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
17 # Token to be included in all healthcheck requests. Disabled by default.
18 # Server expects request header of the format "Authorization: Bearer xxx"
23 # In each of the service sections below, the keys under
24 # InternalURLs are the endpoints where the service should be
25 # listening, and reachable from other hosts in the cluster.
28 "http://host1.example:12345": {}
29 "http://host2.example:12345":
30 # Rendezvous is normally empty/omitted. When changing the
31 # URL of a Keepstore service, Rendezvous should be set to
32 # the old URL (with trailing slash omitted) to preserve
33 # rendezvous ordering.
68 # Base URL for Workbench inline preview. If blank, use
69 # WebDAVDownload instead, and disable inline preview.
70 # If both are empty, downloading collections from workbench
73 # It is important to properly configure the download service
74 # to migitate cross-site-scripting (XSS) attacks. A HTML page
75 # can be stored in collection. If an attacker causes a victim
76 # to visit that page through Workbench, it will be rendered by
77 # the browser. If all collections are served at the same
78 # domain, the browser will consider collections as coming from
79 # the same origin and having access to the same browsing data,
80 # enabling malicious Javascript on that page to access Arvados
81 # on behalf of the victim.
83 # This is mitigating by having separate domains for each
84 # collection, or limiting preview to circumstances where the
85 # collection is not accessed with the user's regular
88 # Serve preview links using uuid or pdh in subdomain
89 # (requires wildcard DNS and TLS certificate)
90 # https://*.collections.uuid_prefix.arvadosapi.com
92 # Serve preview links using uuid or pdh in main domain
93 # (requires wildcard DNS and TLS certificate)
94 # https://*--collections.uuid_prefix.arvadosapi.com
96 # Serve preview links by setting uuid or pdh in the path.
97 # This configuration only allows previews of public data or
98 # collection-sharing links, because these use the anonymous
99 # user token or the token is already embedded in the URL.
100 # Other data must be handled as downloads via WebDAVDownload:
101 # https://collections.uuid_prefix.arvadosapi.com
107 # Base URL for download links. If blank, serve links to WebDAV
108 # with disposition=attachment query param. Unlike preview links,
109 # browsers do not render attachments, so there is no risk of XSS.
111 # If WebDAVDownload is blank, and WebDAV uses a
112 # single-origin form, then Workbench will show an error page
114 # Serve download links by setting uuid or pdh in the path:
115 # https://download.uuid_prefix.arvadosapi.com
127 # ShellInABox service endpoint URL for a given VM. If empty, do not
128 # offer web shell logins.
130 # E.g., using a path-based proxy server to forward connections to shell hosts:
131 # https://webshell.uuid_prefix.arvadosapi.com
133 # E.g., using a name-based proxy server to forward connections to shell hosts:
134 # https://*.webshell.uuid_prefix.arvadosapi.com
150 # max concurrent connections per arvados server daemon
153 # All parameters here are passed to the PG client library in a connection string;
154 # see https://www.postgresql.org/docs/current/static/libpq-connect.html#LIBPQ-PARAMKEYWORDS
162 # Maximum size (in bytes) allowed for a single API request. This
163 # limit is published in the discovery document for use by clients.
164 # Note: You must separately configure the upstream web server or
165 # proxy to actually enforce the desired maximum request size on the
167 MaxRequestSize: 134217728
169 # Limit the number of bytes read from the database during an index
170 # request (by retrieving and returning fewer rows than would
171 # normally be returned in a single response).
172 # Note 1: This setting never reduces the number of returned rows to
173 # zero, no matter how big the first data row is.
174 # Note 2: Currently, this is only checked against a specific set of
175 # columns that tend to get large (collections.manifest_text,
176 # containers.mounts, workflows.definition). Other fields (e.g.,
177 # "properties" hashes) are not counted against this limit.
178 MaxIndexDatabaseRead: 134217728
180 # Maximum number of items to return when responding to a APIs that
181 # can return partial result sets using limit and offset parameters
182 # (e.g., *.index, groups.contents). If a request specifies a "limit"
183 # parameter higher than this value, this value is used instead.
184 MaxItemsPerResponse: 1000
186 # Maximum number of concurrent requests to accept in a single
187 # service process, or 0 for no limit.
188 MaxConcurrentRequests: 0
190 # Maximum number of 64MiB memory buffers per Keepstore server process, or
191 # 0 for no limit. When this limit is reached, up to
192 # (MaxConcurrentRequests - MaxKeepBlobBuffers) HTTP requests requiring
193 # buffers (like GET and PUT) will wait for buffer space to be released.
194 # Any HTTP requests beyond MaxConcurrentRequests will receive an
195 # immediate 503 response.
197 # MaxKeepBlobBuffers should be set such that (MaxKeepBlobBuffers * 64MiB
198 # * 1.1) fits comfortably in memory. On a host dedicated to running
199 # Keepstore, divide total memory by 88MiB to suggest a suitable value.
200 # For example, if grep MemTotal /proc/meminfo reports MemTotal: 7125440
201 # kB, compute 7125440 / (88 * 1024)=79 and configure MaxBuffers: 79
202 MaxKeepBlobBuffers: 128
204 # API methods to disable. Disabled methods are not listed in the
205 # discovery document, and respond 404 to all requests.
206 # Example: {"jobs.create":{}, "pipeline_instances.create": {}}
209 # Interval (seconds) between asynchronous permission view updates. Any
210 # permission-updating API called with the 'async' parameter schedules a an
211 # update on the permission view in the future, if not already scheduled.
212 AsyncPermissionsUpdateInterval: 20s
214 # Maximum number of concurrent outgoing requests to make while
215 # serving a single incoming multi-cluster (federated) request.
216 MaxRequestAmplification: 4
218 # RailsSessionSecretToken is a string of alphanumeric characters
219 # used by Rails to sign session tokens. IMPORTANT: This is a
220 # site secret. It should be at least 50 characters.
221 RailsSessionSecretToken: ""
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 # In the default configuration, authentication happens through the Arvados SSO
265 # server, which uses OAuth2 against Google's servers, so in that case this
266 # should be an address associated with a Google account.
267 AutoAdminUserWithEmail: ""
269 # If AutoAdminFirstUser is set to true, the first user to log in when no
270 # other admin users exist will automatically become an admin user.
271 AutoAdminFirstUser: false
273 # Email address to notify whenever a user creates a profile for the
275 UserProfileNotificationAddress: ""
276 AdminNotifierEmailFrom: arvados@example.com
277 EmailSubjectPrefix: "[ARVADOS] "
278 UserNotifierEmailFrom: arvados@example.com
279 NewUserNotificationRecipients: {}
280 NewInactiveUserNotificationRecipients: {}
282 # Set AnonymousUserToken to enable anonymous user access. You can get
283 # the token by running "bundle exec ./script/get_anonymous_user_token.rb"
284 # in the directory where your API server is running.
285 AnonymousUserToken: ""
287 # If a new user has an alternate email address (local@domain)
288 # with the domain given here, its local part becomes the new
289 # user's default username. Otherwise, the user's primary email
291 PreferDomainForUsername: ""
294 # Time to keep audit logs, in seconds. (An audit log is a row added
295 # to the "logs" table in the PostgreSQL database each time an
296 # Arvados object is created, modified, or deleted.)
298 # Currently, websocket event notifications rely on audit logs, so
299 # this should not be set lower than 300 (5 minutes).
302 # Maximum number of log rows to delete in a single SQL transaction.
304 # If MaxDeleteBatch is 0, log entries will never be
305 # deleted by Arvados. Cleanup can be done by an external process
306 # without affecting any Arvados system processes, as long as very
307 # recent (<5 minutes old) logs are not deleted.
309 # 100000 is a reasonable batch size for most sites.
312 # Attributes to suppress in events and audit logs. Notably,
313 # specifying {"manifest_text": {}} here typically makes the database
314 # smaller and faster.
316 # Warning: Using any non-empty value here can have undesirable side
317 # effects for any client or component that relies on event logs.
318 # Use at your own risk.
319 UnloggedAttributes: {}
323 # Logging threshold: panic, fatal, error, warn, info, debug, or
327 # Logging format: json or text
330 # Maximum characters of (JSON-encoded) query parameters to include
331 # in each request log entry. When params exceed this size, they will
332 # be JSON-encoded, truncated to this size, and logged as
334 MaxRequestLogParamsSize: 2000
338 # Enable access controls for data stored in Keep. This should
339 # always be set to true on a production cluster.
342 # BlobSigningKey is a string of alphanumeric characters used to
343 # generate permission signatures for Keep locators. It must be
344 # identical to the permission key given to Keep. IMPORTANT: This
345 # is a site secret. It should be at least 50 characters.
347 # Modifying BlobSigningKey will invalidate all existing
348 # signatures, which can cause programs to fail (e.g., arv-put,
349 # arv-get, and Crunch jobs). To avoid errors, rotate keys only
350 # when no such processes are running.
353 # Enable garbage collection of unreferenced blobs in Keep.
356 # Time to leave unreferenced blobs in "trashed" state before
357 # deleting them, or 0 to skip the "trashed" state entirely and
358 # delete unreferenced blobs.
360 # If you use any Amazon S3 buckets as storage volumes, this
361 # must be at least 24h to avoid occasional data loss.
362 BlobTrashLifetime: 336h
364 # How often to check for (and delete) trashed blocks whose
365 # BlobTrashLifetime has expired.
366 BlobTrashCheckInterval: 24h
368 # Maximum number of concurrent "trash blob" and "delete trashed
369 # blob" operations conducted by a single keepstore process. Each
370 # of these can be set to 0 to disable the respective operation.
372 # If BlobTrashLifetime is zero, "trash" and "delete trash"
373 # happen at once, so only the lower of these two values is used.
374 BlobTrashConcurrency: 4
375 BlobDeleteConcurrency: 4
377 # Maximum number of concurrent "create additional replica of
378 # existing blob" operations conducted by a single keepstore
380 BlobReplicateConcurrency: 4
382 # Default replication level for collections. This is used when a
383 # collection's replication_desired attribute is nil.
384 DefaultReplication: 2
386 # BlobSigningTTL determines the minimum lifetime of transient
387 # data, i.e., blocks that are not referenced by
388 # collections. Unreferenced blocks exist for two reasons:
390 # 1) A data block must be written to a disk/cloud backend device
391 # before a collection can be created/updated with a reference to
394 # 2) Deleting or updating a collection can remove the last
395 # remaining reference to a data block.
397 # If BlobSigningTTL is too short, long-running
398 # processes/containers will fail when they take too long (a)
399 # between writing blocks and writing collections that reference
400 # them, or (b) between reading collections and reading the
403 # If BlobSigningTTL is too long, data will still be stored long
404 # after the referring collections are deleted, and you will
405 # needlessly fill up disks or waste money on cloud storage.
407 # Modifying BlobSigningTTL invalidates existing signatures; see
408 # BlobSigningKey note above.
410 # The default is 2 weeks.
413 # When running keep-balance, this is the destination filename for
414 # the list of lost block hashes if there are any, one per line.
415 # Updated automically during each successful run.
416 BlobMissingReport: ""
418 # keep-balance operates periodically, i.e.: do a
419 # scan/balance operation, sleep, repeat.
421 # BalancePeriod determines the interval between start times of
422 # successive scan/balance operations. If a scan/balance operation
423 # takes longer than RunPeriod, the next one will follow it
426 # If SIGUSR1 is received during an idle period between operations,
427 # the next operation will start immediately.
430 # Limits the number of collections retrieved by keep-balance per
431 # API transaction. If this is zero, page size is
432 # determined by the API server's own page size limits (see
433 # API.MaxItemsPerResponse and API.MaxIndexDatabaseRead).
434 BalanceCollectionBatch: 0
436 # The size of keep-balance's internal queue of
437 # collections. Higher values use more memory and improve throughput
438 # by allowing keep-balance to fetch the next page of collections
439 # while the current page is still being processed. If this is zero
440 # or omitted, pages are processed serially.
441 BalanceCollectionBuffers: 1000
443 # Maximum time for a rebalancing run. This ensures keep-balance
444 # eventually gives up and retries if, for example, a network
445 # error causes a hung connection that is never closed by the
446 # OS. It should be long enough that it doesn't interrupt a
447 # long-running balancing operation.
450 # Default lifetime for ephemeral collections: 2 weeks. This must not
451 # be less than BlobSigningTTL.
452 DefaultTrashLifetime: 336h
454 # Interval (seconds) between trash sweeps. During a trash sweep,
455 # collections are marked as trash if their trash_at time has
456 # arrived, and deleted if their delete_at time has arrived.
457 TrashSweepInterval: 60s
459 # If true, enable collection versioning.
460 # When a collection's preserve_version field is true or the current version
461 # is older than the amount of seconds defined on PreserveVersionIfIdle,
462 # a snapshot of the collection's previous state is created and linked to
463 # the current collection.
464 CollectionVersioning: false
466 # 0s = auto-create a new version on every update.
467 # -1s = never auto-create new versions.
468 # > 0s = auto-create a new version when older than the specified number of seconds.
469 PreserveVersionIfIdle: -1s
471 # If non-empty, allow project and collection names to contain
472 # the "/" character (slash/stroke/solidus), and replace "/" with
473 # the given string in the filesystem hierarchy presented by
474 # WebDAV. Example values are "%2f" and "{slash}". Names that
475 # contain the substitution string itself may result in confusing
476 # behavior, so a value like "_" is not recommended.
478 # If the default empty value is used, the server will reject
479 # requests to create or rename a collection when the new name
482 # If the value "/" is used, project and collection names
483 # containing "/" will be allowed, but they will not be
484 # accessible via WebDAV.
486 # Use of this feature is not recommended, if it can be avoided.
487 ForwardSlashNameSubstitution: ""
489 # Include "folder objects" in S3 ListObjects responses.
490 S3FolderObjects: true
492 # Managed collection properties. At creation time, if the client didn't
493 # provide the listed keys, they will be automatically populated following
494 # one of the following behaviors:
496 # * UUID of the user who owns the containing project.
497 # responsible_person_uuid: {Function: original_owner, Protected: true}
499 # * Default concrete value.
500 # foo_bar: {Value: baz, Protected: false}
502 # If Protected is true, only an admin user can modify its value.
504 SAMPLE: {Function: original_owner, Protected: true}
506 # In "trust all content" mode, Workbench will redirect download
507 # requests to WebDAV preview link, even in the cases when
508 # WebDAV would have to expose XSS vulnerabilities in order to
509 # handle the redirect (see discussion on Services.WebDAV).
511 # This setting has no effect in the recommended configuration,
512 # where the WebDAV is configured to have a separate domain for
513 # every collection; in this case XSS protection is provided by
514 # browsers' same-origin policy.
516 # The default setting (false) is appropriate for a multi-user site.
517 TrustAllContent: false
519 # Cache parameters for WebDAV content serving:
520 # * TTL: Maximum time to cache manifests and permission checks.
521 # * UUIDTTL: Maximum time to cache collection state.
522 # * MaxBlockEntries: Maximum number of block cache entries.
523 # * MaxCollectionEntries: Maximum number of collection cache entries.
524 # * MaxCollectionBytes: Approximate memory limit for collection cache.
525 # * MaxPermissionEntries: Maximum number of permission cache entries.
526 # * MaxUUIDEntries: Maximum number of UUID cache entries.
531 MaxCollectionEntries: 1000
532 MaxCollectionBytes: 100000000
533 MaxPermissionEntries: 1000
537 # One of the following mechanisms (SSO, Google, PAM, LDAP, or
538 # LoginCluster) should be enabled; see
539 # https://doc.arvados.org/install/setup-login.html
542 # Authenticate with Google.
545 # Use the Google Cloud console to enable the People API (APIs
546 # and Services > Enable APIs and services > Google People API
547 # > Enable), generate a Client ID and secret (APIs and
548 # Services > Credentials > Create credentials > OAuth client
549 # ID > Web application) and add your controller's /login URL
550 # (e.g., "https://zzzzz.example.com/login") as an authorized
553 # Incompatible with ForceLegacyAPI14. ProviderAppID must be
558 # Allow users to log in to existing accounts using any verified
559 # email address listed by their Google account. If true, the
560 # Google People API must be enabled in order for Google login to
561 # work. If false, only the primary email address will be used.
562 AlternateEmailAddresses: true
565 # Authenticate with an OpenID Connect provider.
568 # Issuer URL, e.g., "https://login.example.com".
570 # This must be exactly equal to the URL returned by the issuer
571 # itself in its config response ("isser" key). If the
572 # configured value is "https://example" and the provider
573 # returns "https://example:443" or "https://example/" then
574 # login will fail, even though those URLs are equivalent
578 # Your client ID and client secret (supplied by the provider).
582 # OpenID claim field containing the user's email
583 # address. Normally "email"; see
584 # https://openid.net/specs/openid-connect-core-1_0.html#StandardClaims
587 # OpenID claim field containing the email verification
588 # flag. Normally "email_verified". To accept every returned
589 # email address without checking a "verified" field at all,
590 # use the empty string "".
591 EmailVerifiedClaim: "email_verified"
593 # OpenID claim field containing the user's preferred
594 # username. If empty, use the mailbox part of the user's email
599 # (Experimental) Use PAM to authenticate users.
602 # PAM service name. PAM will apply the policy in the
603 # corresponding config file (e.g., /etc/pam.d/arvados) or, if
604 # there is none, the default "other" config.
607 # Domain name (e.g., "example.com") to use to construct the
608 # user's email address if PAM authentication returns a
609 # username with no "@". If empty, use the PAM username as the
610 # user's email address, whether or not it contains "@".
612 # Note that the email address is used as the primary key for
613 # user records when logging in. Therefore, if you change
614 # PAMDefaultEmailDomain after the initial installation, you
615 # should also update existing user records to reflect the new
616 # domain. Otherwise, next time those users log in, they will
617 # be given new accounts instead of accessing their existing
619 DefaultEmailDomain: ""
622 # Use an LDAP service to authenticate users.
625 # Server URL, like "ldap://ldapserver.example.com:389" or
626 # "ldaps://ldapserver.example.com:636".
627 URL: "ldap://ldap:389"
629 # Use StartTLS upon connecting to the server.
632 # Skip TLS certificate name verification.
635 # Strip the @domain part if a user supplies an email-style
636 # username with this domain. If "*", strip any user-provided
637 # domain. If "", never strip the domain part. Example:
641 # If, after applying StripDomain, the username contains no "@"
642 # character, append this domain to form an email-style
643 # username. Example: "example.com"
646 # The LDAP attribute to filter on when looking up a username
647 # (after applying StripDomain and AppendDomain).
650 # Bind with this username (DN or UPN) and password when
651 # looking up the user record.
653 # Example user: "cn=admin,dc=example,dc=com"
655 SearchBindPassword: ""
657 # Directory base for username lookup. Example:
658 # "ou=Users,dc=example,dc=com"
661 # Additional filters to apply when looking up users' LDAP
662 # entries. This can be used to restrict access to a subset of
663 # LDAP users, or to disambiguate users from other directory
664 # entries that have the SearchAttribute present.
666 # Special characters in assertion values must be escaped (see
669 # Example: "(objectClass=person)"
672 # LDAP attribute to use as the user's email address.
674 # Important: This must not be an attribute whose value can be
675 # edited in the directory by the users themselves. Otherwise,
676 # users can take over other users' Arvados accounts trivially
677 # (email address is the primary key for Arvados accounts.)
680 # LDAP attribute to use as the preferred Arvados username. If
681 # no value is found (or this config is empty) the username
682 # originally supplied by the user will be used.
683 UsernameAttribute: uid
686 # Authenticate with a separate SSO server. (Deprecated)
689 # ProviderAppID and ProviderAppSecret are generated during SSO
691 # https://doc.arvados.org/v2.0/install/install-sso.html#update-config
693 ProviderAppSecret: ""
695 # The cluster ID to delegate the user database. When set,
696 # logins on this cluster will be redirected to the login cluster
697 # (login cluster must appear in RemoteClusters with Proxy: true)
700 # How long a cached token belonging to a remote cluster will
701 # remain valid before it needs to be revalidated.
702 RemoteTokenRefresh: 5m
705 # Path to git or gitolite-shell executable. Each authenticated
706 # request will execute this program with the single argument "http-backend"
707 GitCommand: /usr/bin/git
709 # Path to Gitolite's home directory. If a non-empty path is given,
710 # the CGI environment will be set up to support the use of
711 # gitolite-shell as a GitCommand: for example, if GitoliteHome is
712 # "/gh", then the CGI environment will have GITOLITE_HTTP_HOME=/gh,
713 # PATH=$PATH:/gh/bin, and GL_BYPASS_ACCESS_CHECKS=1.
716 # Git repositories must be readable by api server, or you won't be
717 # able to submit crunch jobs. To pass the test suites, put a clone
718 # of the arvados tree in {git_repositories_dir}/arvados.git or
719 # {git_repositories_dir}/arvados/.git
720 Repositories: /var/lib/arvados/git/repositories
728 # List of supported Docker Registry image formats that compute nodes
729 # are able to use. `arv keep docker` will error out if a user tries
730 # to store an image with an unsupported format. Use an empty array
731 # to skip the compatibility check (and display a warning message to
734 # Example for sites running docker < 1.10: {"v1": {}}
735 # Example for sites running docker >= 1.10: {"v2": {}}
736 # Example for disabling check: {}
737 SupportedDockerImageFormats:
741 # Include details about job reuse decisions in the server log. This
742 # causes additional database queries to run, so it should not be
743 # enabled unless you expect to examine the resulting logs for
744 # troubleshooting purposes.
745 LogReuseDecisions: false
747 # Default value for keep_cache_ram of a container's runtime_constraints.
748 DefaultKeepCacheRAM: 268435456
750 # Number of times a container can be unlocked before being
751 # automatically cancelled.
752 MaxDispatchAttempts: 5
754 # Default value for container_count_max for container requests. This is the
755 # number of times Arvados will create a new container to satisfy a container
756 # request. If a container is cancelled it will retry a new container if
757 # container_count < container_count_max on any container requests associated
758 # with the cancelled container.
761 # The maximum number of compute nodes that can be in use simultaneously
762 # If this limit is reduced, any existing nodes with slot number >= new limit
763 # will not be counted against the new limit. In other words, the new limit
764 # won't be strictly enforced until those nodes with higher slot numbers
768 # Preemptible instance support (e.g. AWS Spot Instances)
769 # When true, child containers will get created with the preemptible
770 # scheduling parameter parameter set.
771 UsePreemptibleInstances: false
773 # PEM encoded SSH key (RSA, DSA, or ECDSA) used by the
774 # (experimental) cloud dispatcher for executing containers on
775 # worker VMs. Begins with "-----BEGIN RSA PRIVATE KEY-----\n"
776 # and ends with "\n-----END RSA PRIVATE KEY-----\n".
777 DispatchPrivateKey: ""
779 # Maximum time to wait for workers to come up before abandoning
780 # stale locks from a previous dispatch process.
783 # The crunch-run command to manage the container on a node
784 CrunchRunCommand: "crunch-run"
786 # Extra arguments to add to crunch-run invocation
787 # Example: ["--cgroup-parent-subsystem=memory"]
788 CrunchRunArgumentsList: []
790 # Extra RAM to reserve on the node, in addition to
791 # the amount specified in the container's RuntimeConstraints
792 ReserveExtraRAM: 256MiB
794 # Minimum time between two attempts to run the same container
798 # When you run the db:delete_old_container_logs task, it will find
799 # containers that have been finished for at least this many seconds,
800 # and delete their stdout, stderr, arv-mount, crunch-run, and
801 # crunchstat logs from the logs table.
804 # These two settings control how frequently log events are flushed to the
805 # database. Log lines are buffered until either crunch_log_bytes_per_event
806 # has been reached or crunch_log_seconds_between_events has elapsed since
808 LogBytesPerEvent: 4096
809 LogSecondsBetweenEvents: 5s
811 # The sample period for throttling logs.
812 LogThrottlePeriod: 60s
814 # Maximum number of bytes that job can log over crunch_log_throttle_period
815 # before being silenced until the end of the period.
816 LogThrottleBytes: 65536
818 # Maximum number of lines that job can log over crunch_log_throttle_period
819 # before being silenced until the end of the period.
820 LogThrottleLines: 1024
822 # Maximum bytes that may be logged by a single job. Log bytes that are
823 # silenced by throttling are not counted against this total.
824 LimitLogBytesPerJob: 67108864
826 LogPartialLineThrottlePeriod: 5s
828 # Container logs are written to Keep and saved in a
829 # collection, which is updated periodically while the
830 # container runs. This value sets the interval between
831 # collection updates.
834 # The log collection is also updated when the specified amount of
835 # log data (given in bytes) is produced in less than one update
841 SbatchArgumentsList: []
842 SbatchEnvironmentVariables:
845 # Path to dns server configuration directory
846 # (e.g. /etc/unbound.d/conf.d). If false, do not write any config
847 # files or touch restart.txt (see below).
850 # Template file for the dns server host snippets. See
851 # unbound.template in this directory for an example. If false, do
852 # not write any config files.
853 DNSServerConfTemplate: ""
855 # String to write to {dns_server_conf_dir}/restart.txt (with a
856 # trailing newline) after updating local data. If false, do not
857 # open or write the restart.txt file.
858 DNSServerReloadCommand: ""
860 # Command to run after each DNS update. Template variables will be
861 # substituted; see the "unbound" example below. If false, do not run
863 DNSServerUpdateCommand: ""
865 ComputeNodeDomain: ""
866 ComputeNodeNameservers:
870 # Hostname to assign to a compute node when it sends a "ping" and the
871 # hostname in its Node record is nil.
872 # During bootstrapping, the "ping" script is expected to notice the
873 # hostname given in the ping response, and update its unix hostname
875 # If false, leave the hostname alone (this is appropriate if your compute
876 # nodes' hostnames are already assigned by some other mechanism).
878 # One way or another, the hostnames of your node records should agree
879 # with your DNS records and your /etc/slurm-llnl/slurm.conf files.
881 # Example for compute0000, compute0001, ....:
882 # assign_node_hostname: compute%<slot_number>04d
883 # (See http://ruby-doc.org/core-2.2.2/Kernel.html#method-i-format for more.)
884 AssignNodeHostname: "compute%<slot_number>d"
887 # Enable the legacy 'jobs' API (crunch v1). This value must be a string.
889 # Note: this only enables read-only access, creating new
890 # legacy jobs and pipelines is not supported.
892 # 'auto' -- (default) enable the Jobs API only if it has been used before
893 # (i.e., there are job records in the database)
894 # 'true' -- enable the Jobs API despite lack of existing records.
895 # 'false' -- disable the Jobs API despite presence of existing records.
898 # Git repositories must be readable by api server, or you won't be
899 # able to submit crunch jobs. To pass the test suites, put a clone
900 # of the arvados tree in {git_repositories_dir}/arvados.git or
901 # {git_repositories_dir}/arvados/.git
902 GitInternalDir: /var/lib/arvados/internal.git
905 # Enable the cloud scheduler (experimental).
908 # Name/number of port where workers' SSH services listen.
911 # Interval between queue polls.
914 # Shell command to execute on each worker to determine whether
915 # the worker is booted and ready to run containers. It should
916 # exit zero if the worker is ready.
917 BootProbeCommand: "docker ps -q"
919 # Minimum interval between consecutive probes to a single
923 # Maximum probes per second, across all workers in a pool.
924 MaxProbesPerSecond: 10
926 # Time before repeating SIGTERM when killing a container.
929 # Time to give up on SIGTERM and write off the worker.
932 # Maximum create/destroy-instance operations per second (0 =
934 MaxCloudOpsPerSecond: 0
936 # Interval between cloud provider syncs/updates ("list all
940 # Time to leave an idle worker running (in case new containers
941 # appear in the queue that it can run) before shutting it
945 # Time to wait for a new worker to boot (i.e., pass
946 # BootProbeCommand) before giving up and shutting it down.
949 # Maximum time a worker can stay alive with no successful
950 # probes before being automatically shut down.
953 # Time after shutting down a worker to retry the
954 # shutdown/destroy operation.
957 # Worker VM image ID.
960 # An executable file (located on the dispatcher host) to be
961 # copied to cloud instances at runtime and used as the
962 # container runner/supervisor. The default value is the
963 # dispatcher program itself.
965 # Use the empty string to disable this step: nothing will be
966 # copied, and cloud instances are assumed to have a suitable
967 # version of crunch-run installed.
968 DeployRunnerBinary: "/proc/self/exe"
970 # Tags to add on all resources (VMs, NICs, disks) created by
971 # the container dispatcher. (Arvados's own tags --
972 # InstanceType, IdleBehavior, and InstanceSecret -- will also
977 # Prefix for predefined tags used by Arvados (InstanceSetID,
978 # InstanceType, InstanceSecret, IdleBehavior). With the
979 # default value "Arvados", tags are "ArvadosInstanceSetID",
980 # "ArvadosInstanceSecret", etc.
982 # This should only be changed while no cloud resources are in
983 # use and the cloud dispatcher is not running. Otherwise,
984 # VMs/resources that were added using the old tag prefix will
985 # need to be detected and cleaned up manually.
986 TagKeyPrefix: Arvados
988 # Cloud driver: "azure" (Microsoft Azure) or "ec2" (Amazon AWS).
991 # Cloud-specific driver parameters.
998 # (ec2) Instance configuration.
1004 AdminUsername: debian
1006 # (azure) Credentials.
1012 # (azure) Instance configuration.
1013 CloudEnvironment: AzurePublicCloud
1020 DeleteDanglingResourcesAfter: 20s
1021 AdminUsername: arvados
1025 # Use the instance type name as the key (in place of "SAMPLE" in
1026 # this sample entry).
1028 # Cloud provider's instance type. Defaults to the configured type name.
1032 IncludedScratch: 16GB
1039 # AccessViaHosts specifies which keepstore processes can read
1040 # and write data on the volume.
1042 # For a local filesystem, AccessViaHosts has one entry,
1043 # indicating which server the filesystem is located on.
1045 # For a network-attached backend accessible by all keepstore
1046 # servers, like a cloud storage bucket or an NFS mount,
1047 # AccessViaHosts can be empty/omitted.
1049 # Further info/examples:
1050 # https://doc.arvados.org/install/configure-fs-storage.html
1051 # https://doc.arvados.org/install/configure-s3-object-storage.html
1052 # https://doc.arvados.org/install/configure-azure-blob-storage.html
1056 "http://host1.example:25107": {}
1064 # for s3 driver -- see
1065 # https://doc.arvados.org/install/configure-s3-object-storage.html
1072 LocationConstraint: false
1079 # For S3 driver, potentially unsafe tuning parameter,
1080 # intentionally excluded from main documentation.
1082 # Enable deletion (garbage collection) even when the
1083 # configured BlobTrashLifetime is zero. WARNING: eventual
1084 # consistency may result in race conditions that can cause
1085 # data loss. Do not enable this unless you understand and
1089 # for azure driver -- see
1090 # https://doc.arvados.org/install/configure-azure-blob-storage.html
1091 StorageAccountName: aaaaa
1092 StorageAccountKey: aaaaa
1093 StorageBaseURL: core.windows.net
1094 ContainerName: aaaaa
1096 ListBlobsRetryDelay: 10s
1097 ListBlobsMaxAttempts: 10
1099 WriteRaceInterval: 15s
1100 WriteRacePollTime: 1s
1102 # for local directory driver -- see
1103 # https://doc.arvados.org/install/configure-fs-storage.html
1104 Root: /var/lib/arvados/keep-data
1106 # For local directory driver, potentially confusing tuning
1107 # parameter, intentionally excluded from main documentation.
1109 # When true, read and write operations (for whole 64MiB
1110 # blocks) on an individual volume will queued and issued
1111 # serially. When false, read and write operations will be
1112 # issued concurrently.
1114 # May possibly improve throughput if you have physical spinning disks
1115 # and experience contention when there are multiple requests
1116 # to the same volume.
1118 # Otherwise, when using SSDs, RAID, or a shared network filesystem, you
1119 # should leave this alone.
1125 SendUserSetupNotificationEmail: true
1127 # Bug/issue report notification to and from addresses
1128 IssueReporterEmailFrom: "arvados@example.com"
1129 IssueReporterEmailTo: "arvados@example.com"
1130 SupportEmailAddress: "arvados@example.com"
1132 # Generic issue email from
1133 EmailFrom: "arvados@example.com"
1140 ActivateUsers: false
1142 # API endpoint host or host:port; default is {id}.arvadosapi.com
1143 Host: sample.arvadosapi.com
1145 # Perform a proxy request when a local client requests an
1146 # object belonging to this remote.
1149 # Default "https". Can be set to "http" for testing.
1152 # Disable TLS verify. Can be set to true for testing.
1155 # When users present tokens issued by this remote cluster, and
1156 # their accounts are active on the remote cluster, activate
1157 # them on this cluster too.
1158 ActivateUsers: false
1161 # Workbench1 configs
1163 ActivationContactLink: mailto:info@arvados.org
1164 ArvadosDocsite: https://doc.arvados.org
1165 ArvadosPublicDataDocURL: https://playground.arvados.org/projects/public
1166 ShowUserAgreementInline: false
1169 # Scratch directory used by the remote repository browsing
1170 # feature. If it doesn't exist, it (and any missing parents) will be
1171 # created using mkdir_p.
1172 RepositoryCache: /var/www/arvados-workbench/current/tmp/git
1174 # Below is a sample setting of user_profile_form_fields config parameter.
1175 # This configuration parameter should be set to either false (to disable) or
1176 # to a map as shown below.
1177 # Configure the map of input fields to be displayed in the profile page
1178 # using the attribute "key" for each of the input fields.
1179 # This sample shows configuration with one required and one optional form fields.
1180 # For each of these input fields:
1181 # You can specify "Type" as "text" or "select".
1182 # List the "Options" to be displayed for each of the "select" menu.
1183 # Set "Required" as "true" for any of these fields to make them required.
1184 # If any of the required fields are missing in the user's profile, the user will be
1185 # redirected to the profile page before they can access any Workbench features.
1186 UserProfileFormFields:
1189 FormFieldTitle: Best color
1190 FormFieldDescription: your favorite color
1199 # exampleTextValue: # key that will be set in properties
1201 # FormFieldTitle: ""
1202 # FormFieldDescription: ""
1205 # exampleOptionsValue:
1207 # FormFieldTitle: ""
1208 # FormFieldDescription: ""
1216 # Use "UserProfileFormMessage to configure the message you want
1217 # to display on the profile page.
1218 UserProfileFormMessage: 'Welcome to Arvados. All <span style="color:red">required fields</span> must be completed before you can proceed.'
1220 # Mimetypes of applications for which the view icon
1221 # would be enabled in a collection's show page.
1222 # It is sufficient to list only applications here.
1223 # No need to list text and image types.
1224 ApplicationMimetypesWithViewIcon:
1242 # The maximum number of bytes to load in the log viewer
1243 LogViewerMaxBytes: 1M
1245 # When anonymous_user_token is configured, show public projects page
1246 EnablePublicProjectsPage: true
1248 # By default, disable the "Getting Started" popup which is specific to Arvados playground
1249 EnableGettingStartedPopup: false
1251 # Ask Arvados API server to compress its response payloads.
1252 APIResponseCompression: true
1254 # Timeouts for API requests.
1255 APIClientConnectTimeout: 2m
1256 APIClientReceiveTimeout: 5m
1258 # Maximum number of historic log records of a running job to fetch
1259 # and display in the Log tab, while subscribing to web sockets.
1260 RunningJobLogRecordsToFetch: 2000
1262 # In systems with many shared projects, loading of dashboard and topnav
1263 # can be slow due to collections indexing; use the following parameters
1264 # to suppress these properties
1265 ShowRecentCollectionsOnDashboard: true
1266 ShowUserNotifications: true
1268 # Enable/disable "multi-site search" in top nav ("true"/"false"), or
1269 # a link to the multi-site search page on a "home" Workbench site.
1272 # https://workbench.qr1hi.arvadosapi.com/collections/multisite
1275 # Should workbench allow management of local git repositories? Set to false if
1276 # the jobs api is disabled and there are no local git repositories.
1279 SiteName: Arvados Workbench
1280 ProfilingEnabled: false
1282 # This is related to obsolete Google OpenID 1.0 login
1283 # but some workbench stuff still expects it to be set.
1284 DefaultOpenIdPrefix: "https://www.google.com/accounts/o8/id"
1286 # Workbench2 configs
1288 FileViewersConfigURL: ""
1290 # Workbench welcome screen, this is HTML text that will be
1291 # incorporated directly onto the page.
1293 <img src="/arvados-logo-big.png" style="width: 20%; float: right; padding: 1em;" />
1294 <h2>Please log in.</h2>
1296 <p>The "Log in" button below will show you a sign-in
1297 page. After you log in, you will be redirected back to
1298 Arvados Workbench.</p>
1300 <p>If you have never used Arvados Workbench before, logging in
1301 for the first time will automatically create a new
1304 <i>Arvados Workbench uses your name and email address only for
1305 identification, and does not retrieve any other personal
1308 # Workbench screen displayed to inactive users. This is HTML
1309 # text that will be incorporated directly onto the page.
1311 <img src="/arvados-logo-big.png" style="width: 20%; float: right; padding: 1em;" />
1312 <h3>Hi! You're logged in, but...</h3>
1313 <p>Your account is inactive.</p>
1314 <p>An administrator must activate your account before you can get
1317 # Connecting to Arvados shell VMs tends to be site-specific.
1318 # Put any special instructions here. This is HTML text that will
1319 # be incorporated directly onto the Workbench page.
1321 <a href="https://doc.arvados.org/user/getting_started/ssh-access-unix.html">Accessing an Arvados VM with SSH</a> (generic instructions).
1322 Site configurations vary. Contact your local cluster administrator if you have difficulty accessing an Arvados shell node.
1324 # Sample text if you are using a "switchyard" ssh proxy.
1325 # Replace "zzzzz" with your Cluster ID.
1327 # <p>Add a section like this to your SSH configuration file ( <i>~/.ssh/config</i>):</p>
1330 # ServerAliveInterval 60
1331 # ProxyCommand ssh -p2222 turnout@switchyard.zzzzz.arvadosapi.com -x -a $SSH_PROXY_FLAGS %h
1334 # If you are using a switchyard ssh proxy, shell node hostnames
1335 # may require a special hostname suffix. In the sample ssh
1336 # configuration above, this would be ".zzzzz"
1337 # This is added to the hostname in the "command line" column
1338 # the Workbench "shell VMs" page.
1340 # If your shell nodes are directly accessible by users without a
1341 # proxy and have fully qualified host names, you should leave
1343 SSHHelpHostSuffix: ""
1345 # Bypass new (Arvados 1.5) API implementations, and hand off
1346 # requests directly to Rails instead. This can provide a temporary
1347 # workaround for clients that are incompatible with the new API
1348 # implementation. Note that it also disables some new federation
1349 # features and will be removed in a future release.
1350 ForceLegacyAPI14: false
1352 # (Experimental) Restart services automatically when config file
1353 # changes are detected. Only supported by `arvados-server boot` in
1355 AutoReloadConfig: false