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"
73 # max concurrent connections per arvados server daemon
76 # All parameters here are passed to the PG client library in a connection string;
77 # see https://www.postgresql.org/docs/current/static/libpq-connect.html#LIBPQ-PARAMKEYWORDS
84 # Maximum size (in bytes) allowed for a single API request. This
85 # limit is published in the discovery document for use by clients.
86 # Note: You must separately configure the upstream web server or
87 # proxy to actually enforce the desired maximum request size on the
89 MaxRequestSize: 134217728
91 # Limit the number of bytes read from the database during an index
92 # request (by retrieving and returning fewer rows than would
93 # normally be returned in a single response).
94 # Note 1: This setting never reduces the number of returned rows to
95 # zero, no matter how big the first data row is.
96 # Note 2: Currently, this is only checked against a specific set of
97 # columns that tend to get large (collections.manifest_text,
98 # containers.mounts, workflows.definition). Other fields (e.g.,
99 # "properties" hashes) are not counted against this limit.
100 MaxIndexDatabaseRead: 134217728
102 # Maximum number of items to return when responding to a APIs that
103 # can return partial result sets using limit and offset parameters
104 # (e.g., *.index, groups.contents). If a request specifies a "limit"
105 # parameter higher than this value, this value is used instead.
106 MaxItemsPerResponse: 1000
108 # API methods to disable. Disabled methods are not listed in the
109 # discovery document, and respond 404 to all requests.
110 # Example: ["jobs.create", "pipeline_instances.create"]
113 # Interval (seconds) between asynchronous permission view updates. Any
114 # permission-updating API called with the 'async' parameter schedules a an
115 # update on the permission view in the future, if not already scheduled.
116 AsyncPermissionsUpdateInterval: 20
118 # Maximum number of concurrent outgoing requests to make while
119 # serving a single incoming multi-cluster (federated) request.
120 MaxRequestAmplification: 4
122 # RailsSessionSecretToken is a string of alphanumeric characters
123 # used by Rails to sign session tokens. IMPORTANT: This is a
124 # site secret. It should be at least 50 characters.
125 RailsSessionSecretToken: ""
127 # Maximum wall clock time to spend handling an incoming request.
131 # Config parameters to automatically setup new users. If enabled,
132 # this users will be able to self-activate. Enable this if you want
133 # to run an open instance where anyone can create an account and use
134 # the system without requiring manual approval.
136 # The params auto_setup_new_users_with_* are meaningful only when auto_setup_new_users is turned on.
137 # auto_setup_name_blacklist is a list of usernames to be blacklisted for auto setup.
138 AutoSetupNewUsers: false
139 AutoSetupNewUsersWithVmUUID: ""
140 AutoSetupNewUsersWithRepository: false
141 AutoSetupUsernameBlacklist: [arvados, git, gitolite, gitolite-admin, root, syslog]
143 # When new_users_are_active is set to true, new users will be active
144 # immediately. This skips the "self-activate" step which enforces
145 # user agreements. Should only be enabled for development.
146 NewUsersAreActive: false
148 # The e-mail address of the user you would like to become marked as an admin
149 # user on their first login.
150 # In the default configuration, authentication happens through the Arvados SSO
151 # server, which uses OAuth2 against Google's servers, so in that case this
152 # should be an address associated with a Google account.
153 AutoAdminUserWithEmail: ""
155 # If auto_admin_first_user is set to true, the first user to log in when no
156 # other admin users exist will automatically become an admin user.
157 AutoAdminFirstUser: false
159 # Email address to notify whenever a user creates a profile for the
161 UserProfileNotificationAddress: ""
162 AdminNotifierEmailFrom: arvados@example.com
163 EmailSubjectPrefix: "[ARVADOS] "
164 UserNotifierEmailFrom: arvados@example.com
165 NewUserNotificationRecipients: []
166 NewInactiveUserNotificationRecipients: []
169 # Time to keep audit logs, in seconds. (An audit log is a row added
170 # to the "logs" table in the PostgreSQL database each time an
171 # Arvados object is created, modified, or deleted.)
173 # Currently, websocket event notifications rely on audit logs, so
174 # this should not be set lower than 300 (5 minutes).
177 # Maximum number of log rows to delete in a single SQL transaction.
179 # If max_audit_log_delete_batch is 0, log entries will never be
180 # deleted by Arvados. Cleanup can be done by an external process
181 # without affecting any Arvados system processes, as long as very
182 # recent (<5 minutes old) logs are not deleted.
184 # 100000 is a reasonable batch size for most sites.
187 # Attributes to suppress in events and audit logs. Notably,
188 # specifying ["manifest_text"] here typically makes the database
189 # smaller and faster.
191 # Warning: Using any non-empty value here can have undesirable side
192 # effects for any client or component that relies on event logs.
193 # Use at your own risk.
194 UnloggedAttributes: []
198 # Logging threshold: panic, fatal, error, warn, info, debug, or
202 # Logging format: json or text
205 # Maximum characters of (JSON-encoded) query parameters to include
206 # in each request log entry. When params exceed this size, they will
207 # be JSON-encoded, truncated to this size, and logged as
209 MaxRequestLogParamsSize: 2000
212 # Allow clients to create collections by providing a manifest with
213 # unsigned data blob locators. IMPORTANT: This effectively disables
214 # access controls for data stored in Keep: a client who knows a hash
215 # can write a manifest that references the hash, pass it to
216 # collections.create (which will create a permission link), use
217 # collections.get to obtain a signature for that data locator, and
218 # use that signed locator to retrieve the data from Keep. Therefore,
219 # do not turn this on if your users expect to keep data private from
223 # blob_signing_key is a string of alphanumeric characters used to
224 # generate permission signatures for Keep locators. It must be
225 # identical to the permission key given to Keep. IMPORTANT: This is
226 # a site secret. It should be at least 50 characters.
228 # Modifying blob_signing_key will invalidate all existing
229 # signatures, which can cause programs to fail (e.g., arv-put,
230 # arv-get, and Crunch jobs). To avoid errors, rotate keys only when
231 # no such processes are running.
234 # Default replication level for collections. This is used when a
235 # collection's replication_desired attribute is nil.
236 DefaultReplication: 2
238 # Lifetime (in seconds) of blob permission signatures generated by
239 # the API server. This determines how long a client can take (after
240 # retrieving a collection record) to retrieve the collection data
241 # from Keep. If the client needs more time than that (assuming the
242 # collection still has the same content and the relevant user/token
243 # still has permission) the client can retrieve the collection again
244 # to get fresh signatures.
246 # This must be exactly equal to the -blob-signature-ttl flag used by
247 # keepstore servers. Otherwise, reading data blocks and saving
248 # collections will fail with HTTP 403 permission errors.
250 # Modifying blob_signature_ttl invalidates existing signatures; see
251 # blob_signing_key note above.
253 # The default is 2 weeks.
256 # Default lifetime for ephemeral collections: 2 weeks. This must not
257 # be less than blob_signature_ttl.
258 DefaultTrashLifetime: 336h
260 # Interval (seconds) between trash sweeps. During a trash sweep,
261 # collections are marked as trash if their trash_at time has
262 # arrived, and deleted if their delete_at time has arrived.
263 TrashSweepInterval: 60
265 # If true, enable collection versioning.
266 # When a collection's preserve_version field is true or the current version
267 # is older than the amount of seconds defined on preserve_version_if_idle,
268 # a snapshot of the collection's previous state is created and linked to
269 # the current collection.
270 CollectionVersioning: false
272 # 0 = auto-create a new version on every update.
273 # -1 = never auto-create new versions.
274 # > 0 = auto-create a new version when older than the specified number of seconds.
275 PreserveVersionIfIdle: -1
277 # Managed collection properties. At creation time, if the client didn't
278 # provide the listed keys, they will be automatically populated following
279 # one of the following behaviors:
281 # * UUID of the user who owns the containing project.
282 # responsible_person_uuid: {function: original_owner, protected: true}
284 # * Default concrete value.
285 # foo_bar: {value: baz, protected: false}
287 # If protected is true, only an admin user can modify its value.
288 ManagedProperties: {}
291 # These settings are provided by your OAuth2 provider (e.g.,
293 ProviderAppSecret: ""
297 # Git repositories must be readable by api server, or you won't be
298 # able to submit crunch jobs. To pass the test suites, put a clone
299 # of the arvados tree in {git_repositories_dir}/arvados.git or
300 # {git_repositories_dir}/arvados/.git
301 Repositories: /var/lib/arvados/git/repositories
309 # List of supported Docker Registry image formats that compute nodes
310 # are able to use. `arv keep docker` will error out if a user tries
311 # to store an image with an unsupported format. Use an empty array
312 # to skip the compatibility check (and display a warning message to
315 # Example for sites running docker < 1.10: ["v1"]
316 # Example for sites running docker >= 1.10: ["v2"]
317 # Example for disabling check: []
318 SupportedDockerImageFormats: ["v2"]
320 # Include details about job reuse decisions in the server log. This
321 # causes additional database queries to run, so it should not be
322 # enabled unless you expect to examine the resulting logs for
323 # troubleshooting purposes.
324 LogReuseDecisions: false
326 # Default value for keep_cache_ram of a container's runtime_constraints.
327 DefaultKeepCacheRAM: 268435456
329 # Number of times a container can be unlocked before being
330 # automatically cancelled.
331 MaxDispatchAttempts: 5
333 # Default value for container_count_max for container requests. This is the
334 # number of times Arvados will create a new container to satisfy a container
335 # request. If a container is cancelled it will retry a new container if
336 # container_count < container_count_max on any container requests associated
337 # with the cancelled container.
340 # The maximum number of compute nodes that can be in use simultaneously
341 # If this limit is reduced, any existing nodes with slot number >= new limit
342 # will not be counted against the new limit. In other words, the new limit
343 # won't be strictly enforced until those nodes with higher slot numbers
347 # Preemptible instance support (e.g. AWS Spot Instances)
348 # When true, child containers will get created with the preemptible
349 # scheduling parameter parameter set.
350 UsePreemptibleInstances: false
352 # Include details about job reuse decisions in the server log. This
353 # causes additional database queries to run, so it should not be
354 # enabled unless you expect to examine the resulting logs for
355 # troubleshooting purposes.
356 LogReuseDecisions: false
358 # PEM encoded SSH key (RSA, DSA, or ECDSA) used by the
359 # (experimental) cloud dispatcher for executing containers on
360 # worker VMs. Begins with "-----BEGIN RSA PRIVATE KEY-----\n"
361 # and ends with "\n-----END RSA PRIVATE KEY-----\n".
362 DispatchPrivateKey: none
364 # Maximum time to wait for workers to come up before abandoning
365 # stale locks from a previous dispatch process.
369 # When you run the db:delete_old_container_logs task, it will find
370 # containers that have been finished for at least this many seconds,
371 # and delete their stdout, stderr, arv-mount, crunch-run, and
372 # crunchstat logs from the logs table.
375 # These two settings control how frequently log events are flushed to the
376 # database. Log lines are buffered until either crunch_log_bytes_per_event
377 # has been reached or crunch_log_seconds_between_events has elapsed since
379 LogBytesPerEvent: 4096
380 LogSecondsBetweenEvents: 1
382 # The sample period for throttling logs, in seconds.
383 LogThrottlePeriod: 60
385 # Maximum number of bytes that job can log over crunch_log_throttle_period
386 # before being silenced until the end of the period.
387 LogThrottleBytes: 65536
389 # Maximum number of lines that job can log over crunch_log_throttle_period
390 # before being silenced until the end of the period.
391 LogThrottleLines: 1024
393 # Maximum bytes that may be logged by a single job. Log bytes that are
394 # silenced by throttling are not counted against this total.
395 LimitLogBytesPerJob: 67108864
397 LogPartialLineThrottlePeriod: 5
399 # Container logs are written to Keep and saved in a collection,
400 # which is updated periodically while the container runs. This
401 # value sets the interval (given in seconds) between collection
403 LogUpdatePeriod: 1800
405 # The log collection is also updated when the specified amount of
406 # log data (given in bytes) is produced in less than one update
408 LogUpdateSize: 33554432
412 # Path to dns server configuration directory
413 # (e.g. /etc/unbound.d/conf.d). If false, do not write any config
414 # files or touch restart.txt (see below).
417 # Template file for the dns server host snippets. See
418 # unbound.template in this directory for an example. If false, do
419 # not write any config files.
420 DNSServerConfTemplate: ""
422 # String to write to {dns_server_conf_dir}/restart.txt (with a
423 # trailing newline) after updating local data. If false, do not
424 # open or write the restart.txt file.
425 DNSServerReloadCommand: ""
427 # Command to run after each DNS update. Template variables will be
428 # substituted; see the "unbound" example below. If false, do not run
430 DNSServerUpdateCommand: ""
432 ComputeNodeDomain: ""
433 ComputeNodeNameservers:
436 # Hostname to assign to a compute node when it sends a "ping" and the
437 # hostname in its Node record is nil.
438 # During bootstrapping, the "ping" script is expected to notice the
439 # hostname given in the ping response, and update its unix hostname
441 # If false, leave the hostname alone (this is appropriate if your compute
442 # nodes' hostnames are already assigned by some other mechanism).
444 # One way or another, the hostnames of your node records should agree
445 # with your DNS records and your /etc/slurm-llnl/slurm.conf files.
447 # Example for compute0000, compute0001, ....:
448 # assign_node_hostname: compute%<slot_number>04d
449 # (See http://ruby-doc.org/core-2.2.2/Kernel.html#method-i-format for more.)
450 AssignNodeHostname: "compute%<slot_number>d"
453 # Enable the legacy Jobs API. This value must be a string.
454 # 'auto' -- (default) enable the Jobs API only if it has been used before
455 # (i.e., there are job records in the database)
456 # 'true' -- enable the Jobs API despite lack of existing records.
457 # 'false' -- disable the Jobs API despite presence of existing records.
460 # Git repositories must be readable by api server, or you won't be
461 # able to submit crunch jobs. To pass the test suites, put a clone
462 # of the arvados tree in {git_repositories_dir}/arvados.git or
463 # {git_repositories_dir}/arvados/.git
464 GitInternalDir: /var/lib/arvados/internal.git
466 # Docker image to be used when none found in runtime_constraints of a job
467 DefaultDockerImage: ""
469 # none or slurm_immediate
470 CrunchJobWrapper: none
472 # username, or false = do not set uid when running jobs.
473 CrunchJobUser: crunch
475 # The web service must be able to create/write this file, and
476 # crunch-job must be able to stat() it.
477 CrunchRefreshTrigger: /tmp/crunch_refresh_trigger
479 # Control job reuse behavior when two completed jobs match the
480 # search criteria and have different outputs.
482 # If true, in case of a conflict, reuse the earliest job (this is
483 # similar to container reuse behavior).
485 # If false, in case of a conflict, do not reuse any completed job,
486 # but do reuse an already-running job if available (this is the
487 # original job reuse behavior, and is still the default).
488 ReuseJobIfOutputsDiffer: false
491 # Enable the cloud scheduler (experimental).
494 # Name/number of port where workers' SSH services listen.
497 # Interval between queue polls.
500 # Shell command to execute on each worker to determine whether
501 # the worker is booted and ready to run containers. It should
502 # exit zero if the worker is ready.
503 BootProbeCommand: "docker ps"
505 # Minimum interval between consecutive probes to a single
509 # Maximum probes per second, across all workers in a pool.
510 MaxProbesPerSecond: 10
512 # Time before repeating SIGTERM when killing a container.
515 # Time to give up on SIGTERM and write off the worker.
518 # Maximum create/destroy-instance operations per second (0 =
520 MaxCloudOpsPerSecond: 0
522 # Interval between cloud provider syncs/updates ("list all
526 # Time to leave an idle worker running (in case new containers
527 # appear in the queue that it can run) before shutting it
531 # Time to wait for a new worker to boot (i.e., pass
532 # BootProbeCommand) before giving up and shutting it down.
535 # Maximum time a worker can stay alive with no successful
536 # probes before being automatically shut down.
539 # Time after shutting down a worker to retry the
540 # shutdown/destroy operation.
543 # Worker VM image ID.
544 ImageID: ami-01234567890abcdef
546 # Tags to add on all resources (VMs, NICs, disks) created by
547 # the container dispatcher. (Arvados's own tags --
548 # InstanceType, IdleBehavior, and InstanceSecret -- will also
553 # Prefix for predefined tags used by Arvados (InstanceSetID,
554 # InstanceType, InstanceSecret, IdleBehavior). With the
555 # default value "Arvados", tags are "ArvadosInstanceSetID",
556 # "ArvadosInstanceSecret", etc.
558 # This should only be changed while no cloud resources are in
559 # use and the cloud dispatcher is not running. Otherwise,
560 # VMs/resources that were added using the old tag prefix will
561 # need to be detected and cleaned up manually.
562 TagKeyPrefix: Arvados
564 # Cloud driver: "azure" (Microsoft Azure) or "ec2" (Amazon AWS).
567 # Cloud-specific driver parameters.
574 # (ec2) Instance configuration.
580 AdminUsername: debian
582 # (azure) Credentials.
588 # (azure) Instance configuration.
589 CloudEnvironment: AzurePublicCloud
596 DeleteDanglingResourcesAfter: 20s
597 AdminUsername: arvados
601 # Use the instance type name as the key (in place of "SAMPLE" in
602 # this sample entry).
604 # Cloud provider's instance type. Defaults to the configured type name.
608 IncludedScratch: 16GB
616 SendUserSetupNotificationEmail: ""
617 IssueReporterEmailFrom: ""
618 IssueReporterEmailTo: ""
619 SupportEmailAddress: ""
629 Host: sample.arvadosapi.com