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"
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
119 # Config parameters to automatically setup new users. If enabled,
120 # this users will be able to self-activate. Enable this if you want
121 # to run an open instance where anyone can create an account and use
122 # the system without requiring manual approval.
124 # The params auto_setup_new_users_with_* are meaningful only when auto_setup_new_users is turned on.
125 # auto_setup_name_blacklist is a list of usernames to be blacklisted for auto setup.
126 AutoSetupNewUsers: false
127 AutoSetupNewUsersWithVmUUID: ""
128 AutoSetupNewUsersWithRepository: false
129 AutoSetupUsernameBlacklist: [arvados, git, gitolite, gitolite-admin, root, syslog]
131 # When new_users_are_active is set to true, new users will be active
132 # immediately. This skips the "self-activate" step which enforces
133 # user agreements. Should only be enabled for development.
134 NewUsersAreActive: false
136 # The e-mail address of the user you would like to become marked as an admin
137 # user on their first login.
138 # In the default configuration, authentication happens through the Arvados SSO
139 # server, which uses OAuth2 against Google's servers, so in that case this
140 # should be an address associated with a Google account.
141 AutoAdminUserWithEmail: ""
143 # If auto_admin_first_user is set to true, the first user to log in when no
144 # other admin users exist will automatically become an admin user.
145 AutoAdminFirstUser: false
147 # Email address to notify whenever a user creates a profile for the
149 UserProfileNotificationAddress: ""
150 AdminNotifierEmailFrom: arvados@example.com
151 EmailSubjectPrefix: "[ARVADOS] "
152 UserNotifierEmailFrom: arvados@example.com
153 NewUserNotificationRecipients: []
154 NewInactiveUserNotificationRecipients: []
157 # Time to keep audit logs, in seconds. (An audit log is a row added
158 # to the "logs" table in the PostgreSQL database each time an
159 # Arvados object is created, modified, or deleted.)
161 # Currently, websocket event notifications rely on audit logs, so
162 # this should not be set lower than 600 (5 minutes).
165 # Maximum number of log rows to delete in a single SQL transaction.
167 # If max_audit_log_delete_batch is 0, log entries will never be
168 # deleted by Arvados. Cleanup can be done by an external process
169 # without affecting any Arvados system processes, as long as very
170 # recent (<5 minutes old) logs are not deleted.
172 # 100000 is a reasonable batch size for most sites.
175 # Attributes to suppress in events and audit logs. Notably,
176 # specifying ["manifest_text"] here typically makes the database
177 # smaller and faster.
179 # Warning: Using any non-empty value here can have undesirable side
180 # effects for any client or component that relies on event logs.
181 # Use at your own risk.
182 UnloggedAttributes: []
185 # Maximum characters of (JSON-encoded) query parameters to include
186 # in each request log entry. When params exceed this size, they will
187 # be JSON-encoded, truncated to this size, and logged as
189 MaxRequestLogParamsSize: 2000
192 # Allow clients to create collections by providing a manifest with
193 # unsigned data blob locators. IMPORTANT: This effectively disables
194 # access controls for data stored in Keep: a client who knows a hash
195 # can write a manifest that references the hash, pass it to
196 # collections.create (which will create a permission link), use
197 # collections.get to obtain a signature for that data locator, and
198 # use that signed locator to retrieve the data from Keep. Therefore,
199 # do not turn this on if your users expect to keep data private from
203 # blob_signing_key is a string of alphanumeric characters used to
204 # generate permission signatures for Keep locators. It must be
205 # identical to the permission key given to Keep. IMPORTANT: This is
206 # a site secret. It should be at least 50 characters.
208 # Modifying blob_signing_key will invalidate all existing
209 # signatures, which can cause programs to fail (e.g., arv-put,
210 # arv-get, and Crunch jobs). To avoid errors, rotate keys only when
211 # no such processes are running.
214 # Default replication level for collections. This is used when a
215 # collection's replication_desired attribute is nil.
216 DefaultReplication: 2
218 # Lifetime (in seconds) of blob permission signatures generated by
219 # the API server. This determines how long a client can take (after
220 # retrieving a collection record) to retrieve the collection data
221 # from Keep. If the client needs more time than that (assuming the
222 # collection still has the same content and the relevant user/token
223 # still has permission) the client can retrieve the collection again
224 # to get fresh signatures.
226 # This must be exactly equal to the -blob-signature-ttl flag used by
227 # keepstore servers. Otherwise, reading data blocks and saving
228 # collections will fail with HTTP 403 permission errors.
230 # Modifying blob_signature_ttl invalidates existing signatures; see
231 # blob_signing_key note above.
233 # The default is 2 weeks.
234 BlobSigningTTL: 1209600
236 # Default lifetime for ephemeral collections: 2 weeks. This must not
237 # be less than blob_signature_ttl.
238 DefaultTrashLifetime: 1209600
240 # Interval (seconds) between trash sweeps. During a trash sweep,
241 # collections are marked as trash if their trash_at time has
242 # arrived, and deleted if their delete_at time has arrived.
243 TrashSweepInterval: 60
245 # If true, enable collection versioning.
246 # When a collection's preserve_version field is true or the current version
247 # is older than the amount of seconds defined on preserve_version_if_idle,
248 # a snapshot of the collection's previous state is created and linked to
249 # the current collection.
250 CollectionVersioning: false
252 # 0 = auto-create a new version on every update.
253 # -1 = never auto-create new versions.
254 # > 0 = auto-create a new version when older than the specified number of seconds.
255 PreserveVersionIfIdle: -1
258 # These settings are provided by your OAuth2 provider (e.g.,
260 ProviderAppSecret: ""
264 # Git repositories must be readable by api server, or you won't be
265 # able to submit crunch jobs. To pass the test suites, put a clone
266 # of the arvados tree in {git_repositories_dir}/arvados.git or
267 # {git_repositories_dir}/arvados/.git
268 Repositories: /var/lib/arvados/git/repositories
274 # List of supported Docker Registry image formats that compute nodes
275 # are able to use. ` + "`" + `arv keep docker` + "`" + ` will error out if a user tries
276 # to store an image with an unsupported format. Use an empty array
277 # to skip the compatibility check (and display a warning message to
280 # Example for sites running docker < 1.10: ["v1"]
281 # Example for sites running docker >= 1.10: ["v2"]
282 # Example for disabling check: []
283 SupportedDockerImageFormats: ["v2"]
285 # Include details about job reuse decisions in the server log. This
286 # causes additional database queries to run, so it should not be
287 # enabled unless you expect to examine the resulting logs for
288 # troubleshooting purposes.
289 LogReuseDecisions: false
291 # Default value for keep_cache_ram of a container's runtime_constraints.
292 DefaultKeepCacheRAM: 268435456
294 # Number of times a container can be unlocked before being
295 # automatically cancelled.
296 MaxDispatchAttempts: 5
298 # Default value for container_count_max for container requests. This is the
299 # number of times Arvados will create a new container to satisfy a container
300 # request. If a container is cancelled it will retry a new container if
301 # container_count < container_count_max on any container requests associated
302 # with the cancelled container.
305 # The maximum number of compute nodes that can be in use simultaneously
306 # If this limit is reduced, any existing nodes with slot number >= new limit
307 # will not be counted against the new limit. In other words, the new limit
308 # won't be strictly enforced until those nodes with higher slot numbers
312 # Preemptible instance support (e.g. AWS Spot Instances)
313 # When true, child containers will get created with the preemptible
314 # scheduling parameter parameter set.
315 UsePreemptibleInstances: false
317 # Include details about job reuse decisions in the server log. This
318 # causes additional database queries to run, so it should not be
319 # enabled unless you expect to examine the resulting logs for
320 # troubleshooting purposes.
321 LogReuseDecisions: false
324 # When you run the db:delete_old_container_logs task, it will find
325 # containers that have been finished for at least this many seconds,
326 # and delete their stdout, stderr, arv-mount, crunch-run, and
327 # crunchstat logs from the logs table.
330 # These two settings control how frequently log events are flushed to the
331 # database. Log lines are buffered until either crunch_log_bytes_per_event
332 # has been reached or crunch_log_seconds_between_events has elapsed since
334 LogBytesPerEvent: 4096
335 LogSecondsBetweenEvents: 1
337 # The sample period for throttling logs, in seconds.
338 LogThrottlePeriod: 60
340 # Maximum number of bytes that job can log over crunch_log_throttle_period
341 # before being silenced until the end of the period.
342 LogThrottleBytes: 65536
344 # Maximum number of lines that job can log over crunch_log_throttle_period
345 # before being silenced until the end of the period.
346 LogThrottleLines: 1024
348 # Maximum bytes that may be logged by a single job. Log bytes that are
349 # silenced by throttling are not counted against this total.
350 LimitLogBytesPerJob: 67108864
352 LogPartialLineThrottlePeriod: 5
354 # Container logs are written to Keep and saved in a collection,
355 # which is updated periodically while the container runs. This
356 # value sets the interval (given in seconds) between collection
358 LogUpdatePeriod: 1800
360 # The log collection is also updated when the specified amount of
361 # log data (given in bytes) is produced in less than one update
363 LogUpdateSize: 33554432
367 # Path to dns server configuration directory
368 # (e.g. /etc/unbound.d/conf.d). If false, do not write any config
369 # files or touch restart.txt (see below).
372 # Template file for the dns server host snippets. See
373 # unbound.template in this directory for an example. If false, do
374 # not write any config files.
375 DNSServerConfTemplate: ""
377 # String to write to {dns_server_conf_dir}/restart.txt (with a
378 # trailing newline) after updating local data. If false, do not
379 # open or write the restart.txt file.
380 DNSServerReloadCommand: ""
382 # Command to run after each DNS update. Template variables will be
383 # substituted; see the "unbound" example below. If false, do not run
385 DNSServerUpdateCommand: ""
387 ComputeNodeDomain: ""
388 ComputeNodeNameservers:
391 # Hostname to assign to a compute node when it sends a "ping" and the
392 # hostname in its Node record is nil.
393 # During bootstrapping, the "ping" script is expected to notice the
394 # hostname given in the ping response, and update its unix hostname
396 # If false, leave the hostname alone (this is appropriate if your compute
397 # nodes' hostnames are already assigned by some other mechanism).
399 # One way or another, the hostnames of your node records should agree
400 # with your DNS records and your /etc/slurm-llnl/slurm.conf files.
402 # Example for compute0000, compute0001, ....:
403 # assign_node_hostname: compute%<slot_number>04d
404 # (See http://ruby-doc.org/core-2.2.2/Kernel.html#method-i-format for more.)
405 AssignNodeHostname: "compute%<slot_number>d"
408 # Enable the legacy Jobs API. This value must be a string.
409 # 'auto' -- (default) enable the Jobs API only if it has been used before
410 # (i.e., there are job records in the database)
411 # 'true' -- enable the Jobs API despite lack of existing records.
412 # 'false' -- disable the Jobs API despite presence of existing records.
415 # Git repositories must be readable by api server, or you won't be
416 # able to submit crunch jobs. To pass the test suites, put a clone
417 # of the arvados tree in {git_repositories_dir}/arvados.git or
418 # {git_repositories_dir}/arvados/.git
419 GitInternalDir: /var/lib/arvados/internal.git
421 # Docker image to be used when none found in runtime_constraints of a job
422 DefaultDockerImage: ""
424 # none or slurm_immediate
425 CrunchJobWrapper: none
427 # username, or false = do not set uid when running jobs.
428 CrunchJobUser: crunch
430 # The web service must be able to create/write this file, and
431 # crunch-job must be able to stat() it.
432 CrunchRefreshTrigger: /tmp/crunch_refresh_trigger
434 # Control job reuse behavior when two completed jobs match the
435 # search criteria and have different outputs.
437 # If true, in case of a conflict, reuse the earliest job (this is
438 # similar to container reuse behavior).
440 # If false, in case of a conflict, do not reuse any completed job,
441 # but do reuse an already-running job if available (this is the
442 # original job reuse behavior, and is still the default).
443 ReuseJobIfOutputsDiffer: false
448 SendUserSetupNotificationEmail: ""
449 IssueReporterEmailFrom: ""
450 IssueReporterEmailTo: ""
451 SupportEmailAddress: ""