1 // Copyright (C) The Arvados Authors. All rights reserved.
3 // SPDX-License-Identifier: AGPL-3.0
7 var DefaultYAML = []byte(`# Copyright (C) The Arvados Authors. All rights reserved.
9 # SPDX-License-Identifier: AGPL-3.0
11 # Do not use this file for site configuration. Create
12 # /etc/arvados/config.yml instead.
14 # The order of precedence (highest to lowest):
15 # 1. Legacy component-specific config files (deprecated)
16 # 2. /etc/arvados/config.yml
17 # 3. config.default.yml
23 # Token to be included in all healthcheck requests. Disabled by default.
24 # Server expects request header of the format "Authorization: Bearer xxx"
79 # max concurrent connections per arvados server daemon
82 # All parameters here are passed to the PG client library in a connection string;
83 # see https://www.postgresql.org/docs/current/static/libpq-connect.html#LIBPQ-PARAMKEYWORDS
90 # Maximum size (in bytes) allowed for a single API request. This
91 # limit is published in the discovery document for use by clients.
92 # Note: You must separately configure the upstream web server or
93 # proxy to actually enforce the desired maximum request size on the
95 MaxRequestSize: 134217728
97 # Limit the number of bytes read from the database during an index
98 # request (by retrieving and returning fewer rows than would
99 # normally be returned in a single response).
100 # Note 1: This setting never reduces the number of returned rows to
101 # zero, no matter how big the first data row is.
102 # Note 2: Currently, this is only checked against a specific set of
103 # columns that tend to get large (collections.manifest_text,
104 # containers.mounts, workflows.definition). Other fields (e.g.,
105 # "properties" hashes) are not counted against this limit.
106 MaxIndexDatabaseRead: 134217728
108 # Maximum number of items to return when responding to a APIs that
109 # can return partial result sets using limit and offset parameters
110 # (e.g., *.index, groups.contents). If a request specifies a "limit"
111 # parameter higher than this value, this value is used instead.
112 MaxItemsPerResponse: 1000
114 # API methods to disable. Disabled methods are not listed in the
115 # discovery document, and respond 404 to all requests.
116 # Example: ["jobs.create", "pipeline_instances.create"]
119 # Interval (seconds) between asynchronous permission view updates. Any
120 # permission-updating API called with the 'async' parameter schedules a an
121 # update on the permission view in the future, if not already scheduled.
122 AsyncPermissionsUpdateInterval: 20
124 # Maximum number of concurrent outgoing requests to make while
125 # serving a single incoming multi-cluster (federated) request.
126 MaxRequestAmplification: 4
128 # RailsSessionSecretToken is a string of alphanumeric characters
129 # used by Rails to sign session tokens. IMPORTANT: This is a
130 # site secret. It should be at least 50 characters.
131 RailsSessionSecretToken: ""
133 # Maximum wall clock time to spend handling an incoming request.
137 # Config parameters to automatically setup new users. If enabled,
138 # this users will be able to self-activate. Enable this if you want
139 # to run an open instance where anyone can create an account and use
140 # the system without requiring manual approval.
142 # The params auto_setup_new_users_with_* are meaningful only when auto_setup_new_users is turned on.
143 # auto_setup_name_blacklist is a list of usernames to be blacklisted for auto setup.
144 AutoSetupNewUsers: false
145 AutoSetupNewUsersWithVmUUID: ""
146 AutoSetupNewUsersWithRepository: false
147 AutoSetupUsernameBlacklist: [arvados, git, gitolite, gitolite-admin, root, syslog]
149 # When new_users_are_active is set to true, new users will be active
150 # immediately. This skips the "self-activate" step which enforces
151 # user agreements. Should only be enabled for development.
152 NewUsersAreActive: false
154 # The e-mail address of the user you would like to become marked as an admin
155 # user on their first login.
156 # In the default configuration, authentication happens through the Arvados SSO
157 # server, which uses OAuth2 against Google's servers, so in that case this
158 # should be an address associated with a Google account.
159 AutoAdminUserWithEmail: ""
161 # If auto_admin_first_user is set to true, the first user to log in when no
162 # other admin users exist will automatically become an admin user.
163 AutoAdminFirstUser: false
165 # Email address to notify whenever a user creates a profile for the
167 UserProfileNotificationAddress: ""
168 AdminNotifierEmailFrom: arvados@example.com
169 EmailSubjectPrefix: "[ARVADOS] "
170 UserNotifierEmailFrom: arvados@example.com
171 NewUserNotificationRecipients: []
172 NewInactiveUserNotificationRecipients: []
175 # Time to keep audit logs, in seconds. (An audit log is a row added
176 # to the "logs" table in the PostgreSQL database each time an
177 # Arvados object is created, modified, or deleted.)
179 # Currently, websocket event notifications rely on audit logs, so
180 # this should not be set lower than 300 (5 minutes).
183 # Maximum number of log rows to delete in a single SQL transaction.
185 # If max_audit_log_delete_batch is 0, log entries will never be
186 # deleted by Arvados. Cleanup can be done by an external process
187 # without affecting any Arvados system processes, as long as very
188 # recent (<5 minutes old) logs are not deleted.
190 # 100000 is a reasonable batch size for most sites.
193 # Attributes to suppress in events and audit logs. Notably,
194 # specifying ["manifest_text"] here typically makes the database
195 # smaller and faster.
197 # Warning: Using any non-empty value here can have undesirable side
198 # effects for any client or component that relies on event logs.
199 # Use at your own risk.
200 UnloggedAttributes: []
204 # Logging threshold: panic, fatal, error, warn, info, debug, or
208 # Logging format: json or text
211 # Maximum characters of (JSON-encoded) query parameters to include
212 # in each request log entry. When params exceed this size, they will
213 # be JSON-encoded, truncated to this size, and logged as
215 MaxRequestLogParamsSize: 2000
218 # Allow clients to create collections by providing a manifest with
219 # unsigned data blob locators. IMPORTANT: This effectively disables
220 # access controls for data stored in Keep: a client who knows a hash
221 # can write a manifest that references the hash, pass it to
222 # collections.create (which will create a permission link), use
223 # collections.get to obtain a signature for that data locator, and
224 # use that signed locator to retrieve the data from Keep. Therefore,
225 # do not turn this on if your users expect to keep data private from
229 # blob_signing_key is a string of alphanumeric characters used to
230 # generate permission signatures for Keep locators. It must be
231 # identical to the permission key given to Keep. IMPORTANT: This is
232 # a site secret. It should be at least 50 characters.
234 # Modifying blob_signing_key will invalidate all existing
235 # signatures, which can cause programs to fail (e.g., arv-put,
236 # arv-get, and Crunch jobs). To avoid errors, rotate keys only when
237 # no such processes are running.
240 # Default replication level for collections. This is used when a
241 # collection's replication_desired attribute is nil.
242 DefaultReplication: 2
244 # Lifetime (in seconds) of blob permission signatures generated by
245 # the API server. This determines how long a client can take (after
246 # retrieving a collection record) to retrieve the collection data
247 # from Keep. If the client needs more time than that (assuming the
248 # collection still has the same content and the relevant user/token
249 # still has permission) the client can retrieve the collection again
250 # to get fresh signatures.
252 # This must be exactly equal to the -blob-signature-ttl flag used by
253 # keepstore servers. Otherwise, reading data blocks and saving
254 # collections will fail with HTTP 403 permission errors.
256 # Modifying blob_signature_ttl invalidates existing signatures; see
257 # blob_signing_key note above.
259 # The default is 2 weeks.
262 # Default lifetime for ephemeral collections: 2 weeks. This must not
263 # be less than blob_signature_ttl.
264 DefaultTrashLifetime: 336h
266 # Interval (seconds) between trash sweeps. During a trash sweep,
267 # collections are marked as trash if their trash_at time has
268 # arrived, and deleted if their delete_at time has arrived.
269 TrashSweepInterval: 60
271 # If true, enable collection versioning.
272 # When a collection's preserve_version field is true or the current version
273 # is older than the amount of seconds defined on preserve_version_if_idle,
274 # a snapshot of the collection's previous state is created and linked to
275 # the current collection.
276 CollectionVersioning: false
278 # 0 = auto-create a new version on every update.
279 # -1 = never auto-create new versions.
280 # > 0 = auto-create a new version when older than the specified number of seconds.
281 PreserveVersionIfIdle: -1
283 # Default collection properties. At creation time, if the client didn't
284 # provide the listed keys, they will be automatically populated following
285 # one of the following behaviors:
287 # * UUID of the user who owns the containing project.
288 # responsible_person_uuid: {function: original_owner, protected: true}
290 # * Default concrete value.
291 # foo_bar: {value: baz, protected: false}
293 # If protected is true, only an admin user can modify its value.
294 DefaultProperties: {}
297 # These settings are provided by your OAuth2 provider (e.g.,
299 ProviderAppSecret: ""
303 # Git repositories must be readable by api server, or you won't be
304 # able to submit crunch jobs. To pass the test suites, put a clone
305 # of the arvados tree in {git_repositories_dir}/arvados.git or
306 # {git_repositories_dir}/arvados/.git
307 Repositories: /var/lib/arvados/git/repositories
315 # List of supported Docker Registry image formats that compute nodes
316 # are able to use. ` + "`" + `arv keep docker` + "`" + ` will error out if a user tries
317 # to store an image with an unsupported format. Use an empty array
318 # to skip the compatibility check (and display a warning message to
321 # Example for sites running docker < 1.10: ["v1"]
322 # Example for sites running docker >= 1.10: ["v2"]
323 # Example for disabling check: []
324 SupportedDockerImageFormats: ["v2"]
326 # Include details about job reuse decisions in the server log. This
327 # causes additional database queries to run, so it should not be
328 # enabled unless you expect to examine the resulting logs for
329 # troubleshooting purposes.
330 LogReuseDecisions: false
332 # Default value for keep_cache_ram of a container's runtime_constraints.
333 DefaultKeepCacheRAM: 268435456
335 # Number of times a container can be unlocked before being
336 # automatically cancelled.
337 MaxDispatchAttempts: 5
339 # Default value for container_count_max for container requests. This is the
340 # number of times Arvados will create a new container to satisfy a container
341 # request. If a container is cancelled it will retry a new container if
342 # container_count < container_count_max on any container requests associated
343 # with the cancelled container.
346 # The maximum number of compute nodes that can be in use simultaneously
347 # If this limit is reduced, any existing nodes with slot number >= new limit
348 # will not be counted against the new limit. In other words, the new limit
349 # won't be strictly enforced until those nodes with higher slot numbers
353 # Preemptible instance support (e.g. AWS Spot Instances)
354 # When true, child containers will get created with the preemptible
355 # scheduling parameter parameter set.
356 UsePreemptibleInstances: false
358 # Include details about job reuse decisions in the server log. This
359 # causes additional database queries to run, so it should not be
360 # enabled unless you expect to examine the resulting logs for
361 # troubleshooting purposes.
362 LogReuseDecisions: false
364 # PEM encoded SSH key (RSA, DSA, or ECDSA) used by the
365 # (experimental) cloud dispatcher for executing containers on
366 # worker VMs. Begins with "-----BEGIN RSA PRIVATE KEY-----\n"
367 # and ends with "\n-----END RSA PRIVATE KEY-----\n".
368 DispatchPrivateKey: none
370 # Maximum time to wait for workers to come up before abandoning
371 # stale locks from a previous dispatch process.
375 # When you run the db:delete_old_container_logs task, it will find
376 # containers that have been finished for at least this many seconds,
377 # and delete their stdout, stderr, arv-mount, crunch-run, and
378 # crunchstat logs from the logs table.
381 # These two settings control how frequently log events are flushed to the
382 # database. Log lines are buffered until either crunch_log_bytes_per_event
383 # has been reached or crunch_log_seconds_between_events has elapsed since
385 LogBytesPerEvent: 4096
386 LogSecondsBetweenEvents: 1
388 # The sample period for throttling logs, in seconds.
389 LogThrottlePeriod: 60
391 # Maximum number of bytes that job can log over crunch_log_throttle_period
392 # before being silenced until the end of the period.
393 LogThrottleBytes: 65536
395 # Maximum number of lines that job can log over crunch_log_throttle_period
396 # before being silenced until the end of the period.
397 LogThrottleLines: 1024
399 # Maximum bytes that may be logged by a single job. Log bytes that are
400 # silenced by throttling are not counted against this total.
401 LimitLogBytesPerJob: 67108864
403 LogPartialLineThrottlePeriod: 5
405 # Container logs are written to Keep and saved in a collection,
406 # which is updated periodically while the container runs. This
407 # value sets the interval (given in seconds) between collection
409 LogUpdatePeriod: 1800
411 # The log collection is also updated when the specified amount of
412 # log data (given in bytes) is produced in less than one update
414 LogUpdateSize: 33554432
418 # Path to dns server configuration directory
419 # (e.g. /etc/unbound.d/conf.d). If false, do not write any config
420 # files or touch restart.txt (see below).
423 # Template file for the dns server host snippets. See
424 # unbound.template in this directory for an example. If false, do
425 # not write any config files.
426 DNSServerConfTemplate: ""
428 # String to write to {dns_server_conf_dir}/restart.txt (with a
429 # trailing newline) after updating local data. If false, do not
430 # open or write the restart.txt file.
431 DNSServerReloadCommand: ""
433 # Command to run after each DNS update. Template variables will be
434 # substituted; see the "unbound" example below. If false, do not run
436 DNSServerUpdateCommand: ""
438 ComputeNodeDomain: ""
439 ComputeNodeNameservers:
442 # Hostname to assign to a compute node when it sends a "ping" and the
443 # hostname in its Node record is nil.
444 # During bootstrapping, the "ping" script is expected to notice the
445 # hostname given in the ping response, and update its unix hostname
447 # If false, leave the hostname alone (this is appropriate if your compute
448 # nodes' hostnames are already assigned by some other mechanism).
450 # One way or another, the hostnames of your node records should agree
451 # with your DNS records and your /etc/slurm-llnl/slurm.conf files.
453 # Example for compute0000, compute0001, ....:
454 # assign_node_hostname: compute%<slot_number>04d
455 # (See http://ruby-doc.org/core-2.2.2/Kernel.html#method-i-format for more.)
456 AssignNodeHostname: "compute%<slot_number>d"
459 # Enable the legacy Jobs API. This value must be a string.
460 # 'auto' -- (default) enable the Jobs API only if it has been used before
461 # (i.e., there are job records in the database)
462 # 'true' -- enable the Jobs API despite lack of existing records.
463 # 'false' -- disable the Jobs API despite presence of existing records.
466 # Git repositories must be readable by api server, or you won't be
467 # able to submit crunch jobs. To pass the test suites, put a clone
468 # of the arvados tree in {git_repositories_dir}/arvados.git or
469 # {git_repositories_dir}/arvados/.git
470 GitInternalDir: /var/lib/arvados/internal.git
472 # Docker image to be used when none found in runtime_constraints of a job
473 DefaultDockerImage: ""
475 # none or slurm_immediate
476 CrunchJobWrapper: none
478 # username, or false = do not set uid when running jobs.
479 CrunchJobUser: crunch
481 # The web service must be able to create/write this file, and
482 # crunch-job must be able to stat() it.
483 CrunchRefreshTrigger: /tmp/crunch_refresh_trigger
485 # Control job reuse behavior when two completed jobs match the
486 # search criteria and have different outputs.
488 # If true, in case of a conflict, reuse the earliest job (this is
489 # similar to container reuse behavior).
491 # If false, in case of a conflict, do not reuse any completed job,
492 # but do reuse an already-running job if available (this is the
493 # original job reuse behavior, and is still the default).
494 ReuseJobIfOutputsDiffer: false
497 # Enable the cloud scheduler (experimental).
500 # Name/number of port where workers' SSH services listen.
503 # Interval between queue polls.
506 # Shell command to execute on each worker to determine whether
507 # the worker is booted and ready to run containers. It should
508 # exit zero if the worker is ready.
509 BootProbeCommand: "docker ps"
511 # Minimum interval between consecutive probes to a single
515 # Maximum probes per second, across all workers in a pool.
516 MaxProbesPerSecond: 10
518 # Time before repeating SIGTERM when killing a container.
521 # Time to give up on SIGTERM and write off the worker.
524 # Maximum create/destroy-instance operations per second (0 =
526 MaxCloudOpsPerSecond: 0
528 # Interval between cloud provider syncs/updates ("list all
532 # Time to leave an idle worker running (in case new containers
533 # appear in the queue that it can run) before shutting it
537 # Time to wait for a new worker to boot (i.e., pass
538 # BootProbeCommand) before giving up and shutting it down.
541 # Maximum time a worker can stay alive with no successful
542 # probes before being automatically shut down.
545 # Time after shutting down a worker to retry the
546 # shutdown/destroy operation.
549 # Worker VM image ID.
550 ImageID: ami-01234567890abcdef
552 # Tags to add on all resources (VMs, NICs, disks) created by
553 # the container dispatcher. (Arvados's own tags --
554 # InstanceType, IdleBehavior, and InstanceSecret -- will also
559 # Prefix for predefined tags used by Arvados (InstanceSetID,
560 # InstanceType, InstanceSecret, IdleBehavior). With the
561 # default value "Arvados", tags are "ArvadosInstanceSetID",
562 # "ArvadosInstanceSecret", etc.
564 # This should only be changed while no cloud resources are in
565 # use and the cloud dispatcher is not running. Otherwise,
566 # VMs/resources that were added using the old tag prefix will
567 # need to be detected and cleaned up manually.
568 TagKeyPrefix: Arvados
570 # Cloud driver: "azure" (Microsoft Azure) or "ec2" (Amazon AWS).
573 # Cloud-specific driver parameters.
580 # (ec2) Instance configuration.
586 AdminUsername: debian
588 # (azure) Credentials.
594 # (azure) Instance configuration.
595 CloudEnvironment: AzurePublicCloud
602 DeleteDanglingResourcesAfter: 20s
603 AdminUsername: arvados
607 # Use the instance type name as the key (in place of "SAMPLE" in
608 # this sample entry).
610 # Cloud provider's instance type. Defaults to the configured type name.
614 IncludedScratch: 16GB
622 SendUserSetupNotificationEmail: ""
623 IssueReporterEmailFrom: ""
624 IssueReporterEmailTo: ""
625 SupportEmailAddress: ""
635 Host: sample.arvadosapi.com