1 # Copyright (C) The Arvados Authors. All rights reserved.
3 # SPDX-License-Identifier: AGPL-3.0
5 # Do not use this file for site configuration. Create
6 # /etc/arvados/config.yml instead.
8 # The order of precedence (highest to lowest):
9 # 1. Legacy component-specific config files (deprecated)
10 # 2. /etc/arvados/config.yml
11 # 3. config.default.yml
15 # Token used internally by Arvados components to authenticate to
16 # one another. Use a string of at least 50 random alphanumerics.
19 # Token to be included in all healthcheck requests. Disabled by default.
20 # Server expects request header of the format "Authorization: Bearer xxx"
25 # In each of the service sections below, the keys under
26 # InternalURLs are the endpoints where the service should be
27 # listening, and reachable from other hosts in the
31 # "http://host1.example:12345": {}
32 # "http://host2.example:12345": {}
35 InternalURLs: {SAMPLE: {}}
38 InternalURLs: {SAMPLE: {}}
41 InternalURLs: {SAMPLE: {}}
44 InternalURLs: {SAMPLE: {}}
47 InternalURLs: {SAMPLE: {}}
50 InternalURLs: {SAMPLE: {}}
53 InternalURLs: {SAMPLE: {}}
56 InternalURLs: {SAMPLE: {}}
59 InternalURLs: {SAMPLE: {}}
62 InternalURLs: {SAMPLE: {}}
63 # Base URL for Workbench inline preview. If blank, use
64 # WebDAVDownload instead, and disable inline preview.
65 # If both are empty, downloading collections from workbench
68 # It is important to properly configure the download service
69 # to migitate cross-site-scripting (XSS) attacks. A HTML page
70 # can be stored in collection. If an attacker causes a victim
71 # to visit that page through Workbench, it will be rendered by
72 # the browser. If all collections are served at the same
73 # domain, the browser will consider collections as coming from
74 # the same origin and having access to the same browsing data,
75 # enabling malicious Javascript on that page to access Arvados
76 # on behalf of the victim.
78 # This is mitigating by having separate domains for each
79 # collection, or limiting preview to circumstances where the
80 # collection is not accessed with the user's regular
83 # Serve preview links using uuid or pdh in subdomain
84 # (requires wildcard DNS and TLS certificate)
85 # https://*.collections.uuid_prefix.arvadosapi.com
87 # Serve preview links using uuid or pdh in main domain
88 # (requires wildcard DNS and TLS certificate)
89 # https://*--collections.uuid_prefix.arvadosapi.com
91 # Serve preview links by setting uuid or pdh in the path.
92 # This configuration only allows previews of public data or
93 # collection-sharing links, because these use the anonymous
94 # user token or the token is already embedded in the URL.
95 # Other data must be handled as downloads via WebDAVDownload:
96 # https://collections.uuid_prefix.arvadosapi.com
101 InternalURLs: {SAMPLE: {}}
102 # Base URL for download links. If blank, serve links to WebDAV
103 # with disposition=attachment query param. Unlike preview links,
104 # browsers do not render attachments, so there is no risk of XSS.
106 # If WebDAVDownload is blank, and WebDAV uses a
107 # single-origin form, then Workbench will show an error page
109 # Serve download links by setting uuid or pdh in the path:
110 # https://download.uuid_prefix.arvadosapi.com
117 # Rendezvous is normally empty/omitted. When changing the
118 # URL of a Keepstore service, Rendezvous should be set to
119 # the old URL (with trailing slash omitted) to preserve
120 # rendezvous ordering.
124 InternalURLs: {SAMPLE: {}}
127 InternalURLs: {SAMPLE: {}}
128 # ShellInABox service endpoint URL for a given VM. If empty, do not
129 # offer web shell logins.
131 # E.g., using a path-based proxy server to forward connections to shell hosts:
132 # https://webshell.uuid_prefix.arvadosapi.com
134 # E.g., using a name-based proxy server to forward connections to shell hosts:
135 # https://*.webshell.uuid_prefix.arvadosapi.com
138 InternalURLs: {SAMPLE: {}}
141 InternalURLs: {SAMPLE: {}}
144 InternalURLs: {SAMPLE: {}}
148 # max concurrent connections per arvados server daemon
151 # All parameters here are passed to the PG client library in a connection string;
152 # see https://www.postgresql.org/docs/current/static/libpq-connect.html#LIBPQ-PARAMKEYWORDS
160 # Limits for how long a client token created by regular users can be valid,
161 # and also is used as a default expiration policy when no expiration date is
163 # Default value zero means token expirations don't get clamped and no
164 # default expiration is set.
167 # Maximum size (in bytes) allowed for a single API request. This
168 # limit is published in the discovery document for use by clients.
169 # Note: You must separately configure the upstream web server or
170 # proxy to actually enforce the desired maximum request size on the
172 MaxRequestSize: 134217728
174 # Limit the number of bytes read from the database during an index
175 # request (by retrieving and returning fewer rows than would
176 # normally be returned in a single response).
177 # Note 1: This setting never reduces the number of returned rows to
178 # zero, no matter how big the first data row is.
179 # Note 2: Currently, this is only checked against a specific set of
180 # columns that tend to get large (collections.manifest_text,
181 # containers.mounts, workflows.definition). Other fields (e.g.,
182 # "properties" hashes) are not counted against this limit.
183 MaxIndexDatabaseRead: 134217728
185 # Maximum number of items to return when responding to a APIs that
186 # can return partial result sets using limit and offset parameters
187 # (e.g., *.index, groups.contents). If a request specifies a "limit"
188 # parameter higher than this value, this value is used instead.
189 MaxItemsPerResponse: 1000
191 # Maximum number of concurrent requests to accept in a single
192 # service process, or 0 for no limit.
193 MaxConcurrentRequests: 0
195 # Maximum number of 64MiB memory buffers per Keepstore server process, or
196 # 0 for no limit. When this limit is reached, up to
197 # (MaxConcurrentRequests - MaxKeepBlobBuffers) HTTP requests requiring
198 # buffers (like GET and PUT) will wait for buffer space to be released.
199 # Any HTTP requests beyond MaxConcurrentRequests will receive an
200 # immediate 503 response.
202 # MaxKeepBlobBuffers should be set such that (MaxKeepBlobBuffers * 64MiB
203 # * 1.1) fits comfortably in memory. On a host dedicated to running
204 # Keepstore, divide total memory by 88MiB to suggest a suitable value.
205 # For example, if grep MemTotal /proc/meminfo reports MemTotal: 7125440
206 # kB, compute 7125440 / (88 * 1024)=79 and set MaxKeepBlobBuffers: 79
207 MaxKeepBlobBuffers: 128
209 # API methods to disable. Disabled methods are not listed in the
210 # discovery document, and respond 404 to all requests.
211 # Example: {"jobs.create":{}, "pipeline_instances.create": {}}
214 # Interval (seconds) between asynchronous permission view updates. Any
215 # permission-updating API called with the 'async' parameter schedules a an
216 # update on the permission view in the future, if not already scheduled.
217 AsyncPermissionsUpdateInterval: 20s
219 # Maximum number of concurrent outgoing requests to make while
220 # serving a single incoming multi-cluster (federated) request.
221 MaxRequestAmplification: 4
223 # Maximum wall clock time to spend handling an incoming request.
226 # Websocket will send a periodic empty event after 'SendTimeout'
227 # if there is no other activity to maintain the connection /
228 # detect dropped connections.
231 WebsocketClientEventQueue: 64
232 WebsocketServerEventQueue: 4
234 # Timeout on requests to internal Keep services.
235 KeepServiceRequestTimeout: 15s
238 # Config parameters to automatically setup new users. If enabled,
239 # this users will be able to self-activate. Enable this if you want
240 # to run an open instance where anyone can create an account and use
241 # the system without requiring manual approval.
243 # The params AutoSetupNewUsersWith* are meaningful only when AutoSetupNewUsers is turned on.
244 # AutoSetupUsernameBlacklist is a list of usernames to be blacklisted for auto setup.
245 AutoSetupNewUsers: false
246 AutoSetupNewUsersWithVmUUID: ""
247 AutoSetupNewUsersWithRepository: false
248 AutoSetupUsernameBlacklist:
257 # When NewUsersAreActive is set to true, new users will be active
258 # immediately. This skips the "self-activate" step which enforces
259 # user agreements. Should only be enabled for development.
260 NewUsersAreActive: false
262 # The e-mail address of the user you would like to become marked as an admin
263 # user on their first login.
264 AutoAdminUserWithEmail: ""
266 # If AutoAdminFirstUser is set to true, the first user to log in when no
267 # other admin users exist will automatically become an admin user.
268 AutoAdminFirstUser: false
270 # Email address to notify whenever a user creates a profile for the
272 UserProfileNotificationAddress: ""
273 AdminNotifierEmailFrom: arvados@example.com
274 EmailSubjectPrefix: "[ARVADOS] "
275 UserNotifierEmailFrom: arvados@example.com
276 NewUserNotificationRecipients: {}
277 NewInactiveUserNotificationRecipients: {}
279 # Set AnonymousUserToken to enable anonymous user access. Populate this
280 # field with a long random string. Then run "bundle exec
281 # ./script/get_anonymous_user_token.rb" in the directory where your API
282 # server is running to record the token in the database.
283 AnonymousUserToken: ""
285 # If a new user has an alternate email address (local@domain)
286 # with the domain given here, its local part becomes the new
287 # user's default username. Otherwise, the user's primary email
289 PreferDomainForUsername: ""
292 <% if not @user.full_name.empty? -%>
293 <%= @user.full_name %>,
298 Your Arvados account has been set up. You can log in at
300 <%= Rails.configuration.Services.Workbench1.ExternalURL %>
303 Your Arvados administrator.
306 # Time to keep audit logs, in seconds. (An audit log is a row added
307 # to the "logs" table in the PostgreSQL database each time an
308 # Arvados object is created, modified, or deleted.)
310 # Currently, websocket event notifications rely on audit logs, so
311 # this should not be set lower than 300 (5 minutes).
314 # Maximum number of log rows to delete in a single SQL transaction.
316 # If MaxDeleteBatch is 0, log entries will never be
317 # deleted by Arvados. Cleanup can be done by an external process
318 # without affecting any Arvados system processes, as long as very
319 # recent (<5 minutes old) logs are not deleted.
321 # 100000 is a reasonable batch size for most sites.
324 # Attributes to suppress in events and audit logs. Notably,
325 # specifying {"manifest_text": {}} here typically makes the database
326 # smaller and faster.
328 # Warning: Using any non-empty value here can have undesirable side
329 # effects for any client or component that relies on event logs.
330 # Use at your own risk.
331 UnloggedAttributes: {}
335 # Logging threshold: panic, fatal, error, warn, info, debug, or
339 # Logging format: json or text
342 # Maximum characters of (JSON-encoded) query parameters to include
343 # in each request log entry. When params exceed this size, they will
344 # be JSON-encoded, truncated to this size, and logged as
346 MaxRequestLogParamsSize: 2000
350 # Enable access controls for data stored in Keep. This should
351 # always be set to true on a production cluster.
354 # BlobSigningKey is a string of alphanumeric characters used to
355 # generate permission signatures for Keep locators. It must be
356 # identical to the permission key given to Keep. IMPORTANT: This
357 # is a site secret. It should be at least 50 characters.
359 # Modifying BlobSigningKey will invalidate all existing
360 # signatures, which can cause programs to fail (e.g., arv-put,
361 # arv-get, and Crunch jobs). To avoid errors, rotate keys only
362 # when no such processes are running.
365 # Enable garbage collection of unreferenced blobs in Keep.
368 # Time to leave unreferenced blobs in "trashed" state before
369 # deleting them, or 0 to skip the "trashed" state entirely and
370 # delete unreferenced blobs.
372 # If you use any Amazon S3 buckets as storage volumes, this
373 # must be at least 24h to avoid occasional data loss.
374 BlobTrashLifetime: 336h
376 # How often to check for (and delete) trashed blocks whose
377 # BlobTrashLifetime has expired.
378 BlobTrashCheckInterval: 24h
380 # Maximum number of concurrent "trash blob" and "delete trashed
381 # blob" operations conducted by a single keepstore process. Each
382 # of these can be set to 0 to disable the respective operation.
384 # If BlobTrashLifetime is zero, "trash" and "delete trash"
385 # happen at once, so only the lower of these two values is used.
386 BlobTrashConcurrency: 4
387 BlobDeleteConcurrency: 4
389 # Maximum number of concurrent "create additional replica of
390 # existing blob" operations conducted by a single keepstore
392 BlobReplicateConcurrency: 4
394 # Default replication level for collections. This is used when a
395 # collection's replication_desired attribute is nil.
396 DefaultReplication: 2
398 # BlobSigningTTL determines the minimum lifetime of transient
399 # data, i.e., blocks that are not referenced by
400 # collections. Unreferenced blocks exist for two reasons:
402 # 1) A data block must be written to a disk/cloud backend device
403 # before a collection can be created/updated with a reference to
406 # 2) Deleting or updating a collection can remove the last
407 # remaining reference to a data block.
409 # If BlobSigningTTL is too short, long-running
410 # processes/containers will fail when they take too long (a)
411 # between writing blocks and writing collections that reference
412 # them, or (b) between reading collections and reading the
415 # If BlobSigningTTL is too long, data will still be stored long
416 # after the referring collections are deleted, and you will
417 # needlessly fill up disks or waste money on cloud storage.
419 # Modifying BlobSigningTTL invalidates existing signatures; see
420 # BlobSigningKey note above.
422 # The default is 2 weeks.
425 # When running keep-balance, this is the destination filename for
426 # the list of lost block hashes if there are any, one per line.
427 # Updated automically during each successful run.
428 BlobMissingReport: ""
430 # keep-balance operates periodically, i.e.: do a
431 # scan/balance operation, sleep, repeat.
433 # BalancePeriod determines the interval between start times of
434 # successive scan/balance operations. If a scan/balance operation
435 # takes longer than BalancePeriod, the next one will follow it
438 # If SIGUSR1 is received during an idle period between operations,
439 # the next operation will start immediately.
442 # Limits the number of collections retrieved by keep-balance per
443 # API transaction. If this is zero, page size is
444 # determined by the API server's own page size limits (see
445 # API.MaxItemsPerResponse and API.MaxIndexDatabaseRead).
446 BalanceCollectionBatch: 0
448 # The size of keep-balance's internal queue of
449 # collections. Higher values use more memory and improve throughput
450 # by allowing keep-balance to fetch the next page of collections
451 # while the current page is still being processed. If this is zero
452 # or omitted, pages are processed serially.
453 BalanceCollectionBuffers: 1000
455 # Maximum time for a rebalancing run. This ensures keep-balance
456 # eventually gives up and retries if, for example, a network
457 # error causes a hung connection that is never closed by the
458 # OS. It should be long enough that it doesn't interrupt a
459 # long-running balancing operation.
462 # Default lifetime for ephemeral collections: 2 weeks. This must not
463 # be less than BlobSigningTTL.
464 DefaultTrashLifetime: 336h
466 # Interval (seconds) between trash sweeps. During a trash sweep,
467 # collections are marked as trash if their trash_at time has
468 # arrived, and deleted if their delete_at time has arrived.
469 TrashSweepInterval: 60s
471 # If true, enable collection versioning.
472 # When a collection's preserve_version field is true or the current version
473 # is older than the amount of seconds defined on PreserveVersionIfIdle,
474 # a snapshot of the collection's previous state is created and linked to
475 # the current collection.
476 CollectionVersioning: false
478 # 0s = auto-create a new version on every update.
479 # -1s = never auto-create new versions.
480 # > 0s = auto-create a new version when older than the specified number of seconds.
481 PreserveVersionIfIdle: -1s
483 # If non-empty, allow project and collection names to contain
484 # the "/" character (slash/stroke/solidus), and replace "/" with
485 # the given string in the filesystem hierarchy presented by
486 # WebDAV. Example values are "%2f" and "{slash}". Names that
487 # contain the substitution string itself may result in confusing
488 # behavior, so a value like "_" is not recommended.
490 # If the default empty value is used, the server will reject
491 # requests to create or rename a collection when the new name
494 # If the value "/" is used, project and collection names
495 # containing "/" will be allowed, but they will not be
496 # accessible via WebDAV.
498 # Use of this feature is not recommended, if it can be avoided.
499 ForwardSlashNameSubstitution: ""
501 # Include "folder objects" in S3 ListObjects responses.
502 S3FolderObjects: true
504 # Managed collection properties. At creation time, if the client didn't
505 # provide the listed keys, they will be automatically populated following
506 # one of the following behaviors:
508 # * UUID of the user who owns the containing project.
509 # responsible_person_uuid: {Function: original_owner, Protected: true}
511 # * Default concrete value.
512 # foo_bar: {Value: baz, Protected: false}
514 # If Protected is true, only an admin user can modify its value.
516 SAMPLE: {Function: original_owner, Protected: true}
518 # In "trust all content" mode, Workbench will redirect download
519 # requests to WebDAV preview link, even in the cases when
520 # WebDAV would have to expose XSS vulnerabilities in order to
521 # handle the redirect (see discussion on Services.WebDAV).
523 # This setting has no effect in the recommended configuration,
524 # where the WebDAV is configured to have a separate domain for
525 # every collection; in this case XSS protection is provided by
526 # browsers' same-origin policy.
528 # The default setting (false) is appropriate for a multi-user site.
529 TrustAllContent: false
531 # Cache parameters for WebDAV content serving:
533 # Time to cache manifests, permission checks, and sessions.
536 # Time to cache collection state.
539 # Block cache entries. Each block consumes up to 64 MiB RAM.
542 # Collection cache entries.
543 MaxCollectionEntries: 1000
545 # Approximate memory limit (in bytes) for collection cache.
546 MaxCollectionBytes: 100000000
548 # Permission cache entries.
549 MaxPermissionEntries: 1000
551 # UUID cache entries.
554 # Persistent sessions.
558 # One of the following mechanisms (SSO, Google, PAM, LDAP, or
559 # LoginCluster) should be enabled; see
560 # https://doc.arvados.org/install/setup-login.html
563 # Authenticate with Google.
566 # Use the Google Cloud console to enable the People API (APIs
567 # and Services > Enable APIs and services > Google People API
568 # > Enable), generate a Client ID and secret (APIs and
569 # Services > Credentials > Create credentials > OAuth client
570 # ID > Web application) and add your controller's /login URL
571 # (e.g., "https://zzzzz.example.com/login") as an authorized
576 # Allow users to log in to existing accounts using any verified
577 # email address listed by their Google account. If true, the
578 # Google People API must be enabled in order for Google login to
579 # work. If false, only the primary email address will be used.
580 AlternateEmailAddresses: true
582 # Send additional parameters with authentication requests. See
583 # https://developers.google.com/identity/protocols/oauth2/openid-connect#authenticationuriparameters
584 # for a list of supported parameters.
585 AuthenticationRequestParameters:
586 # Show the "choose which Google account" page, even if the
587 # client is currently logged in to exactly one Google
589 prompt: select_account
594 # Authenticate with an OpenID Connect provider.
597 # Issuer URL, e.g., "https://login.example.com".
599 # This must be exactly equal to the URL returned by the issuer
600 # itself in its config response ("isser" key). If the
601 # configured value is "https://example" and the provider
602 # returns "https://example:443" or "https://example/" then
603 # login will fail, even though those URLs are equivalent
607 # Your client ID and client secret (supplied by the provider).
611 # OpenID claim field containing the user's email
612 # address. Normally "email"; see
613 # https://openid.net/specs/openid-connect-core-1_0.html#StandardClaims
616 # OpenID claim field containing the email verification
617 # flag. Normally "email_verified". To accept every returned
618 # email address without checking a "verified" field at all,
619 # use the empty string "".
620 EmailVerifiedClaim: "email_verified"
622 # OpenID claim field containing the user's preferred
623 # username. If empty, use the mailbox part of the user's email
627 # Send additional parameters with authentication requests,
628 # like {display: page, prompt: consent}. See
629 # https://openid.net/specs/openid-connect-core-1_0.html#AuthRequest
630 # and refer to your provider's documentation for supported
632 AuthenticationRequestParameters:
635 # Accept an OIDC access token as an API token if the OIDC
636 # provider's UserInfo endpoint accepts it.
638 # AcceptAccessTokenScope should also be used when enabling
640 AcceptAccessToken: false
642 # Before accepting an OIDC access token as an API token, first
643 # check that it is a JWT whose "scope" value includes this
644 # value. Example: "https://zzzzz.example.com/" (your Arvados
647 # If this value is empty and AcceptAccessToken is true, all
648 # access tokens will be accepted regardless of scope,
649 # including non-JWT tokens. This is not recommended.
650 AcceptAccessTokenScope: ""
653 # (Experimental) Use PAM to authenticate users.
656 # PAM service name. PAM will apply the policy in the
657 # corresponding config file (e.g., /etc/pam.d/arvados) or, if
658 # there is none, the default "other" config.
661 # Domain name (e.g., "example.com") to use to construct the
662 # user's email address if PAM authentication returns a
663 # username with no "@". If empty, use the PAM username as the
664 # user's email address, whether or not it contains "@".
666 # Note that the email address is used as the primary key for
667 # user records when logging in. Therefore, if you change
668 # PAMDefaultEmailDomain after the initial installation, you
669 # should also update existing user records to reflect the new
670 # domain. Otherwise, next time those users log in, they will
671 # be given new accounts instead of accessing their existing
673 DefaultEmailDomain: ""
676 # Use an LDAP service to authenticate users.
679 # Server URL, like "ldap://ldapserver.example.com:389" or
680 # "ldaps://ldapserver.example.com:636".
681 URL: "ldap://ldap:389"
683 # Use StartTLS upon connecting to the server.
686 # Skip TLS certificate name verification.
689 # Strip the @domain part if a user supplies an email-style
690 # username with this domain. If "*", strip any user-provided
691 # domain. If "", never strip the domain part. Example:
695 # If, after applying StripDomain, the username contains no "@"
696 # character, append this domain to form an email-style
697 # username. Example: "example.com"
700 # The LDAP attribute to filter on when looking up a username
701 # (after applying StripDomain and AppendDomain).
704 # Bind with this username (DN or UPN) and password when
705 # looking up the user record.
707 # Example user: "cn=admin,dc=example,dc=com"
709 SearchBindPassword: ""
711 # Directory base for username lookup. Example:
712 # "ou=Users,dc=example,dc=com"
715 # Additional filters to apply when looking up users' LDAP
716 # entries. This can be used to restrict access to a subset of
717 # LDAP users, or to disambiguate users from other directory
718 # entries that have the SearchAttribute present.
720 # Special characters in assertion values must be escaped (see
723 # Example: "(objectClass=person)"
726 # LDAP attribute to use as the user's email address.
728 # Important: This must not be an attribute whose value can be
729 # edited in the directory by the users themselves. Otherwise,
730 # users can take over other users' Arvados accounts trivially
731 # (email address is the primary key for Arvados accounts.)
734 # LDAP attribute to use as the preferred Arvados username. If
735 # no value is found (or this config is empty) the username
736 # originally supplied by the user will be used.
737 UsernameAttribute: uid
740 # Authenticate with a separate SSO server. (Deprecated)
743 # ProviderAppID and ProviderAppSecret are generated during SSO
745 # https://doc.arvados.org/v2.0/install/install-sso.html#update-config
747 ProviderAppSecret: ""
750 # Authenticate users listed here in the config file. This
751 # feature is intended to be used in test environments, and
752 # should not be used in production.
756 Email: alice@example.com
759 # The cluster ID to delegate the user database. When set,
760 # logins on this cluster will be redirected to the login cluster
761 # (login cluster must appear in RemoteClusters with Proxy: true)
764 # How long a cached token belonging to a remote cluster will
765 # remain valid before it needs to be revalidated.
766 RemoteTokenRefresh: 5m
768 # How long a client token created from a login flow will be valid without
769 # asking the user to re-login. Example values: 60m, 8h.
770 # Default value zero means tokens don't have expiration.
773 # If true (default) tokens issued through login are allowed to create
775 # If false, tokens issued through login are not allowed to
776 # viewing/creating other tokens. New tokens can only be created
777 # by going through login again.
778 IssueTrustedTokens: true
780 # When the token is returned to a client, the token itself may
781 # be restricted from viewing/creating other tokens based on whether
782 # the client is "trusted" or not. The local Workbench1 and
783 # Workbench2 are trusted by default, but if this is a
784 # LoginCluster, you probably want to include the other Workbench
785 # instances in the federation in this list.
788 "https://workbench.federate1.example": {}
789 "https://workbench.federate2.example": {}
792 # Path to git or gitolite-shell executable. Each authenticated
793 # request will execute this program with the single argument "http-backend"
794 GitCommand: /usr/bin/git
796 # Path to Gitolite's home directory. If a non-empty path is given,
797 # the CGI environment will be set up to support the use of
798 # gitolite-shell as a GitCommand: for example, if GitoliteHome is
799 # "/gh", then the CGI environment will have GITOLITE_HTTP_HOME=/gh,
800 # PATH=$PATH:/gh/bin, and GL_BYPASS_ACCESS_CHECKS=1.
803 # Git repositories must be readable by api server, or you won't be
804 # able to submit crunch jobs. To pass the test suites, put a clone
805 # of the arvados tree in {git_repositories_dir}/arvados.git or
806 # {git_repositories_dir}/arvados/.git
807 Repositories: /var/lib/arvados/git/repositories
815 # List of supported Docker Registry image formats that compute nodes
816 # are able to use. `arv keep docker` will error out if a user tries
817 # to store an image with an unsupported format. Use an empty array
818 # to skip the compatibility check (and display a warning message to
821 # Example for sites running docker < 1.10: {"v1": {}}
822 # Example for sites running docker >= 1.10: {"v2": {}}
823 # Example for disabling check: {}
824 SupportedDockerImageFormats:
828 # Include details about job reuse decisions in the server log. This
829 # causes additional database queries to run, so it should not be
830 # enabled unless you expect to examine the resulting logs for
831 # troubleshooting purposes.
832 LogReuseDecisions: false
834 # Default value for keep_cache_ram of a container's runtime_constraints.
835 DefaultKeepCacheRAM: 268435456
837 # Number of times a container can be unlocked before being
838 # automatically cancelled.
839 MaxDispatchAttempts: 5
841 # Default value for container_count_max for container requests. This is the
842 # number of times Arvados will create a new container to satisfy a container
843 # request. If a container is cancelled it will retry a new container if
844 # container_count < container_count_max on any container requests associated
845 # with the cancelled container.
848 # The maximum number of compute nodes that can be in use simultaneously
849 # If this limit is reduced, any existing nodes with slot number >= new limit
850 # will not be counted against the new limit. In other words, the new limit
851 # won't be strictly enforced until those nodes with higher slot numbers
855 # Preemptible instance support (e.g. AWS Spot Instances)
856 # When true, child containers will get created with the preemptible
857 # scheduling parameter parameter set.
858 UsePreemptibleInstances: false
860 # PEM encoded SSH key (RSA, DSA, or ECDSA) used by the
861 # (experimental) cloud dispatcher for executing containers on
862 # worker VMs. Begins with "-----BEGIN RSA PRIVATE KEY-----\n"
863 # and ends with "\n-----END RSA PRIVATE KEY-----\n".
864 DispatchPrivateKey: ""
866 # Maximum time to wait for workers to come up before abandoning
867 # stale locks from a previous dispatch process.
870 # The crunch-run command used to start a container on a worker node.
872 # When dispatching to cloud VMs, this is used only if
873 # DeployRunnerBinary in the CloudVMs section is set to the empty
875 CrunchRunCommand: "crunch-run"
877 # Extra arguments to add to crunch-run invocation
878 # Example: ["--cgroup-parent-subsystem=memory"]
879 CrunchRunArgumentsList: []
881 # Extra RAM to reserve on the node, in addition to
882 # the amount specified in the container's RuntimeConstraints
883 ReserveExtraRAM: 256MiB
885 # Minimum time between two attempts to run the same container
888 # Container runtime: "docker" (default) or "singularity" (experimental)
889 RuntimeEngine: docker
892 # When you run the db:delete_old_container_logs task, it will find
893 # containers that have been finished for at least this many seconds,
894 # and delete their stdout, stderr, arv-mount, crunch-run, and
895 # crunchstat logs from the logs table.
898 # These two settings control how frequently log events are flushed to the
899 # database. Log lines are buffered until either crunch_log_bytes_per_event
900 # has been reached or crunch_log_seconds_between_events has elapsed since
902 LogBytesPerEvent: 4096
903 LogSecondsBetweenEvents: 5s
905 # The sample period for throttling logs.
906 LogThrottlePeriod: 60s
908 # Maximum number of bytes that job can log over crunch_log_throttle_period
909 # before being silenced until the end of the period.
910 LogThrottleBytes: 65536
912 # Maximum number of lines that job can log over crunch_log_throttle_period
913 # before being silenced until the end of the period.
914 LogThrottleLines: 1024
916 # Maximum bytes that may be logged by a single job. Log bytes that are
917 # silenced by throttling are not counted against this total.
918 LimitLogBytesPerJob: 67108864
920 LogPartialLineThrottlePeriod: 5s
922 # Container logs are written to Keep and saved in a
923 # collection, which is updated periodically while the
924 # container runs. This value sets the interval between
925 # collection updates.
928 # The log collection is also updated when the specified amount of
929 # log data (given in bytes) is produced in less than one update
934 # An admin user can use "arvados-client shell" to start an
935 # interactive shell (with any user ID) in any running
939 # Any user can use "arvados-client shell" to start an
940 # interactive shell (with any user ID) in any running
941 # container that they started, provided it isn't also
942 # associated with a different user's container request.
944 # Interactive sessions make it easy to alter the container's
945 # runtime environment in ways that aren't recorded or
946 # reproducible. Consider the implications for automatic
947 # container reuse before enabling and using this feature. In
948 # particular, note that starting an interactive session does
949 # not disqualify a container from being reused by a different
950 # user/workflow in the future.
955 SbatchArgumentsList: []
956 SbatchEnvironmentVariables:
959 # Path to dns server configuration directory
960 # (e.g. /etc/unbound.d/conf.d). If false, do not write any config
961 # files or touch restart.txt (see below).
964 # Template file for the dns server host snippets. See
965 # unbound.template in this directory for an example. If false, do
966 # not write any config files.
967 DNSServerConfTemplate: ""
969 # String to write to {dns_server_conf_dir}/restart.txt (with a
970 # trailing newline) after updating local data. If false, do not
971 # open or write the restart.txt file.
972 DNSServerReloadCommand: ""
974 # Command to run after each DNS update. Template variables will be
975 # substituted; see the "unbound" example below. If false, do not run
977 DNSServerUpdateCommand: ""
979 ComputeNodeDomain: ""
980 ComputeNodeNameservers:
984 # Hostname to assign to a compute node when it sends a "ping" and the
985 # hostname in its Node record is nil.
986 # During bootstrapping, the "ping" script is expected to notice the
987 # hostname given in the ping response, and update its unix hostname
989 # If false, leave the hostname alone (this is appropriate if your compute
990 # nodes' hostnames are already assigned by some other mechanism).
992 # One way or another, the hostnames of your node records should agree
993 # with your DNS records and your /etc/slurm-llnl/slurm.conf files.
995 # Example for compute0000, compute0001, ....:
996 # assign_node_hostname: compute%<slot_number>04d
997 # (See http://ruby-doc.org/core-2.2.2/Kernel.html#method-i-format for more.)
998 AssignNodeHostname: "compute%<slot_number>d"
1001 # Enable the legacy 'jobs' API (crunch v1). This value must be a string.
1003 # Note: this only enables read-only access, creating new
1004 # legacy jobs and pipelines is not supported.
1006 # 'auto' -- (default) enable the Jobs API only if it has been used before
1007 # (i.e., there are job records in the database)
1008 # 'true' -- enable the Jobs API despite lack of existing records.
1009 # 'false' -- disable the Jobs API despite presence of existing records.
1012 # Git repositories must be readable by api server, or you won't be
1013 # able to submit crunch jobs. To pass the test suites, put a clone
1014 # of the arvados tree in {git_repositories_dir}/arvados.git or
1015 # {git_repositories_dir}/arvados/.git
1016 GitInternalDir: /var/lib/arvados/internal.git
1019 # Enable the cloud scheduler (experimental).
1022 # Name/number of port where workers' SSH services listen.
1025 # Interval between queue polls.
1028 # Shell command to execute on each worker to determine whether
1029 # the worker is booted and ready to run containers. It should
1030 # exit zero if the worker is ready.
1031 BootProbeCommand: "docker ps -q"
1033 # Minimum interval between consecutive probes to a single
1037 # Maximum probes per second, across all workers in a pool.
1038 MaxProbesPerSecond: 10
1040 # Time before repeating SIGTERM when killing a container.
1043 # Time to give up on a process (most likely arv-mount) that
1044 # still holds a container lockfile after its main supervisor
1045 # process has exited, and declare the instance broken.
1046 TimeoutStaleRunLock: 5s
1048 # Time to give up on SIGTERM and write off the worker.
1051 # Maximum create/destroy-instance operations per second (0 =
1053 MaxCloudOpsPerSecond: 0
1055 # Maximum concurrent node creation operations (0 = unlimited). This is
1056 # recommended by Azure in certain scenarios (see
1057 # https://docs.microsoft.com/en-us/azure/virtual-machines/linux/capture-image)
1058 # and can be used with other cloud providers too, if desired.
1059 MaxConcurrentInstanceCreateOps: 0
1061 # Interval between cloud provider syncs/updates ("list all
1065 # Time to leave an idle worker running (in case new containers
1066 # appear in the queue that it can run) before shutting it
1070 # Time to wait for a new worker to boot (i.e., pass
1071 # BootProbeCommand) before giving up and shutting it down.
1074 # Maximum time a worker can stay alive with no successful
1075 # probes before being automatically shut down.
1078 # Time after shutting down a worker to retry the
1079 # shutdown/destroy operation.
1080 TimeoutShutdown: 10s
1082 # Worker VM image ID.
1083 # (aws) AMI identifier
1084 # (azure) managed disks: the name of the managed disk image
1085 # (azure) shared image gallery: the name of the image definition. Also
1086 # see the SharedImageGalleryName and SharedImageGalleryImageVersion fields.
1087 # (azure) unmanaged disks (deprecated): the complete URI of the VHD, e.g.
1088 # https://xxxxx.blob.core.windows.net/system/Microsoft.Compute/Images/images/xxxxx.vhd
1091 # An executable file (located on the dispatcher host) to be
1092 # copied to cloud instances at runtime and used as the
1093 # container runner/supervisor. The default value is the
1094 # dispatcher program itself.
1096 # Use the empty string to disable this step: nothing will be
1097 # copied, and cloud instances are assumed to have a suitable
1098 # version of crunch-run installed; see CrunchRunCommand above.
1099 DeployRunnerBinary: "/proc/self/exe"
1101 # Tags to add on all resources (VMs, NICs, disks) created by
1102 # the container dispatcher. (Arvados's own tags --
1103 # InstanceType, IdleBehavior, and InstanceSecret -- will also
1108 # Prefix for predefined tags used by Arvados (InstanceSetID,
1109 # InstanceType, InstanceSecret, IdleBehavior). With the
1110 # default value "Arvados", tags are "ArvadosInstanceSetID",
1111 # "ArvadosInstanceSecret", etc.
1113 # This should only be changed while no cloud resources are in
1114 # use and the cloud dispatcher is not running. Otherwise,
1115 # VMs/resources that were added using the old tag prefix will
1116 # need to be detected and cleaned up manually.
1117 TagKeyPrefix: Arvados
1119 # Cloud driver: "azure" (Microsoft Azure) or "ec2" (Amazon AWS).
1122 # Cloud-specific driver parameters.
1125 # (ec2) Credentials. Omit or leave blank if using IAM role.
1129 # (ec2) Instance configuration.
1135 AdminUsername: debian
1137 # (azure) Credentials.
1143 # (azure) Instance configuration.
1144 CloudEnvironment: AzurePublicCloud
1147 # (azure) The resource group where the VM and virtual NIC will be
1151 # (azure) The resource group of the Network to use for the virtual
1152 # NIC (if different from ResourceGroup)
1153 NetworkResourceGroup: ""
1157 # (azure) managed disks: The resource group where the managed disk
1158 # image can be found (if different from ResourceGroup).
1159 ImageResourceGroup: ""
1161 # (azure) shared image gallery: the name of the gallery
1162 SharedImageGalleryName: ""
1163 # (azure) shared image gallery: the version of the image definition
1164 SharedImageGalleryImageVersion: ""
1166 # (azure) unmanaged disks (deprecated): Where to store the VM VHD blobs
1170 # (azure) How long to wait before deleting VHD and NIC
1171 # objects that are no longer being used.
1172 DeleteDanglingResourcesAfter: 20s
1174 # Account (that already exists in the VM image) that will be
1175 # set up with an ssh authorized key to allow the compute
1176 # dispatcher to connect.
1177 AdminUsername: arvados
1181 # Use the instance type name as the key (in place of "SAMPLE" in
1182 # this sample entry).
1184 # Cloud provider's instance type. Defaults to the configured type name.
1188 IncludedScratch: 16GB
1195 # AccessViaHosts specifies which keepstore processes can read
1196 # and write data on the volume.
1198 # For a local filesystem, AccessViaHosts has one entry,
1199 # indicating which server the filesystem is located on.
1201 # For a network-attached backend accessible by all keepstore
1202 # servers, like a cloud storage bucket or an NFS mount,
1203 # AccessViaHosts can be empty/omitted.
1205 # Further info/examples:
1206 # https://doc.arvados.org/install/configure-fs-storage.html
1207 # https://doc.arvados.org/install/configure-s3-object-storage.html
1208 # https://doc.arvados.org/install/configure-azure-blob-storage.html
1212 "http://host1.example:25107": {}
1220 # for s3 driver -- see
1221 # https://doc.arvados.org/install/configure-s3-object-storage.html
1224 SecretAccessKey: aaaaa
1228 LocationConstraint: false
1234 # Use aws-s3-go (v2) instead of goamz
1235 UseAWSS3v2Driver: false
1237 # For S3 driver, potentially unsafe tuning parameter,
1238 # intentionally excluded from main documentation.
1240 # Enable deletion (garbage collection) even when the
1241 # configured BlobTrashLifetime is zero. WARNING: eventual
1242 # consistency may result in race conditions that can cause
1243 # data loss. Do not enable this unless you understand and
1247 # for azure driver -- see
1248 # https://doc.arvados.org/install/configure-azure-blob-storage.html
1249 StorageAccountName: aaaaa
1250 StorageAccountKey: aaaaa
1251 StorageBaseURL: core.windows.net
1252 ContainerName: aaaaa
1254 ListBlobsRetryDelay: 10s
1255 ListBlobsMaxAttempts: 10
1257 WriteRaceInterval: 15s
1258 WriteRacePollTime: 1s
1260 # for local directory driver -- see
1261 # https://doc.arvados.org/install/configure-fs-storage.html
1262 Root: /var/lib/arvados/keep-data
1264 # For local directory driver, potentially confusing tuning
1265 # parameter, intentionally excluded from main documentation.
1267 # When true, read and write operations (for whole 64MiB
1268 # blocks) on an individual volume will queued and issued
1269 # serially. When false, read and write operations will be
1270 # issued concurrently.
1272 # May possibly improve throughput if you have physical spinning disks
1273 # and experience contention when there are multiple requests
1274 # to the same volume.
1276 # Otherwise, when using SSDs, RAID, or a shared network filesystem, you
1277 # should leave this alone.
1283 SendUserSetupNotificationEmail: true
1285 # Bug/issue report notification to and from addresses
1286 IssueReporterEmailFrom: "arvados@example.com"
1287 IssueReporterEmailTo: "arvados@example.com"
1288 SupportEmailAddress: "arvados@example.com"
1290 # Generic issue email from
1291 EmailFrom: "arvados@example.com"
1298 ActivateUsers: false
1300 # API endpoint host or host:port; default is {id}.arvadosapi.com
1301 Host: sample.arvadosapi.com
1303 # Perform a proxy request when a local client requests an
1304 # object belonging to this remote.
1307 # Default "https". Can be set to "http" for testing.
1310 # Disable TLS verify. Can be set to true for testing.
1313 # When users present tokens issued by this remote cluster, and
1314 # their accounts are active on the remote cluster, activate
1315 # them on this cluster too.
1316 ActivateUsers: false
1319 # Workbench1 configs
1321 ActivationContactLink: mailto:info@arvados.org
1322 ArvadosDocsite: https://doc.arvados.org
1323 ArvadosPublicDataDocURL: https://playground.arvados.org/projects/public
1324 ShowUserAgreementInline: false
1327 # Scratch directory used by the remote repository browsing
1328 # feature. If it doesn't exist, it (and any missing parents) will be
1329 # created using mkdir_p.
1330 RepositoryCache: /var/www/arvados-workbench/current/tmp/git
1332 # Below is a sample setting of user_profile_form_fields config parameter.
1333 # This configuration parameter should be set to either false (to disable) or
1334 # to a map as shown below.
1335 # Configure the map of input fields to be displayed in the profile page
1336 # using the attribute "key" for each of the input fields.
1337 # This sample shows configuration with one required and one optional form fields.
1338 # For each of these input fields:
1339 # You can specify "Type" as "text" or "select".
1340 # List the "Options" to be displayed for each of the "select" menu.
1341 # Set "Required" as "true" for any of these fields to make them required.
1342 # If any of the required fields are missing in the user's profile, the user will be
1343 # redirected to the profile page before they can access any Workbench features.
1344 UserProfileFormFields:
1347 FormFieldTitle: Best color
1348 FormFieldDescription: your favorite color
1357 # exampleTextValue: # key that will be set in properties
1359 # FormFieldTitle: ""
1360 # FormFieldDescription: ""
1363 # exampleOptionsValue:
1365 # FormFieldTitle: ""
1366 # FormFieldDescription: ""
1374 # Use "UserProfileFormMessage to configure the message you want
1375 # to display on the profile page.
1376 UserProfileFormMessage: 'Welcome to Arvados. All <span style="color:red">required fields</span> must be completed before you can proceed.'
1378 # Mimetypes of applications for which the view icon
1379 # would be enabled in a collection's show page.
1380 # It is sufficient to list only applications here.
1381 # No need to list text and image types.
1382 ApplicationMimetypesWithViewIcon:
1400 # The maximum number of bytes to load in the log viewer
1401 LogViewerMaxBytes: 1M
1403 # When anonymous_user_token is configured, show public projects page
1404 EnablePublicProjectsPage: true
1406 # By default, disable the "Getting Started" popup which is specific to Arvados playground
1407 EnableGettingStartedPopup: false
1409 # Ask Arvados API server to compress its response payloads.
1410 APIResponseCompression: true
1412 # Timeouts for API requests.
1413 APIClientConnectTimeout: 2m
1414 APIClientReceiveTimeout: 5m
1416 # Maximum number of historic log records of a running job to fetch
1417 # and display in the Log tab, while subscribing to web sockets.
1418 RunningJobLogRecordsToFetch: 2000
1420 # In systems with many shared projects, loading of dashboard and topnav
1421 # can be slow due to collections indexing; use the following parameters
1422 # to suppress these properties
1423 ShowRecentCollectionsOnDashboard: true
1424 ShowUserNotifications: true
1426 # Enable/disable "multi-site search" in top nav ("true"/"false"), or
1427 # a link to the multi-site search page on a "home" Workbench site.
1430 # https://workbench.zzzzz.arvadosapi.com/collections/multisite
1433 # Should workbench allow management of local git repositories? Set to false if
1434 # the jobs api is disabled and there are no local git repositories.
1437 SiteName: Arvados Workbench
1438 ProfilingEnabled: false
1440 # This is related to obsolete Google OpenID 1.0 login
1441 # but some workbench stuff still expects it to be set.
1442 DefaultOpenIdPrefix: "https://www.google.com/accounts/o8/id"
1444 # Workbench2 configs
1446 FileViewersConfigURL: ""
1448 # Idle time after which the user's session will be auto closed.
1449 # This feature is disabled when set to zero.
1452 # Workbench welcome screen, this is HTML text that will be
1453 # incorporated directly onto the page.
1455 <img src="/arvados-logo-big.png" style="width: 20%; float: right; padding: 1em;" />
1456 <h2>Please log in.</h2>
1458 <p>If you have never used Arvados Workbench before, logging in
1459 for the first time will automatically create a new
1462 <i>Arvados Workbench uses your information only for
1463 identification, and does not retrieve any other personal
1466 # Workbench screen displayed to inactive users. This is HTML
1467 # text that will be incorporated directly onto the page.
1469 <img src="/arvados-logo-big.png" style="width: 20%; float: right; padding: 1em;" />
1470 <h3>Hi! You're logged in, but...</h3>
1471 <p>Your account is inactive.</p>
1472 <p>An administrator must activate your account before you can get
1475 # Connecting to Arvados shell VMs tends to be site-specific.
1476 # Put any special instructions here. This is HTML text that will
1477 # be incorporated directly onto the Workbench page.
1479 <a href="https://doc.arvados.org/user/getting_started/ssh-access-unix.html">Accessing an Arvados VM with SSH</a> (generic instructions).
1480 Site configurations vary. Contact your local cluster administrator if you have difficulty accessing an Arvados shell node.
1482 # Sample text if you are using a "switchyard" ssh proxy.
1483 # Replace "zzzzz" with your Cluster ID.
1485 # <p>Add a section like this to your SSH configuration file ( <i>~/.ssh/config</i>):</p>
1488 # ServerAliveInterval 60
1489 # ProxyCommand ssh -p2222 turnout@switchyard.zzzzz.arvadosapi.com -x -a $SSH_PROXY_FLAGS %h
1492 # If you are using a switchyard ssh proxy, shell node hostnames
1493 # may require a special hostname suffix. In the sample ssh
1494 # configuration above, this would be ".zzzzz"
1495 # This is added to the hostname in the "command line" column
1496 # the Workbench "shell VMs" page.
1498 # If your shell nodes are directly accessible by users without a
1499 # proxy and have fully qualified host names, you should leave
1501 SSHHelpHostSuffix: ""
1503 # (Experimental) Restart services automatically when config file
1504 # changes are detected. Only supported by `arvados-server boot` in
1506 AutoReloadConfig: false