3 navsection: installguide
4 title: Multi host Arvados
7 Copyright (C) The Arvados Authors. All rights reserved.
9 SPDX-License-Identifier: CC-BY-SA-3.0
12 # "Hosts preparation":#hosts_preparation
13 ## "Hosts setup using terraform (experimental)":#hosts_setup_using_terraform
14 ## "Create a compute image":#create_a_compute_image
15 # "Multi host install using the provision.sh script":#multi_host
16 # "Choose the desired configuration":#choose_configuration
17 ## "Multiple hosts / multiple hostnames":#multi_host_multi_hostnames
18 ## "Further customization of the installation (modifying the salt pillars and states)":#further_customization
19 # "Installation order":#installation_order
20 # "Run the provision.sh script":#run_provision_script
21 # "Initial user and login":#initial_user
22 # "Test the installed cluster running a simple workflow":#test_install
24 h2(#hosts_preparation). Hosts preparation
26 In order to run Arvados on a multi-host installation, there are a few requirements that your infrastructure has to fulfill.
28 These instructions explain how to setup a multi-host environment that is suitable for production use of Arvados.
30 We suggest distributing the Arvados components in the following way, creating at least 6 hosts:
38 ## arvados cloud dispatcher
46 # KEEPSTOREs (at least 2)
48 # SHELL node (optional):
51 Note that these hosts can be virtual machines in your infrastructure and they don't need to be physical machines.
53 h3(#hosts_setup_using_terraform). Hosts setup using terraform (experimental)
55 We added a few "terraform":https://terraform.io/ scripts (https://github.com/arvados/arvados/tree/master/tools/terraform) to let you create these instances easier.
56 Check "the Arvados terraform documentation":/doc/install/terraform.html for more details.
58 h2(#multi_host). Multi host install using the provision.sh script
60 {% if site.current_version %}
61 {% assign branchname = site.current_version | slice: 1, 5 | append: '-dev' %}
63 {% assign branchname = 'master' %}
66 This is a package-based installation method. Start with the @provision.sh@ script which is available by cloning the @{{ branchname }}@ branch from "https://git.arvados.org/arvados.git":https://git.arvados.org/arvados.git . The @provision.sh@ script and its supporting files can be found in the "arvados/tools/salt-install":https://git.arvados.org/arvados.git/tree/refs/heads/{{ branchname }}:/tools/salt-install directory in the Arvados git repository.
68 This procedure will install all the main Arvados components to get you up and running in a multi-host environment.
70 The @provision.sh@ script will help you deploy Arvados by preparing your environment to be able to run the installer, then running it. The actual installer is located at "arvados-formula":https://git.arvados.org/arvados-formula.git/tree/refs/heads/{{ branchname }} and will be cloned during the running of the @provision.sh@ script. The installer is built using "Saltstack":https://saltproject.io/ and @provision.sh@ performs the install using master-less mode.
72 After setting up a few variables in a config file (next step), you'll be ready to run it and get Arvados deployed.
74 h3(#create_a_compute_image). Create a compute image
76 In a multi-host installation, containers are dispatched in docker daemons running in the <i>compute instances</i>, which need some special setup. We provide a "compute image builder script":https://github.com/arvados/arvados/tree/master/tools/compute-images that you can use to build a template image following "these instructions":https://doc.arvados.org/main/install/crunch2-cloud/install-compute-node.html . Once you have that image created, you can use the image reference in the Arvados configuration in the next steps.
78 h2(#choose_configuration). Choose the desired configuration
80 For documentation's sake, we will use the cluster name <i>arva2</i> and the domain <i>arv.local</i>. If you don't change them as required in the next steps, installation won't proceed.
82 We will try to provide a few Arvados' multi host installation configurations examples for different infrastructure providers. Currently only AWS is available but they can be used with almost any provider with little changes.
84 You need to copy one of the example configuration files and directory, and edit them to suit your needs.
86 h3(#multi_host_multi_hostnames). Multiple hosts / multiple hostnames
88 <pre><code>cp local.params.example.multiple_hosts local.params
89 cp -r config_examples/multi_host/aws local_config_dir
93 Edit the variables in the <i>local.params</i> file. Pay attention to the <b>*_INT_IP, *_TOKEN</b> and <b>*KEY</b> variables. Those variables will be used to do a search and replace on the <i>pillars/*</i> in place of any matching __VARIABLE__.
95 The <i>multi_host</i> include LetsEncrypt salt code to automatically request and install the certificates for the public-facing hosts (API, Workbench) so it will need the hostnames to be reachable from the Internet. If this cluster will not be the case, please set the variable <i>USE_LETSENCRYPT=no</i>.
97 h3(#further_customization). Further customization of the installation (modifying the salt pillars and states)
99 You will need further customization to suit your environment, which can be done editing the Saltstack pillars and states files. Pay particular attention to the <i>pillars/arvados.sls</i> file, where you will need to provide some information that can be retrieved as output of the terraform run.
101 Any extra <i>state</i> file you add under <i>local_config_dir/states</i> will be added to the salt run and applied to the hosts.
103 h2(#installation_order). Installation order
105 A few Arvados nodes need to be installed in certain order. The required order is
109 #. The other nodes can be installed in any order after the two above
111 h2(#run_provision_script). Run the provision.sh script
113 When you finished customizing the configuration, you are ready to copy the files to the hosts and run the @provision.sh@ script. The script allows you to specify the <i>role/s</i> a node will have and it will install only the Arvados components required for such role. The general format of the command is:
116 <pre><code>scp -r provision.sh local* user@host:
117 ssh user@host sudo ./provision.sh --roles comma,separated,list,of,roles,to,apply
121 and wait for it to finish.
123 If everything goes OK, you'll get some final lines stating something like:
126 <pre><code>arvados: Succeeded: 109 (changed=9)
131 The distribution of role as described above can be applied running these commands:
135 <pre><code>scp -r provision.sh local* user@host:
136 ssh user@host sudo ./provision.sh --config local.params --roles database
142 <pre><code>scp -r provision.sh local* user@host:
143 ssh user@host sudo ./provision.sh --config local.params --roles api,controller,websocket,dispatcher
149 <pre><code>scp -r provision.sh local* user@host:
150 ssh user@host sudo ./provision.sh --config local.params --roles keepstore
156 <pre><code>scp -r provision.sh local* user@host:
157 ssh user@host sudo ./provision.sh --config local.params --roles workbench,workbench2,webshell
161 #. Keepproxy / Keepweb
163 <pre><code>scp -r provision.sh local* user@host:
164 ssh user@host sudo ./provision.sh --config local.params --roles keepproxy,keepweb
168 #. Shell (here we copy the CLI test workflow too)
170 <pre><code>scp -r provision.sh local* tests user@host:
171 ssh user@host sudo ./provision.sh --config local.params --roles shell
175 h2(#initial_user). Initial user and login
177 At this point you should be able to log into the Arvados cluster. The initial URL will be:
179 * https://workbench.arva2.arv.local
181 or, in general, the url format will be:
183 * https://workbench.@<cluster>.<domain>@
185 By default, the provision script creates an initial user for testing purposes. This user is configured as administrator of the newly created cluster.
187 Assuming you didn't change these values in the @local.params@ file, the initial credentials are:
190 * Password: 'password'
191 * Email: 'admin@arva2.arv.local'
193 h2(#test_install). Test the installed cluster running a simple workflow
195 If you followed the instructions above, the @provision.sh@ script saves a simple example test workflow in the @/tmp/cluster_tests@ directory in the @shell@ node. If you want to run it, just ssh to the node, change to that directory and run:
198 <pre><code>cd /tmp/cluster_tests
203 It will create a test user (by default, the same one as the admin user), upload a small workflow and run it. If everything goes OK, the output should similar to this (some output was shortened for clarity):
206 <pre><code>Creating Arvados Standard Docker Images project
207 Arvados project uuid is 'arva2-j7d0g-0prd8cjlk6kfl7y'
210 "uuid":"arva2-o0j2j-n4zu4cak5iifq2a",
211 "owner_uuid":"arva2-tpzed-000000000000000",
214 Uploading arvados/jobs' docker image to the project
215 2.1.1: Pulling from arvados/jobs
216 8559a31e96f4: Pulling fs layer
218 Status: Downloaded newer image for arvados/jobs:2.1.1
219 docker.io/arvados/jobs:2.1.1
220 2020-11-23 21:43:39 arvados.arv_put[32678] INFO: Creating new cache file at /home/vagrant/.cache/arvados/arv-put/c59256eda1829281424c80f588c7cc4d
221 2020-11-23 21:43:46 arvados.arv_put[32678] INFO: Collection saved as 'Docker image arvados jobs:2.1.1 sha256:0dd50'
222 arva2-4zz18-1u5pvbld7cvxuy2
223 Creating initial user ('admin')
224 Setting up user ('admin')
229 "owner_uuid":"arva2-tpzed-000000000000000",
231 "uuid":"arva2-o0j2j-1ownrdne0ok9iox"
235 "owner_uuid":"arva2-tpzed-000000000000000",
237 "uuid":"arva2-o0j2j-1zbeyhcwxc1tvb7"
241 "email":"admin@arva2.arv.local",
243 "owner_uuid":"arva2-tpzed-000000000000000",
246 "uuid":"arva2-tpzed-3wrm93zmzpshrq2",
250 "kind":"arvados#HashList"
252 Activating user 'admin'
255 "email":"admin@arva2.arv.local",
258 "uuid":"arva2-tpzed-3wrm93zmzpshrq2",
261 Running test CWL workflow
262 INFO /usr/bin/cwl-runner 2.1.1, arvados-python-client 2.1.1, cwltool 3.0.20200807132242
263 INFO Resolved 'hasher-workflow.cwl' to 'file:///tmp/cluster_tests/hasher-workflow.cwl'
265 INFO Using cluster arva2 (https://arva2.arv.local:8443/)
266 INFO Upload local files: "test.txt"
267 INFO Uploaded to ea34d971b71d5536b4f6b7d6c69dc7f6+50 (arva2-4zz18-c8uvwqdry4r8jao)
268 INFO Using collection cache size 256 MiB
269 INFO [container hasher-workflow.cwl] submitted container_request arva2-xvhdp-v1bkywd58gyocwm
270 INFO [container hasher-workflow.cwl] arva2-xvhdp-v1bkywd58gyocwm is Final
271 INFO Overall process status is success
272 INFO Final output collection d6c69a88147dde9d52a418d50ef788df+123
275 "basename": "hasher3.md5sum.txt",
277 "location": "keep:d6c69a88147dde9d52a418d50ef788df+123/hasher3.md5sum.txt",
281 INFO Final process status is success