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
147 # max concurrent connections per arvados server daemon
150 # All parameters here are passed to the PG client library in a connection string;
151 # see https://www.postgresql.org/docs/current/static/libpq-connect.html#LIBPQ-PARAMKEYWORDS
159 # Maximum size (in bytes) allowed for a single API request. This
160 # limit is published in the discovery document for use by clients.
161 # Note: You must separately configure the upstream web server or
162 # proxy to actually enforce the desired maximum request size on the
164 MaxRequestSize: 134217728
166 # Limit the number of bytes read from the database during an index
167 # request (by retrieving and returning fewer rows than would
168 # normally be returned in a single response).
169 # Note 1: This setting never reduces the number of returned rows to
170 # zero, no matter how big the first data row is.
171 # Note 2: Currently, this is only checked against a specific set of
172 # columns that tend to get large (collections.manifest_text,
173 # containers.mounts, workflows.definition). Other fields (e.g.,
174 # "properties" hashes) are not counted against this limit.
175 MaxIndexDatabaseRead: 134217728
177 # Maximum number of items to return when responding to a APIs that
178 # can return partial result sets using limit and offset parameters
179 # (e.g., *.index, groups.contents). If a request specifies a "limit"
180 # parameter higher than this value, this value is used instead.
181 MaxItemsPerResponse: 1000
183 # Maximum number of concurrent requests to accept in a single
184 # service process, or 0 for no limit.
185 MaxConcurrentRequests: 0
187 # Maximum number of 64MiB memory buffers per Keepstore server process, or
188 # 0 for no limit. When this limit is reached, up to
189 # (MaxConcurrentRequests - MaxKeepBlobBuffers) HTTP requests requiring
190 # buffers (like GET and PUT) will wait for buffer space to be released.
191 # Any HTTP requests beyond MaxConcurrentRequests will receive an
192 # immediate 503 response.
194 # MaxKeepBlobBuffers should be set such that (MaxKeepBlobBuffers * 64MiB
195 # * 1.1) fits comfortably in memory. On a host dedicated to running
196 # Keepstore, divide total memory by 88MiB to suggest a suitable value.
197 # For example, if grep MemTotal /proc/meminfo reports MemTotal: 7125440
198 # kB, compute 7125440 / (88 * 1024)=79 and configure MaxBuffers: 79
199 MaxKeepBlobBuffers: 128
201 # API methods to disable. Disabled methods are not listed in the
202 # discovery document, and respond 404 to all requests.
203 # Example: {"jobs.create":{}, "pipeline_instances.create": {}}
206 # Interval (seconds) between asynchronous permission view updates. Any
207 # permission-updating API called with the 'async' parameter schedules a an
208 # update on the permission view in the future, if not already scheduled.
209 AsyncPermissionsUpdateInterval: 20s
211 # Maximum number of concurrent outgoing requests to make while
212 # serving a single incoming multi-cluster (federated) request.
213 MaxRequestAmplification: 4
215 # RailsSessionSecretToken is a string of alphanumeric characters
216 # used by Rails to sign session tokens. IMPORTANT: This is a
217 # site secret. It should be at least 50 characters.
218 RailsSessionSecretToken: ""
220 # Maximum wall clock time to spend handling an incoming request.
223 # Websocket will send a periodic empty event after 'SendTimeout'
224 # if there is no other activity to maintain the connection /
225 # detect dropped connections.
228 WebsocketClientEventQueue: 64
229 WebsocketServerEventQueue: 4
231 # Timeout on requests to internal Keep services.
232 KeepServiceRequestTimeout: 15s
235 # Config parameters to automatically setup new users. If enabled,
236 # this users will be able to self-activate. Enable this if you want
237 # to run an open instance where anyone can create an account and use
238 # the system without requiring manual approval.
240 # The params AutoSetupNewUsersWith* are meaningful only when AutoSetupNewUsers is turned on.
241 # AutoSetupUsernameBlacklist is a list of usernames to be blacklisted for auto setup.
242 AutoSetupNewUsers: false
243 AutoSetupNewUsersWithVmUUID: ""
244 AutoSetupNewUsersWithRepository: false
245 AutoSetupUsernameBlacklist:
254 # When NewUsersAreActive is set to true, new users will be active
255 # immediately. This skips the "self-activate" step which enforces
256 # user agreements. Should only be enabled for development.
257 NewUsersAreActive: false
259 # The e-mail address of the user you would like to become marked as an admin
260 # user on their first login.
261 # In the default configuration, authentication happens through the Arvados SSO
262 # server, which uses OAuth2 against Google's servers, so in that case this
263 # should be an address associated with a Google account.
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. You can get
280 # the token by running "bundle exec ./script/get_anonymous_user_token.rb"
281 # in the directory where your API server is running.
282 AnonymousUserToken: ""
284 # If a new user has an alternate email address (local@domain)
285 # with the domain given here, its local part becomes the new
286 # user's default username. Otherwise, the user's primary email
288 PreferDomainForUsername: ""
291 <% if not @user.full_name.empty? -%>
292 <%= @user.full_name %>,
297 Your Arvados account has been set up. You can log in at
299 <%= Rails.configuration.Services.Workbench1.ExternalURL %>
302 Your Arvados administrator.
305 # Time to keep audit logs, in seconds. (An audit log is a row added
306 # to the "logs" table in the PostgreSQL database each time an
307 # Arvados object is created, modified, or deleted.)
309 # Currently, websocket event notifications rely on audit logs, so
310 # this should not be set lower than 300 (5 minutes).
313 # Maximum number of log rows to delete in a single SQL transaction.
315 # If MaxDeleteBatch is 0, log entries will never be
316 # deleted by Arvados. Cleanup can be done by an external process
317 # without affecting any Arvados system processes, as long as very
318 # recent (<5 minutes old) logs are not deleted.
320 # 100000 is a reasonable batch size for most sites.
323 # Attributes to suppress in events and audit logs. Notably,
324 # specifying {"manifest_text": {}} here typically makes the database
325 # smaller and faster.
327 # Warning: Using any non-empty value here can have undesirable side
328 # effects for any client or component that relies on event logs.
329 # Use at your own risk.
330 UnloggedAttributes: {}
334 # Logging threshold: panic, fatal, error, warn, info, debug, or
338 # Logging format: json or text
341 # Maximum characters of (JSON-encoded) query parameters to include
342 # in each request log entry. When params exceed this size, they will
343 # be JSON-encoded, truncated to this size, and logged as
345 MaxRequestLogParamsSize: 2000
349 # Enable access controls for data stored in Keep. This should
350 # always be set to true on a production cluster.
353 # BlobSigningKey is a string of alphanumeric characters used to
354 # generate permission signatures for Keep locators. It must be
355 # identical to the permission key given to Keep. IMPORTANT: This
356 # is a site secret. It should be at least 50 characters.
358 # Modifying BlobSigningKey will invalidate all existing
359 # signatures, which can cause programs to fail (e.g., arv-put,
360 # arv-get, and Crunch jobs). To avoid errors, rotate keys only
361 # when no such processes are running.
364 # Enable garbage collection of unreferenced blobs in Keep.
367 # Time to leave unreferenced blobs in "trashed" state before
368 # deleting them, or 0 to skip the "trashed" state entirely and
369 # delete unreferenced blobs.
371 # If you use any Amazon S3 buckets as storage volumes, this
372 # must be at least 24h to avoid occasional data loss.
373 BlobTrashLifetime: 336h
375 # How often to check for (and delete) trashed blocks whose
376 # BlobTrashLifetime has expired.
377 BlobTrashCheckInterval: 24h
379 # Maximum number of concurrent "trash blob" and "delete trashed
380 # blob" operations conducted by a single keepstore process. Each
381 # of these can be set to 0 to disable the respective operation.
383 # If BlobTrashLifetime is zero, "trash" and "delete trash"
384 # happen at once, so only the lower of these two values is used.
385 BlobTrashConcurrency: 4
386 BlobDeleteConcurrency: 4
388 # Maximum number of concurrent "create additional replica of
389 # existing blob" operations conducted by a single keepstore
391 BlobReplicateConcurrency: 4
393 # Default replication level for collections. This is used when a
394 # collection's replication_desired attribute is nil.
395 DefaultReplication: 2
397 # BlobSigningTTL determines the minimum lifetime of transient
398 # data, i.e., blocks that are not referenced by
399 # collections. Unreferenced blocks exist for two reasons:
401 # 1) A data block must be written to a disk/cloud backend device
402 # before a collection can be created/updated with a reference to
405 # 2) Deleting or updating a collection can remove the last
406 # remaining reference to a data block.
408 # If BlobSigningTTL is too short, long-running
409 # processes/containers will fail when they take too long (a)
410 # between writing blocks and writing collections that reference
411 # them, or (b) between reading collections and reading the
414 # If BlobSigningTTL is too long, data will still be stored long
415 # after the referring collections are deleted, and you will
416 # needlessly fill up disks or waste money on cloud storage.
418 # Modifying BlobSigningTTL invalidates existing signatures; see
419 # BlobSigningKey note above.
421 # The default is 2 weeks.
424 # When running keep-balance, this is the destination filename for
425 # the list of lost block hashes if there are any, one per line.
426 # Updated automically during each successful run.
427 BlobMissingReport: ""
429 # keep-balance operates periodically, i.e.: do a
430 # scan/balance operation, sleep, repeat.
432 # BalancePeriod determines the interval between start times of
433 # successive scan/balance operations. If a scan/balance operation
434 # takes longer than RunPeriod, the next one will follow it
437 # If SIGUSR1 is received during an idle period between operations,
438 # the next operation will start immediately.
441 # Limits the number of collections retrieved by keep-balance per
442 # API transaction. If this is zero, page size is
443 # determined by the API server's own page size limits (see
444 # API.MaxItemsPerResponse and API.MaxIndexDatabaseRead).
445 BalanceCollectionBatch: 0
447 # The size of keep-balance's internal queue of
448 # collections. Higher values use more memory and improve throughput
449 # by allowing keep-balance to fetch the next page of collections
450 # while the current page is still being processed. If this is zero
451 # or omitted, pages are processed serially.
452 BalanceCollectionBuffers: 1000
454 # Maximum time for a rebalancing run. This ensures keep-balance
455 # eventually gives up and retries if, for example, a network
456 # error causes a hung connection that is never closed by the
457 # OS. It should be long enough that it doesn't interrupt a
458 # long-running balancing operation.
461 # Default lifetime for ephemeral collections: 2 weeks. This must not
462 # be less than BlobSigningTTL.
463 DefaultTrashLifetime: 336h
465 # Interval (seconds) between trash sweeps. During a trash sweep,
466 # collections are marked as trash if their trash_at time has
467 # arrived, and deleted if their delete_at time has arrived.
468 TrashSweepInterval: 60s
470 # If true, enable collection versioning.
471 # When a collection's preserve_version field is true or the current version
472 # is older than the amount of seconds defined on PreserveVersionIfIdle,
473 # a snapshot of the collection's previous state is created and linked to
474 # the current collection.
475 CollectionVersioning: false
477 # 0s = auto-create a new version on every update.
478 # -1s = never auto-create new versions.
479 # > 0s = auto-create a new version when older than the specified number of seconds.
480 PreserveVersionIfIdle: -1s
482 # If non-empty, allow project and collection names to contain
483 # the "/" character (slash/stroke/solidus), and replace "/" with
484 # the given string in the filesystem hierarchy presented by
485 # WebDAV. Example values are "%2f" and "{slash}". Names that
486 # contain the substitution string itself may result in confusing
487 # behavior, so a value like "_" is not recommended.
489 # If the default empty value is used, the server will reject
490 # requests to create or rename a collection when the new name
493 # If the value "/" is used, project and collection names
494 # containing "/" will be allowed, but they will not be
495 # accessible via WebDAV.
497 # Use of this feature is not recommended, if it can be avoided.
498 ForwardSlashNameSubstitution: ""
500 # Include "folder objects" in S3 ListObjects responses.
501 S3FolderObjects: true
503 # Managed collection properties. At creation time, if the client didn't
504 # provide the listed keys, they will be automatically populated following
505 # one of the following behaviors:
507 # * UUID of the user who owns the containing project.
508 # responsible_person_uuid: {Function: original_owner, Protected: true}
510 # * Default concrete value.
511 # foo_bar: {Value: baz, Protected: false}
513 # If Protected is true, only an admin user can modify its value.
515 SAMPLE: {Function: original_owner, Protected: true}
517 # In "trust all content" mode, Workbench will redirect download
518 # requests to WebDAV preview link, even in the cases when
519 # WebDAV would have to expose XSS vulnerabilities in order to
520 # handle the redirect (see discussion on Services.WebDAV).
522 # This setting has no effect in the recommended configuration,
523 # where the WebDAV is configured to have a separate domain for
524 # every collection; in this case XSS protection is provided by
525 # browsers' same-origin policy.
527 # The default setting (false) is appropriate for a multi-user site.
528 TrustAllContent: false
530 # Cache parameters for WebDAV content serving:
531 # * TTL: Maximum time to cache manifests and permission checks.
532 # * UUIDTTL: Maximum time to cache collection state.
533 # * MaxBlockEntries: Maximum number of block cache entries.
534 # * MaxCollectionEntries: Maximum number of collection cache entries.
535 # * MaxCollectionBytes: Approximate memory limit for collection cache.
536 # * MaxPermissionEntries: Maximum number of permission cache entries.
537 # * MaxUUIDEntries: Maximum number of UUID cache entries.
542 MaxCollectionEntries: 1000
543 MaxCollectionBytes: 100000000
544 MaxPermissionEntries: 1000
548 # One of the following mechanisms (SSO, Google, PAM, LDAP, or
549 # LoginCluster) should be enabled; see
550 # https://doc.arvados.org/install/setup-login.html
553 # Authenticate with Google.
556 # Use the Google Cloud console to enable the People API (APIs
557 # and Services > Enable APIs and services > Google People API
558 # > Enable), generate a Client ID and secret (APIs and
559 # Services > Credentials > Create credentials > OAuth client
560 # ID > Web application) and add your controller's /login URL
561 # (e.g., "https://zzzzz.example.com/login") as an authorized
564 # Incompatible with ForceLegacyAPI14. ProviderAppID must be
569 # Allow users to log in to existing accounts using any verified
570 # email address listed by their Google account. If true, the
571 # Google People API must be enabled in order for Google login to
572 # work. If false, only the primary email address will be used.
573 AlternateEmailAddresses: true
576 # Authenticate with an OpenID Connect provider.
579 # Issuer URL, e.g., "https://login.example.com".
581 # This must be exactly equal to the URL returned by the issuer
582 # itself in its config response ("isser" key). If the
583 # configured value is "https://example" and the provider
584 # returns "https://example:443" or "https://example/" then
585 # login will fail, even though those URLs are equivalent
589 # Your client ID and client secret (supplied by the provider).
593 # OpenID claim field containing the user's email
594 # address. Normally "email"; see
595 # https://openid.net/specs/openid-connect-core-1_0.html#StandardClaims
598 # OpenID claim field containing the email verification
599 # flag. Normally "email_verified". To accept every returned
600 # email address without checking a "verified" field at all,
601 # use the empty string "".
602 EmailVerifiedClaim: "email_verified"
604 # OpenID claim field containing the user's preferred
605 # username. If empty, use the mailbox part of the user's email
610 # (Experimental) Use PAM to authenticate users.
613 # PAM service name. PAM will apply the policy in the
614 # corresponding config file (e.g., /etc/pam.d/arvados) or, if
615 # there is none, the default "other" config.
618 # Domain name (e.g., "example.com") to use to construct the
619 # user's email address if PAM authentication returns a
620 # username with no "@". If empty, use the PAM username as the
621 # user's email address, whether or not it contains "@".
623 # Note that the email address is used as the primary key for
624 # user records when logging in. Therefore, if you change
625 # PAMDefaultEmailDomain after the initial installation, you
626 # should also update existing user records to reflect the new
627 # domain. Otherwise, next time those users log in, they will
628 # be given new accounts instead of accessing their existing
630 DefaultEmailDomain: ""
633 # Use an LDAP service to authenticate users.
636 # Server URL, like "ldap://ldapserver.example.com:389" or
637 # "ldaps://ldapserver.example.com:636".
638 URL: "ldap://ldap:389"
640 # Use StartTLS upon connecting to the server.
643 # Skip TLS certificate name verification.
646 # Strip the @domain part if a user supplies an email-style
647 # username with this domain. If "*", strip any user-provided
648 # domain. If "", never strip the domain part. Example:
652 # If, after applying StripDomain, the username contains no "@"
653 # character, append this domain to form an email-style
654 # username. Example: "example.com"
657 # The LDAP attribute to filter on when looking up a username
658 # (after applying StripDomain and AppendDomain).
661 # Bind with this username (DN or UPN) and password when
662 # looking up the user record.
664 # Example user: "cn=admin,dc=example,dc=com"
666 SearchBindPassword: ""
668 # Directory base for username lookup. Example:
669 # "ou=Users,dc=example,dc=com"
672 # Additional filters to apply when looking up users' LDAP
673 # entries. This can be used to restrict access to a subset of
674 # LDAP users, or to disambiguate users from other directory
675 # entries that have the SearchAttribute present.
677 # Special characters in assertion values must be escaped (see
680 # Example: "(objectClass=person)"
683 # LDAP attribute to use as the user's email address.
685 # Important: This must not be an attribute whose value can be
686 # edited in the directory by the users themselves. Otherwise,
687 # users can take over other users' Arvados accounts trivially
688 # (email address is the primary key for Arvados accounts.)
691 # LDAP attribute to use as the preferred Arvados username. If
692 # no value is found (or this config is empty) the username
693 # originally supplied by the user will be used.
694 UsernameAttribute: uid
697 # Authenticate with a separate SSO server. (Deprecated)
700 # ProviderAppID and ProviderAppSecret are generated during SSO
702 # https://doc.arvados.org/v2.0/install/install-sso.html#update-config
704 ProviderAppSecret: ""
707 # Authenticate users listed here in the config file. This
708 # feature is intended to be used in test environments, and
709 # should not be used in production.
713 Email: alice@example.com
716 # The cluster ID to delegate the user database. When set,
717 # logins on this cluster will be redirected to the login cluster
718 # (login cluster must appear in RemoteClusters with Proxy: true)
721 # How long a cached token belonging to a remote cluster will
722 # remain valid before it needs to be revalidated.
723 RemoteTokenRefresh: 5m
725 # How long a client token created from a login flow will be valid without
726 # asking the user to re-login. Example values: 60m, 8h.
727 # Default value zero means tokens don't have expiration.
730 # When the token is returned to a client, the token itself may
731 # be restricted from manipulating other tokens based on whether
732 # the client is "trusted" or not. The local Workbench1 and
733 # Workbench2 are trusted by default, but if this is a
734 # LoginCluster, you probably want to include the other Workbench
735 # instances in the federation in this list.
738 "https://workbench.federate1.example": {}
739 "https://workbench.federate2.example": {}
742 # Path to git or gitolite-shell executable. Each authenticated
743 # request will execute this program with the single argument "http-backend"
744 GitCommand: /usr/bin/git
746 # Path to Gitolite's home directory. If a non-empty path is given,
747 # the CGI environment will be set up to support the use of
748 # gitolite-shell as a GitCommand: for example, if GitoliteHome is
749 # "/gh", then the CGI environment will have GITOLITE_HTTP_HOME=/gh,
750 # PATH=$PATH:/gh/bin, and GL_BYPASS_ACCESS_CHECKS=1.
753 # Git repositories must be readable by api server, or you won't be
754 # able to submit crunch jobs. To pass the test suites, put a clone
755 # of the arvados tree in {git_repositories_dir}/arvados.git or
756 # {git_repositories_dir}/arvados/.git
757 Repositories: /var/lib/arvados/git/repositories
765 # List of supported Docker Registry image formats that compute nodes
766 # are able to use. `arv keep docker` will error out if a user tries
767 # to store an image with an unsupported format. Use an empty array
768 # to skip the compatibility check (and display a warning message to
771 # Example for sites running docker < 1.10: {"v1": {}}
772 # Example for sites running docker >= 1.10: {"v2": {}}
773 # Example for disabling check: {}
774 SupportedDockerImageFormats:
778 # Include details about job reuse decisions in the server log. This
779 # causes additional database queries to run, so it should not be
780 # enabled unless you expect to examine the resulting logs for
781 # troubleshooting purposes.
782 LogReuseDecisions: false
784 # Default value for keep_cache_ram of a container's runtime_constraints.
785 DefaultKeepCacheRAM: 268435456
787 # Number of times a container can be unlocked before being
788 # automatically cancelled.
789 MaxDispatchAttempts: 5
791 # Default value for container_count_max for container requests. This is the
792 # number of times Arvados will create a new container to satisfy a container
793 # request. If a container is cancelled it will retry a new container if
794 # container_count < container_count_max on any container requests associated
795 # with the cancelled container.
798 # The maximum number of compute nodes that can be in use simultaneously
799 # If this limit is reduced, any existing nodes with slot number >= new limit
800 # will not be counted against the new limit. In other words, the new limit
801 # won't be strictly enforced until those nodes with higher slot numbers
805 # Preemptible instance support (e.g. AWS Spot Instances)
806 # When true, child containers will get created with the preemptible
807 # scheduling parameter parameter set.
808 UsePreemptibleInstances: false
810 # PEM encoded SSH key (RSA, DSA, or ECDSA) used by the
811 # (experimental) cloud dispatcher for executing containers on
812 # worker VMs. Begins with "-----BEGIN RSA PRIVATE KEY-----\n"
813 # and ends with "\n-----END RSA PRIVATE KEY-----\n".
814 DispatchPrivateKey: ""
816 # Maximum time to wait for workers to come up before abandoning
817 # stale locks from a previous dispatch process.
820 # The crunch-run command to manage the container on a node
821 CrunchRunCommand: "crunch-run"
823 # Extra arguments to add to crunch-run invocation
824 # Example: ["--cgroup-parent-subsystem=memory"]
825 CrunchRunArgumentsList: []
827 # Extra RAM to reserve on the node, in addition to
828 # the amount specified in the container's RuntimeConstraints
829 ReserveExtraRAM: 256MiB
831 # Minimum time between two attempts to run the same container
835 # When you run the db:delete_old_container_logs task, it will find
836 # containers that have been finished for at least this many seconds,
837 # and delete their stdout, stderr, arv-mount, crunch-run, and
838 # crunchstat logs from the logs table.
841 # These two settings control how frequently log events are flushed to the
842 # database. Log lines are buffered until either crunch_log_bytes_per_event
843 # has been reached or crunch_log_seconds_between_events has elapsed since
845 LogBytesPerEvent: 4096
846 LogSecondsBetweenEvents: 5s
848 # The sample period for throttling logs.
849 LogThrottlePeriod: 60s
851 # Maximum number of bytes that job can log over crunch_log_throttle_period
852 # before being silenced until the end of the period.
853 LogThrottleBytes: 65536
855 # Maximum number of lines that job can log over crunch_log_throttle_period
856 # before being silenced until the end of the period.
857 LogThrottleLines: 1024
859 # Maximum bytes that may be logged by a single job. Log bytes that are
860 # silenced by throttling are not counted against this total.
861 LimitLogBytesPerJob: 67108864
863 LogPartialLineThrottlePeriod: 5s
865 # Container logs are written to Keep and saved in a
866 # collection, which is updated periodically while the
867 # container runs. This value sets the interval between
868 # collection updates.
871 # The log collection is also updated when the specified amount of
872 # log data (given in bytes) is produced in less than one update
878 SbatchArgumentsList: []
879 SbatchEnvironmentVariables:
882 # Path to dns server configuration directory
883 # (e.g. /etc/unbound.d/conf.d). If false, do not write any config
884 # files or touch restart.txt (see below).
887 # Template file for the dns server host snippets. See
888 # unbound.template in this directory for an example. If false, do
889 # not write any config files.
890 DNSServerConfTemplate: ""
892 # String to write to {dns_server_conf_dir}/restart.txt (with a
893 # trailing newline) after updating local data. If false, do not
894 # open or write the restart.txt file.
895 DNSServerReloadCommand: ""
897 # Command to run after each DNS update. Template variables will be
898 # substituted; see the "unbound" example below. If false, do not run
900 DNSServerUpdateCommand: ""
902 ComputeNodeDomain: ""
903 ComputeNodeNameservers:
907 # Hostname to assign to a compute node when it sends a "ping" and the
908 # hostname in its Node record is nil.
909 # During bootstrapping, the "ping" script is expected to notice the
910 # hostname given in the ping response, and update its unix hostname
912 # If false, leave the hostname alone (this is appropriate if your compute
913 # nodes' hostnames are already assigned by some other mechanism).
915 # One way or another, the hostnames of your node records should agree
916 # with your DNS records and your /etc/slurm-llnl/slurm.conf files.
918 # Example for compute0000, compute0001, ....:
919 # assign_node_hostname: compute%<slot_number>04d
920 # (See http://ruby-doc.org/core-2.2.2/Kernel.html#method-i-format for more.)
921 AssignNodeHostname: "compute%<slot_number>d"
924 # Enable the legacy 'jobs' API (crunch v1). This value must be a string.
926 # Note: this only enables read-only access, creating new
927 # legacy jobs and pipelines is not supported.
929 # 'auto' -- (default) enable the Jobs API only if it has been used before
930 # (i.e., there are job records in the database)
931 # 'true' -- enable the Jobs API despite lack of existing records.
932 # 'false' -- disable the Jobs API despite presence of existing records.
935 # Git repositories must be readable by api server, or you won't be
936 # able to submit crunch jobs. To pass the test suites, put a clone
937 # of the arvados tree in {git_repositories_dir}/arvados.git or
938 # {git_repositories_dir}/arvados/.git
939 GitInternalDir: /var/lib/arvados/internal.git
942 # Enable the cloud scheduler (experimental).
945 # Name/number of port where workers' SSH services listen.
948 # Interval between queue polls.
951 # Shell command to execute on each worker to determine whether
952 # the worker is booted and ready to run containers. It should
953 # exit zero if the worker is ready.
954 BootProbeCommand: "docker ps -q"
956 # Minimum interval between consecutive probes to a single
960 # Maximum probes per second, across all workers in a pool.
961 MaxProbesPerSecond: 10
963 # Time before repeating SIGTERM when killing a container.
966 # Time to give up on a process (most likely arv-mount) that
967 # still holds a container lockfile after its main supervisor
968 # process has exited, and declare the instance broken.
969 TimeoutStaleRunLock: 5s
971 # Time to give up on SIGTERM and write off the worker.
974 # Maximum create/destroy-instance operations per second (0 =
976 MaxCloudOpsPerSecond: 0
978 # Maximum concurrent node creation operations (0 = unlimited). This is
979 # recommended by Azure in certain scenarios (see
980 # https://docs.microsoft.com/en-us/azure/virtual-machines/linux/capture-image)
981 # and can be used with other cloud providers too, if desired.
982 MaxConcurrentInstanceCreateOps: 0
984 # Interval between cloud provider syncs/updates ("list all
988 # Time to leave an idle worker running (in case new containers
989 # appear in the queue that it can run) before shutting it
993 # Time to wait for a new worker to boot (i.e., pass
994 # BootProbeCommand) before giving up and shutting it down.
997 # Maximum time a worker can stay alive with no successful
998 # probes before being automatically shut down.
1001 # Time after shutting down a worker to retry the
1002 # shutdown/destroy operation.
1003 TimeoutShutdown: 10s
1005 # Worker VM image ID.
1006 # (aws) AMI identifier
1007 # (azure) managed disks: the name of the managed disk image
1008 # (azure) shared image gallery: the name of the image definition. Also
1009 # see the SharedImageGalleryName and SharedImageGalleryImageVersion fields.
1010 # (azure) unmanaged disks (deprecated): the complete URI of the VHD, e.g.
1011 # https://xxxxx.blob.core.windows.net/system/Microsoft.Compute/Images/images/xxxxx.vhd
1014 # An executable file (located on the dispatcher host) to be
1015 # copied to cloud instances at runtime and used as the
1016 # container runner/supervisor. The default value is the
1017 # dispatcher program itself.
1019 # Use the empty string to disable this step: nothing will be
1020 # copied, and cloud instances are assumed to have a suitable
1021 # version of crunch-run installed.
1022 DeployRunnerBinary: "/proc/self/exe"
1024 # Tags to add on all resources (VMs, NICs, disks) created by
1025 # the container dispatcher. (Arvados's own tags --
1026 # InstanceType, IdleBehavior, and InstanceSecret -- will also
1031 # Prefix for predefined tags used by Arvados (InstanceSetID,
1032 # InstanceType, InstanceSecret, IdleBehavior). With the
1033 # default value "Arvados", tags are "ArvadosInstanceSetID",
1034 # "ArvadosInstanceSecret", etc.
1036 # This should only be changed while no cloud resources are in
1037 # use and the cloud dispatcher is not running. Otherwise,
1038 # VMs/resources that were added using the old tag prefix will
1039 # need to be detected and cleaned up manually.
1040 TagKeyPrefix: Arvados
1042 # Cloud driver: "azure" (Microsoft Azure) or "ec2" (Amazon AWS).
1045 # Cloud-specific driver parameters.
1048 # (ec2) Credentials.
1052 # (ec2) Instance configuration.
1058 AdminUsername: debian
1060 # (azure) Credentials.
1066 # (azure) Instance configuration.
1067 CloudEnvironment: AzurePublicCloud
1070 # (azure) The resource group where the VM and virtual NIC will be
1074 # (azure) The resource group of the Network to use for the virtual
1075 # NIC (if different from ResourceGroup)
1076 NetworkResourceGroup: ""
1080 # (azure) managed disks: The resource group where the managed disk
1081 # image can be found (if different from ResourceGroup).
1082 ImageResourceGroup: ""
1084 # (azure) shared image gallery: the name of the gallery
1085 SharedImageGalleryName: ""
1086 # (azure) shared image gallery: the version of the image definition
1087 SharedImageGalleryImageVersion: ""
1089 # (azure) unmanaged disks (deprecated): Where to store the VM VHD blobs
1093 # (azure) How long to wait before deleting VHD and NIC
1094 # objects that are no longer being used.
1095 DeleteDanglingResourcesAfter: 20s
1097 # Account (that already exists in the VM image) that will be
1098 # set up with an ssh authorized key to allow the compute
1099 # dispatcher to connect.
1100 AdminUsername: arvados
1104 # Use the instance type name as the key (in place of "SAMPLE" in
1105 # this sample entry).
1107 # Cloud provider's instance type. Defaults to the configured type name.
1111 IncludedScratch: 16GB
1118 # AccessViaHosts specifies which keepstore processes can read
1119 # and write data on the volume.
1121 # For a local filesystem, AccessViaHosts has one entry,
1122 # indicating which server the filesystem is located on.
1124 # For a network-attached backend accessible by all keepstore
1125 # servers, like a cloud storage bucket or an NFS mount,
1126 # AccessViaHosts can be empty/omitted.
1128 # Further info/examples:
1129 # https://doc.arvados.org/install/configure-fs-storage.html
1130 # https://doc.arvados.org/install/configure-s3-object-storage.html
1131 # https://doc.arvados.org/install/configure-azure-blob-storage.html
1135 "http://host1.example:25107": {}
1143 # for s3 driver -- see
1144 # https://doc.arvados.org/install/configure-s3-object-storage.html
1151 LocationConstraint: false
1157 # Use aws-s3-go (v2) instead of goamz
1158 UseAWSS3v2Driver: false
1160 # For S3 driver, potentially unsafe tuning parameter,
1161 # intentionally excluded from main documentation.
1163 # Enable deletion (garbage collection) even when the
1164 # configured BlobTrashLifetime is zero. WARNING: eventual
1165 # consistency may result in race conditions that can cause
1166 # data loss. Do not enable this unless you understand and
1170 # for azure driver -- see
1171 # https://doc.arvados.org/install/configure-azure-blob-storage.html
1172 StorageAccountName: aaaaa
1173 StorageAccountKey: aaaaa
1174 StorageBaseURL: core.windows.net
1175 ContainerName: aaaaa
1177 ListBlobsRetryDelay: 10s
1178 ListBlobsMaxAttempts: 10
1180 WriteRaceInterval: 15s
1181 WriteRacePollTime: 1s
1183 # for local directory driver -- see
1184 # https://doc.arvados.org/install/configure-fs-storage.html
1185 Root: /var/lib/arvados/keep-data
1187 # For local directory driver, potentially confusing tuning
1188 # parameter, intentionally excluded from main documentation.
1190 # When true, read and write operations (for whole 64MiB
1191 # blocks) on an individual volume will queued and issued
1192 # serially. When false, read and write operations will be
1193 # issued concurrently.
1195 # May possibly improve throughput if you have physical spinning disks
1196 # and experience contention when there are multiple requests
1197 # to the same volume.
1199 # Otherwise, when using SSDs, RAID, or a shared network filesystem, you
1200 # should leave this alone.
1206 SendUserSetupNotificationEmail: true
1208 # Bug/issue report notification to and from addresses
1209 IssueReporterEmailFrom: "arvados@example.com"
1210 IssueReporterEmailTo: "arvados@example.com"
1211 SupportEmailAddress: "arvados@example.com"
1213 # Generic issue email from
1214 EmailFrom: "arvados@example.com"
1221 ActivateUsers: false
1223 # API endpoint host or host:port; default is {id}.arvadosapi.com
1224 Host: sample.arvadosapi.com
1226 # Perform a proxy request when a local client requests an
1227 # object belonging to this remote.
1230 # Default "https". Can be set to "http" for testing.
1233 # Disable TLS verify. Can be set to true for testing.
1236 # When users present tokens issued by this remote cluster, and
1237 # their accounts are active on the remote cluster, activate
1238 # them on this cluster too.
1239 ActivateUsers: false
1242 # Workbench1 configs
1244 ActivationContactLink: mailto:info@arvados.org
1245 ArvadosDocsite: https://doc.arvados.org
1246 ArvadosPublicDataDocURL: https://playground.arvados.org/projects/public
1247 ShowUserAgreementInline: false
1250 # Scratch directory used by the remote repository browsing
1251 # feature. If it doesn't exist, it (and any missing parents) will be
1252 # created using mkdir_p.
1253 RepositoryCache: /var/www/arvados-workbench/current/tmp/git
1255 # Below is a sample setting of user_profile_form_fields config parameter.
1256 # This configuration parameter should be set to either false (to disable) or
1257 # to a map as shown below.
1258 # Configure the map of input fields to be displayed in the profile page
1259 # using the attribute "key" for each of the input fields.
1260 # This sample shows configuration with one required and one optional form fields.
1261 # For each of these input fields:
1262 # You can specify "Type" as "text" or "select".
1263 # List the "Options" to be displayed for each of the "select" menu.
1264 # Set "Required" as "true" for any of these fields to make them required.
1265 # If any of the required fields are missing in the user's profile, the user will be
1266 # redirected to the profile page before they can access any Workbench features.
1267 UserProfileFormFields:
1270 FormFieldTitle: Best color
1271 FormFieldDescription: your favorite color
1280 # exampleTextValue: # key that will be set in properties
1282 # FormFieldTitle: ""
1283 # FormFieldDescription: ""
1286 # exampleOptionsValue:
1288 # FormFieldTitle: ""
1289 # FormFieldDescription: ""
1297 # Use "UserProfileFormMessage to configure the message you want
1298 # to display on the profile page.
1299 UserProfileFormMessage: 'Welcome to Arvados. All <span style="color:red">required fields</span> must be completed before you can proceed.'
1301 # Mimetypes of applications for which the view icon
1302 # would be enabled in a collection's show page.
1303 # It is sufficient to list only applications here.
1304 # No need to list text and image types.
1305 ApplicationMimetypesWithViewIcon:
1323 # The maximum number of bytes to load in the log viewer
1324 LogViewerMaxBytes: 1M
1326 # When anonymous_user_token is configured, show public projects page
1327 EnablePublicProjectsPage: true
1329 # By default, disable the "Getting Started" popup which is specific to Arvados playground
1330 EnableGettingStartedPopup: false
1332 # Ask Arvados API server to compress its response payloads.
1333 APIResponseCompression: true
1335 # Timeouts for API requests.
1336 APIClientConnectTimeout: 2m
1337 APIClientReceiveTimeout: 5m
1339 # Maximum number of historic log records of a running job to fetch
1340 # and display in the Log tab, while subscribing to web sockets.
1341 RunningJobLogRecordsToFetch: 2000
1343 # In systems with many shared projects, loading of dashboard and topnav
1344 # can be slow due to collections indexing; use the following parameters
1345 # to suppress these properties
1346 ShowRecentCollectionsOnDashboard: true
1347 ShowUserNotifications: true
1349 # Enable/disable "multi-site search" in top nav ("true"/"false"), or
1350 # a link to the multi-site search page on a "home" Workbench site.
1353 # https://workbench.zzzzz.arvadosapi.com/collections/multisite
1356 # Should workbench allow management of local git repositories? Set to false if
1357 # the jobs api is disabled and there are no local git repositories.
1360 SiteName: Arvados Workbench
1361 ProfilingEnabled: false
1363 # This is related to obsolete Google OpenID 1.0 login
1364 # but some workbench stuff still expects it to be set.
1365 DefaultOpenIdPrefix: "https://www.google.com/accounts/o8/id"
1367 # Workbench2 configs
1369 FileViewersConfigURL: ""
1371 # Idle time after which the user's session will be auto closed.
1372 # This feature is disabled when set to zero.
1375 # Workbench welcome screen, this is HTML text that will be
1376 # incorporated directly onto the page.
1378 <img src="/arvados-logo-big.png" style="width: 20%; float: right; padding: 1em;" />
1379 <h2>Please log in.</h2>
1381 <p>The "Log in" button below will show you a sign-in
1382 page. After you log in, you will be redirected back to
1383 Arvados Workbench.</p>
1385 <p>If you have never used Arvados Workbench before, logging in
1386 for the first time will automatically create a new
1389 <i>Arvados Workbench uses your name and email address only for
1390 identification, and does not retrieve any other personal
1393 # Workbench screen displayed to inactive users. This is HTML
1394 # text that will be incorporated directly onto the page.
1396 <img src="/arvados-logo-big.png" style="width: 20%; float: right; padding: 1em;" />
1397 <h3>Hi! You're logged in, but...</h3>
1398 <p>Your account is inactive.</p>
1399 <p>An administrator must activate your account before you can get
1402 # Connecting to Arvados shell VMs tends to be site-specific.
1403 # Put any special instructions here. This is HTML text that will
1404 # be incorporated directly onto the Workbench page.
1406 <a href="https://doc.arvados.org/user/getting_started/ssh-access-unix.html">Accessing an Arvados VM with SSH</a> (generic instructions).
1407 Site configurations vary. Contact your local cluster administrator if you have difficulty accessing an Arvados shell node.
1409 # Sample text if you are using a "switchyard" ssh proxy.
1410 # Replace "zzzzz" with your Cluster ID.
1412 # <p>Add a section like this to your SSH configuration file ( <i>~/.ssh/config</i>):</p>
1415 # ServerAliveInterval 60
1416 # ProxyCommand ssh -p2222 turnout@switchyard.zzzzz.arvadosapi.com -x -a $SSH_PROXY_FLAGS %h
1419 # If you are using a switchyard ssh proxy, shell node hostnames
1420 # may require a special hostname suffix. In the sample ssh
1421 # configuration above, this would be ".zzzzz"
1422 # This is added to the hostname in the "command line" column
1423 # the Workbench "shell VMs" page.
1425 # If your shell nodes are directly accessible by users without a
1426 # proxy and have fully qualified host names, you should leave
1428 SSHHelpHostSuffix: ""
1430 # Bypass new (Arvados 1.5) API implementations, and hand off
1431 # requests directly to Rails instead. This can provide a temporary
1432 # workaround for clients that are incompatible with the new API
1433 # implementation. Note that it also disables some new federation
1434 # features and will be removed in a future release.
1435 ForceLegacyAPI14: false
1437 # (Experimental) Restart services automatically when config file
1438 # changes are detected. Only supported by `arvados-server boot` in
1440 AutoReloadConfig: false