managed by Arvados that provide ssh access for users to interact with Arvados at the command line. Optional.
+
+{% include 'notebox_begin' %}
+For a small demo installation, it is possible to run all the Arvados services on a single node. Special considerations for single-node installs will be noted in boxes like this.
{% include 'notebox_end' %}
-Retrieve the package list:
+h2(#clusterid). Arvados Cluster ID
-<notextile>
-<pre><code>~$ <span class="userinput">sudo apt-get update</span>
-</code></pre>
-</notextile>
+Each Arvados installation is identified by a cluster identifier, which is a unique 5-character lowercase alphanumeric string. There are 36 5 = 60466176 possible cluster identifiers.
-h2. A unique identifier
+* For automated test purposes, use “z****”
+* For experimental/local-only/private clusters that won’t ever be visible on the public Internet, use “x****”
+* For long-lived clusters, we recommend reserving a cluster id. Contact "info@curii.com":mailto:info@curii.com for more information.
-Each Arvados installation should have a globally unique identifier, which is a unique 5-character lowercase alphanumeric string. For testing purposes, here is one way to make a random 5-character string:
+Here is one way to make a random 5-character string:
<notextile>
<pre><code>~$ <span class="userinput">tr -dc 0-9a-z </dev/urandom | head -c5; echo</span>
</code></pre>
</notextile>
-You may also use a different method to pick the unique identifier. The unique identifier will be part of the hostname of the services in your Arvados cluster. The rest of this documentation will refer to it as your @uuid_prefix@.
-
-
-h2. SSL certificates
-
-There are six public-facing services that require an SSL certificate. If you do not have official SSL certificates, you can use self-signed certificates.
-
-{% include 'notebox_begin' %}
-
-Most Arvados clients and services will accept self-signed certificates when the @ARVADOS_API_HOST_INSECURE@ environment variable is set to @true@. However, web browsers generally do not make it easy for users to accept self-signed certificates from Web sites.
+You may also use a different method to pick the cluster identifier. The cluster identifier will be part of the hostname of the services in your Arvados cluster. The rest of this documentation will refer to it as your @ClusterID@. Whenever @ClusterID@ appears in a configuration example, replace it with your five-character cluster identifier.
-Users who log in through Workbench will visit at least three sites: the SSO server, the API server, and Workbench itself. When a browser visits each of these sites, it will warn the user if the site uses a self-signed certificate, and the user must accept it before continuing. This procedure usually only needs to be done once in a browser.
+h2(#dnstls). DNS entries and TLS certificates
-After that's done, Workbench includes JavaScript clients for other Arvados services. Users are usually not warned if these client connections are refused because the server uses a self-signed certificate, and it is especially difficult to accept those cerficiates:
+The following services are normally public-facing and require DNS entries and corresponding TLS certificates. Get certificates from your preferred TLS certificate provider. We recommend using "Let's Encrypt":https://letsencrypt.org/. You can run several services on the same node, but each distinct DNS name requires a valid, matching TLS certificate.
-* JavaScript connects to the Websockets server to provide incremental page updates and view logs from running jobs.
-* JavaScript connects to the API and Keepproxy servers to upload local files to collections.
-* JavaScript connects to the Keep-web server to download log files.
+This guide uses the following DNS name conventions. A later part of this guide will describe how to set up Nginx virtual hosts.
+It is possible to use custom DNS names for the Arvados services.
-In sum, Workbench will be much less pleasant to use in a cluster that uses self-signed certificates. You should avoid using self-signed certificates unless you plan to deploy a cluster without Workbench; you are deploying only to evaluate Arvados as an individual system administrator; or you can push configuration to users' browsers to trust your self-signed certificates.
+<div class="offset1">
+table(table table-bordered table-condensed).
+|_. Function|_. DNS name|
+|Arvados API|@ClusterID.example.com@|
+|Arvados Webshell|webshell.@ClusterID.example.com@|
+|Arvados Websockets endpoint|ws.@ClusterID.example.com@|
+|Arvados Workbench|workbench.@ClusterID.example.com@|
+|Arvados Workbench 2|workbench2.@ClusterID.example.com@|
+|Arvados Keepproxy server|keep.@ClusterID.example.com@|
+|Arvados Keep-web server|download.@ClusterID.example.com@
+_and_
+*.collections.@ClusterID.example.com@ _or_
+*<notextile>--</notextile>collections.@ClusterID.example.com@ _or_
+collections.@ClusterID.example.com@ (see the "keep-web install docs":install-keep-web.html)|
+</div>
-{% include 'notebox_end' %}
+Setting up Arvados is easiest when Wildcard TLS and wildcard DNS are available. It is also possible to set up Arvados without wildcard TLS and DNS, but not having a wildcard for @keep-web@ (i.e. not having *.collections.@ClusterID.example.com@) comes with a tradeoff: it will disable some features that allow users to view Arvados-hosted data in their browsers. More information on this tradeoff caused by the CORS rules applied by modern browsers is available in the "keep-web URL pattern guide":../api/keep-web-urls.html.
-By convention, we use the following hostname pattern:
+The table below lists the required TLS certificates and DNS names in each scenario.
<div class="offset1">
table(table table-bordered table-condensed).
-|_. Function|_. Hostname|
-|Arvados API|@uuid_prefix@.your.domain|
-|Arvados Git server|git.@uuid_prefix@.your.domain|
-|Arvados Keepproxy server|keep.@uuid_prefix@.your.domain|
-|Arvados Keep-web server|download.@uuid_prefix@.your.domain
-_and_
-*.collections.@uuid_prefix@.your.domain or
-*<notextile>--</notextile>collections.@uuid_prefix@.your.domain or
-collections.@uuid_prefix@.your.domain (see the "keep-web install docs":install-keep-web.html)|
-|Arvados SSO Server|auth.your.domain|
-|Arvados Websockets endpoint|ws.@uuid_prefix@.your.domain|
-|Arvados Workbench|workbench.@uuid_prefix@.your.domain|
+||_. Wildcard TLS and DNS available|_. Wildcard TLS available|_. Other|
+|TLS|@ClusterID.example.com@
+*.@ClusterID.example.com@
+*.collections.@ClusterID.example.com@|*.@ClusterID.example.com@
+@ClusterID.example.com@|@ClusterID.example.com@
+git.@ClusterID.example.com@
+webshell.@ClusterID.example.com@
+ws.@ClusterID.example.com@
+workbench.@ClusterID.example.com@
+workbench2.@ClusterID.example.com@
+keep.@ClusterID.example.com@
+download.@ClusterID.example.com@
+collections.@ClusterID.example.com@|
+|DNS|@ClusterID.example.com@
+git.@ClusterID.example.com@
+webshell.@ClusterID.example.com@
+ws.@ClusterID.example.com@
+workbench.@ClusterID.example.com@
+workbench2.@ClusterID.example.com@
+keep.@ClusterID.example.com@
+download.@ClusterID.example.com@
+*.collections.@ClusterID.example.com@|@ClusterID.example.com@
+git.@ClusterID.example.com@
+webshell.@ClusterID.example.com@
+ws.@ClusterID.example.com@
+workbench.@ClusterID.example.com@
+workbench2.@ClusterID.example.com@
+keep.@ClusterID.example.com@
+download.@ClusterID.example.com@
+collections.@ClusterID.example.com@|@ClusterID.example.com@
+git.@ClusterID.example.com@
+webshell.@ClusterID.example.com@
+ws.@ClusterID.example.com@
+workbench.@ClusterID.example.com@
+workbench2.@ClusterID.example.com@
+keep.@ClusterID.example.com@
+download.@ClusterID.example.com@
+collections.@ClusterID.example.com@|
</div>
+
+{% include 'notebox_begin' %}
+It is also possible to create your own certificate authority, issue server certificates, and install a custom root certificate in the browser. This is out of scope for this guide.
+{% include 'notebox_end' %}
projects / arvados.git / blobdiff