CHANGELOG.md

Changelog

All notable changes to this project will be documented in this file.

The format is based on Keep a Changelog and this project adheres to Semantic Versioning.

Unreleased

Added

• [astarte_realm_management] Allow listing interfaces with mappings using the detailed=true parameter
• [astarte_import] Added support for data types: doublearray, integerarray, booleanarray, longintegerarray, stringarray, datetimearray, binaryblobarray.

[1.3.0] - Unreleased

Changed

• [astarte_pairing] FDO authentication (EXPERIMENTAL feature, disabled by default). New environment variables are needed in order to use FDO:

  • PAIRING_ENABLE_FDO - whether the FDO feature is enabled or not (default: false)
  • PAIRING_FDO_RENDEZVOUS_URL - URL of the rendezvous server (default: "http://rendezvous:8041")
  • ASTARTE_BASE_URL_DOMAIN - domain part of the base URL of astarte, used by devices to connect in TO2 phase (required if FDO enabled)
  • ASTARTE_BASE_URL_PORT - port of the base URL of astarte (required if FDO enabled)
  • ASTARTE_BASE_URL_PROTOCOL - protocol of the base URL of astarte (required if FDO enabled)
  • PAIRING_FDO_RENDEZVOUS_SSL_ENABLED - whether SSL is enabled for the connection to the rendezvous server (default: false)
  • PAIRING_FDO_RENDEZVOUS_SSL_CA_FILE - path to the CA certificate file for the rendezvous server TLS connection; when not specified, the bundled cURL certificate bundle will be used
  • PAIRING_FDO_RENDEZVOUS_SSL_DISABLE_SNI - disable Server Name Indication for the rendezvous server TLS connection (default: false)
  • PAIRING_FDO_RENDEZVOUS_SSL_CUSTOM_SNI - custom SNI hostname for the rendezvous server TLS connection; when not specified, the Rendezvous hostname will be used

• [astarte_housekeeping] AMQP management configuration moved to astarte_events library. Environment variables changed:

  • HOUSEKEEPING_AMQP_SSL_ENABLED is now ASTARTE_EVENTS_AMQP_MANAGEMENT_SSL_ENABLED
  • HOUSEKEEPING_AMQP_SSL_CA_FILE is now ASTARTE_EVENTS_AMQP_MANAGEMENT_SSL_CA_FILE
  • HOUSEKEEPING_AMQP_SSL_DISABLE_SNI is now ASTARTE_EVENTS_AMQP_MANAGEMENT_SSL_DISABLE_SNI
  • HOUSEKEEPING_AMQP_SSL_CUSTOM_SNI is now ASTARTE_EVENTS_AMQP_MANAGEMENT_SSL_CUSTOM_SNI
  • HOUSEKEEPING_AMQP_HOST is now ASTARTE_EVENTS_AMQP_MANAGEMENT_HOST
  • HOUSEKEEPING_AMQP_MANAGEMENT_PORT is now ASTARTE_EVENTS_AMQP_MANAGEMENT_PORT
  • HOUSEKEEPING_AMQP_USERNAME is now ASTARTE_EVENTS_AMQP_MANAGEMENT_USERNAME
  • HOUSEKEEPING_AMQP_PASSWORD is now ASTARTE_EVENTS_AMQP_MANAGEMENT_PASSWORD

[1.3.0-rc.1] - 2026-01-26

Added

• New environment variables for trigger notifications between realm management replicas and realm management -> pairing. These variables are currently being used only by realm management
  • REALM_MANAGEMENT_CLUSTERING_KUBERNETES_SELECTOR. The Endpoint label to query to get realm management instances. Defaults to app=astarte-realm-management.
  • PAIRING_CLUSTERING_KUBERNETES_SELECTOR. The Endpoint label to query to get pairing instances. Defaults to app=astarte-pairing.
• [astarte_pairing] Cluster with realm management using CLUSTERING_STRATEGY and CLUSTERING_KUBERNETES_NAMESPACE
• [astarte_pairing] Add realm-scoped health checks, which can be used to test astarte health or connectivity by devices who are only aware of realm-scoped paths, such as during FDO

Fixed

• Fix crashes in new ScyllaDB versions
• [astarte_appengine_api] Volatile trigger requests to Data Updater Plant now properly encode the volatile trigger type

Changed

• Services now receive trigger installation and deletion notifications, which should reduce the delay between installing the trigger and starting to receive messages

[1.3.0-rc.0] - 2025-11-21

Added

• [astarte_housekeeping] support network topology replication strategy for the astarte keyspace, with the following env vars:
  • HOUSEKEEPING_ASTARTE_KEYSPACE_REPLICATION_STRATEGY - Replication strategy for the astarte keyspace: "SimpleStrategy" or "NetworkTopologyStrategy" (default: "SimpleStrategy")
  • HOUSEKEEPING_ASTARTE_KEYSPACE_REPLICATION_FACTOR - Replication factor when using SimpleStrategy (default: 1)
  • HOUSEKEEPING_ASTARTE_KEYSPACE_NETWORK_REPLICATION_MAP - Datacenter replication map when using NetworkTopologyStrategy (no default, required when using network strategy)
• Added database events handling configuration across all services:
  • DATABASE_EVENTS_HANDLING_METHOD - Controls how database events are handled: "expose" (via telemetry) or "log" (to logs) (default: "expose")
• [astarte_pairing] Added device registration triggers
• [astarte_realm_management] Added device deletion started and device deletion completed triggers
• Allow devices with empty introspection
• Devices can now declare support for optional Astarte MQTT v1 features to Astarte via capabilities
• Support for purge_properties_compression_format capability. possible values are zlib (default) and plaintext

Changed

• BREAKING: Merged API services into main services, eliminating separate containers:
  • astarte-housekeeping-api merged into astarte-housekeeping
  • astarte-pairing-api merged into astarte-pairing
  • astarte-realm-management-api merged into astarte-realm-management
• BREAKING: [astarte_housekeeping] Housekeeping now creates an AMQP vhost for each created realm. Some required configuration was introduced:
  • HOUSEKEEPING_AMQP_HOST - AMQP host for housekeeping operations (required in production, defaults to "localhost" in dev/test)
  • HOUSEKEEPING_AMQP_SSL_ENABLED - Enable SSL for AMQP connections (default: false)
  • HOUSEKEEPING_AMQP_SSL_CA_FILE - CA certificate file for AMQP SSL connections (default: bundled cURL certificate bundle)
  • HOUSEKEEPING_AMQP_SSL_DISABLE_SNI - Disable Server Name Indication (default: false)
  • HOUSEKEEPING_AMQP_SSL_CUSTOM_SNI - Custom SNI hostname (defaults to AMQP host if unset)
  • HOUSEKEEPING_AMQP_USERNAME - AMQP username (default: guest)
  • HOUSEKEEPING_AMQP_PASSWORD - AMQP password (default: guest)
  • HOUSEKEEPING_AMQP_MANAGEMENT_PORT - AMQP management API port (default: 15672)
• BREAKING: [astarte_data_updater_plant] Added a separate the AMQP producer configuration, which is mandatory:
  • ASTARTE_EVENTS_PRODUCER_AMQP_HOST - Host for producer connection (default: "localhost")
  • ASTARTE_EVENTS_PRODUCER_AMQP_USERNAME - Username for producer (default: "guest")
  • ASTARTE_EVENTS_PRODUCER_AMQP_PASSWORD - Password for producer (default: "guest")
  • ASTARTE_EVENTS_PRODUCER_AMQP_VIRTUAL_HOST - Virtual host for internal events (default: "/")
  • ASTARTE_EVENTS_PRODUCER_AMQP_PORT - Port for producer (default: 5672)
  • ASTARTE_EVENTS_PRODUCER_AMQP_SSL_ENABLED - Enable SSL for producer (default: false)
  • ASTARTE_EVENTS_PRODUCER_AMQP_SSL_CA_FILE - CA certificate file for producer SSL (default: bundled cURL certificates)
  • ASTARTE_EVENTS_PRODUCER_AMQP_SSL_DISABLE_SNI - Disable Server Name Indication for producer (default: false)
  • ASTARTE_EVENTS_PRODUCER_AMQP_SSL_CUSTOM_SNI - Custom SNI hostname for producer (falls back to the value of ASTARTE_EVENTS_PRODUCER_AMQP_HOST)
  • ASTARTE_EVENTS_PRODUCER_AMQP_CONNECTION_NUMBER - The number of open connections to RabbitMQ (default: 10)
  • ASTARTE_EVENTS_PRODUCER_AMQP_DATA_QUEUE_TOTAL_COUNT - The total number of data queues in the astarte cluster (default: 128)
  • ASTARTE_EVENTS_PRODUCER_AMQP_EVENTS_EXCHANGE_NAME - The exchange for internal events (default: "astarte_events")
• BREAKING: AMQP triggers are now always sent to an ad-hoc vhost for each realm. the vhost is [astarte_instance_id]_[realm_name], which under normal circumstances is just _[realm_name]
• BREAKING: [astarte_pairing] AMQP Producer configuration is now mandatory using the ASTARTE_EVENTS_PRODUCER_AMQP_* environment variables
• BREAKING: [astarte_realm_management] AMQP Producer configuration is now mandatory using the ASTARTE_EVENTS_PRODUCER_AMQP_* environment variables
• More accurate health checks for astarte services

Fixed

• [astarte_data_updater_plant] Do not crash when transient triggers are installed on devices with outdated introspection
• [astarte_data_updater_plant] Correctly encode values when sending properties to device on connection
• [astarte_realm_management] Allow to delete long-disconnected devices. For this to work, Realm Management needs CLUSTERING_STRATEGY, CLUSTERING_KUBERNETES_NAMESPACE and DATA_UPDATER_PLANT_CLUSTERING_KUBERNETES_SELECTOR to be set, just like AppEngine and DUP. Refer to 1.2.1-rc.0 for additional information on the variables.

[1.2.1] - 2026-03-12

Fixed

• [astarte_realm_management] Insufficient validation for conflicting options in interface aggregate mappings #1072
• [astarte_data_updater_plant] Ensure device deletion rpc always checks updated status

[1.2.1-rc.1] - 2026-02-13

Fixed

• [astarte_realm_management] Bug where devices got stuck in the "in deletion" status: #1493.
• [astarte_data_updater_plant] Correctly reconnect to AMQP after a connection loss
• [astarte_data_updater_plant] Fix possible crash when sending data on interface-specific volatile triggers

[1.2.1-rc.0] - 2025-08-26

Added

• New environment variables to control how clustering work, needed on AppEngine and DUP.
  • CLUSTERING_STRATEGY. Its possible values are:
    • none (default): the service will not look for other nodes/services.
    • docker-compose: this is meant to work in our docker-compose environment and no additional configuration is needed.
    • kubernetes: other nodes/services are found thanks to kubernetes DNS, and the other variables below are relevant.
  • CLUSTERING_KUBERNETES_NAMESPACE. It states under which namespace the Astarte instance has been deployed. Defaults to astarte.
  • DATA_UPDATER_PLANT_CLUSTERING_KUBERNETES_SELECTOR. The Endpoint label to query to get other data updater plant instances. Defaults to app=astarte-data-updater-plant.
  • VERNEMQ_CLUSTERING_KUBERNETES_SELECTOR. The Pod label to use to query Kubernetes to find VerneMQ instances. Defaults to app=astarte-vernemq.
  • VERNEMQ_CLUSTERING_KUBERNETES_SERVICE_NAME. The Service name to use to query Kubernetes to find VerneMQ instances. Defaults to astarte-vernemq.

Changed

• Rework RPC between AppEngine and DUP using Erlang's native clustering and message-passing instead of AMQP queues: #1186. Fix #699.
• Rework RPC between AppEngine and VerneMQ using Erlang's native clustering and message-passing instead of AMQP queues: #1194.
• Rework RPC between DUP and VerneMQ using Erlang's native clustering and message-passing instead of AMQP queues: #1197.
• Changed the database driver from CQEx (unmantained) to (E)xandra
• [astarte_trigger_engine] avoid exposing unknown_fields in mustache templates
• [astarte_trigger_engine] properly handle incoming introspection events

Fixed

• avoid leaving dangling device deletion entries
• [astarte_appengine_api] fix a crash on invalid object update values
• [astarte_appengine_api] fix a crash with empty result and disjoint_tables format for object aggregates
• [astarte_appengine_api] Create needed AMQP exchanges instead of crashing
• [astarte_appengine_api] Handle unset of properties which don't allow being unset
• [astarte_data_updater_plant] fix a crash when performing updates of deleted values
• [astarte_data_updater_plant] Fix a crash while handling device introspection
• [astarte_data_updater_plant] Fix DataUpdater GenServer timeout handling that prevented inactive processes to shut down automatically.
• [astarte_data_updater_plant] Handle unset of properties which don't allow being unset
• [astarte_data_updater_plant] properly discard heartbeat messages when discarding messages
• [astarte_data_updater_plant] Properly reconnect to RabbitMQ in case of disconnection
• [astarte_data_updater_plant] Some queries had the astarte_instance_id applied twice
• [astarte_realm_management] avoid crashing on interface list
• [astarte_realm_management] avoid crash when deleting devices with invalid introspection
• [astarte_realm_management] ensure devices are cleaned up after being deleted

[1.2.1-alpha.0] - 2025-04-10

Changed

• Update the docker-compose configuration to allow both physical and virtual devices to connect to Astarte, provided that the devices and the host are on the same LAN.

Fixed

• [astarte_appengine_api] Correctly handle Cassandra varchars.
• [astarte_data_updater_plant] Correctly handle Cassandra varchars.
• [astarte_housekeeping] Correctly handle Cassandra varchars.
• [astarte_pairing] Correctly handle Cassandra varchars.
• [astarte_realm_management] Correctly handle Cassandra varchars.
• [astarte_trigger_engine] Correctly handle Cassandra varchars.
• [astarte_pairing] Fix a corner case in the realm public key retrieval when connection to the database might fail.
• [astarte_realm_management] Fix a corner case in the realm public key retrieval when connection to the database might fail.
• [astarte_appengine_api] Fix a corner case in the realm public key retrieval when connection to the database might fail.
• [astarte_data_updater_plant] Do not generate redundant disconnection triggers in corner cases when a device is already disconnected. Fix #1014.

[1.2.0] - 2024-07-02

Fixed

• Forward port changes from release-1.1 (connection failure when delivering triggers is handled as an error).

[1.2.0-rc.0] 11-06-2024

Added

• [astarte_trigger_engine] Add trigger_name to event payload and mustache template
• Add support for limiting the number of registered devices in a realm. Existing realms are not affected by this change.
• [astarte_realm_management_api] Allow to read realm's device registration limit using the /config/device_registration_limit endpoint.
• [astarte_housekeeping_api] Allow to read and set a realm's device registration limit using the realm fetch and update API, respectively.
• [astarte_appengine_api] Show deletion status in device details.
• [astarte_appengine_api] Add /version endpoint which returns the microservice application version.
• [astarte_appengine_api] Added group name decoder plug
• [astarte_realm_management_api] Add /version endpoint which returns the microservice application version.
• [astarte_pairing_api] Add /version endpoint which returns the microservice application version.
• [astarte_housekeeping_api] Add /version endpoint which returns the microservice application version.
• [astarte_housekeeping_api] Allow to customize the RPC call timeout with HOUSEKEEPING_API_RPC_TIMEOUT (default: 5 seconds).
• Add API usage metrics.
• Add support for setting the maximum datastream storage retention period in a realm (in seconds). Existing realms are not affected by this change.
• [astarte_housekeeping_api] Allow to read and set a realm's maximum datastream storage retention period using the realm fetch and update API, respectively.
• [astarte_realm_management_api] Allow to read realm's maximum datastream storage retention period with the /config/datastream_maximum_storage_retention endpoint.

Changed

• Forward port changes from release 1.1.
• [astarte_housekeeping_api] Introduce a PATCH-based API for realm update. The old POST-based one is deprecated.
• [astarte_housekeeping] Support the new PATCH-based API for realm update.
• Update Elixir to 1.15.7.
• Update Erlang/OTP to 26.1.
• Update container base image to Debian Bookworm.
• Container user has changed from root to nobody.
• BREAKING: incoming_introspection events display introspection as a map rather than a plaintext string. Revert to the old behaviour by setting DATA_UPDATER_PLANT_GENERATE_LEGACY_INCOMING_INTROSPECTION_EVENTS to true. See astarte-platform/astarte_core#77.
• BREAKING: [astarte_realm_management] do not allow installation of interfaces where database_retention_ttl exceeds the realm's maximum datastream storage retention period, if set.

[1.1.2] - Unreleased

Added

• [astarte_data_updater_plant] customize the number of consumer connections to the AMQP broker using DATA_UPDATER_PLANT_AMQP_CONSUMER_CONNECTION_NUMBER (default: 10).

Fixed

• Forward-port changes from release-1.0 (gracefully handle malformed purge properties messages, fix message count metrics, fix typing issues in payloads).
• [astarte_data_updater_plant] do not leak consumer channels in corner cases.
• [astarte_data_updater_plant] do not leak producer channels in corner cases.
• [astarte_trigger_engine] Always treat event TTL for trigger policies in seconds, not milliseconds.
• [astarte_trigger_engine] ack messages even with unreachable target (see astarte-platform#936)
• [astarte_appengine_api] Handle type casting for longinteger inputs
• [astarte_appengine_api] refactor device.ex file - extract cast_value functions into InterfaceValue module and add tests

[1.1.1] - 2023-11-15

Fixed

• [astarte_data_updater_plant] Don't crash when retrieving the interface version in a device whose introspection is empty, allowing data in astarte-data_ AMQP queues to flow seamlessly.
• [astarte_appengine_api] Don't crash when retrieving the interface version in a device whose introspection is empty.
• [astarte_appengine_api] Return the number of results specified by downsample_to when there are more samples than the default query limit. Fix #824.
• [astarte_appengine_api] Return the number of results specified by downsample_to when used in combination with format=disjoint_tables. Fix #149.
• [astarte-realm_management] Do not crash when some synchronous operations (e.g. interface install) succeed.
• [astarte_data_updater_plant] Don't crash when handling incoming binaryblob values (see astarte-platform/astarte_core#95).
• Forward-port changes from release-1.0 (correctly handle server-owned data, bugfixes in AppEngine, Data Updater Plant and Pairing).

Changed

• [astarte_e2e] Update Elixir to 1.14.5 and Erlang/OTP to 25.3.2. (see astarte-platform#858)

[1.1.0] - 2023-06-20

Fixed

• [astarte_trigger_engine] Allow to decode events that do not contain the deprecated version field.

[1.1.0-rc.0] - 2023-06-09

Changed

• Update Elixir to 1.14.5 and Erlang/OTP to 25.3.2.
• [astarte_data_updater_plant] Use the internal event type for Astarte internal messages. (e.g. device heartbeat).

Fixed

• [astarte_realm_management_api] Provide detailed feedback when a trigger action is malformed. Fix #748.
• [astarte_realm_management_api] Include the policy field when a trigger is returned.

[1.1.0-alpha.0] - 2022-11-24

Added

• [astarte_data_updater_plant] Add support for device introspection triggers.
• [astarte_realm_management] Add support for device introspection triggers.
• [astarte_realm_management_api] Add support for device introspection triggers.
• [astarte_realm_management] Add support for delivery policies for HTTP triggers.
• [astarte_realm_management_api] Add support for delivery policies for HTTP triggers.
• [astarte_data_updater_plant] Add a HTTP trigger routing mechanism based on trigger policies.
• [astarte_trigger_engine] Add a customizable HTTP trigger redelivery mechanism via trigger policies (see #554).

Fixed

• [astarte_appengine_api] Return empty data instead of error when querying properties interfaces which are not fully populated. Fix 531.

[1.0.6] - 2024-04-23

Fixed

• [astarte_appengine_api] Allow to send binaryblobarrays over server owned interfaces.
• [astarte_appengine_api] Doubles and DoubleArrays without decimal part are no longer saved as integer, but a trailing zero is added.
• [astarte_data_updater_plant] Do not crash when receiving a malformed purge properties message.
• [astarte_pairing_api] Gracefully handle HTTP requests with malformed payload.
• [astarte_housekeeping_api] Gracefully handle HTTP requests with malformed payload.
• [astarte_realm_management_api] Gracefully handle HTTP requests with malformed payload.
• [astarte_appengine_api] Expose exchanged_bytes metrics as sum (instead of counter).
• [astarte_pairing] Do not reset total sent messages/bytes when re-registering a device. Fix #776.

[1.0.5] - 2023-09-26

Fixed

• [astarte_appengine_api] Correctly handle binaryblob and datetime in server-owned object aggregated interfaces.
• [astarte_appengine_api] Handle non-array values POSTed to an array endpoint gracefully instead of crashing with an Internal Server Error
• [astarte_appengine_api] Handle updates of objects with invalid keys gracefully instead of crashing with an Internal Server Error.
• [astarte_appengine_api] Return the number of results specified by downsample_to when there are more samples than the default query limit. Fix #824.
• [astarte_appengine_api] Return the number of results specified by downsample_to when used in combination with format=disjoint_tables. Fix #149.
• [astarte_appengine_api] Fix log noise with cassandra during health checks. Fix #817.
• [astarte_pairing] Fix crash when using a custom CA certificate.
• [astarte_data_updater_plant] Remove asynchronous startup of Data Updater, which could lead to Data Updater stalling in some corner cases.

Changed

• [doc] Update the documentation structure. Pages dealing with administrative tasks involving the Astarte Operator and Kubernetes are moved to the astarte-kubernetes-operator repository.

[1.0.4] - 2022-10-25

Changed

• [astarte_appengine_api] Check for device existence before accepting a watch request on Astarte rooms.
• [astarte_data_updater_plant] Check for device existence before installation or deletion of volatile triggers.

[1.0.3] - 2022-07-04

Fixed

• [astarte_appengine_api] Consider allow_bigintegers and allow_safe_bigintegers params when querying the root of individual datastream / properties interfaces. Fix #630.
• [astarte_appengine_api] Correctly return 405 "Cannot write to device owned resource" when POSTing on device-owned interfaces. Fix #264.
• [astarte_appengine_api] Correctly return 405 "Cannot write to read-only resource" when POSTing on incomplete paths of server-owned interfaces.
• [astarte_e2e] Fix ssl options handling so that the e2e client is aware of the CA.

Changed

• [doc] Administrator Guide: bump cert-manager dependency to v1.7.0.
• [data_updater_plant] Increase the declare_exchange timeout to 60 sec.
• [data_updater_plant] Increase the publish timeout to 60 sec for the AMQPEventsProducer.
• [astarte_realm_management_api] Do not crash when receiving trigger errors. Fix 683.
• [astarte_e2e] Allow setting custom subjects for alerting emails.
• [astarte_e2e] Group in a single thread emails referencing the same failure_id.
• [astarte_appengine_api] Make property unset succeed independently of whether there exist a device session on the broker or not. Fix #640.
• [astarte_data_updater_plant] Log the base64-encoded object when receiving an object with unexpected key.

[1.0.2] - 2022-04-01

Added

• [realm_management] Accept retention and expiry updates when updating the minor version of an interface.
• [astarte_realm_management_api] Allow synchronous requests for interface creation, update and deletion using the async_operation option. Default to async calls.
• [astarte_housekeeping_api] Allow synchronous requests for realm creation and deletion using the async_operation option. Default to async calls.

Fixed

• [realm_management] Accept allowed mapping updates in object aggregated interfaces without crashing.
• [astarte_appengine_api] Handle server owned datetimearray values correctly.

Changed

• [astarte_housekeeping] Allow to delete a realm only if all its devices are disconnected. Realm deletion can still only be enabled with an environment variable (defaults to disabled).
• Update CA store to 2022-03-21 version.

[1.0.1] - 2021-12-17

Added

• [data_updater_plant] Add handle_data duration metric.
• [doc] Add documentation for AstarteDefaultIngress.
• [doc] Add deprecation notice for AstarteVoyagerIngress.
• [doc] Add documentation for the handling of Astarte certificates.

Changed

• [doc] Remove astartectl profiles from the possible deployment alternatives.

Fixed

• [astarte_appengine_api] Correctly serialize events containing datetime and array values.
• [astarte_appengine_api] Do not fail when querying datastream interfaces data with since, to, sinceAfter params if result is empty. Fix #552.
• [astarte_appengine_api] Consider microseconds when using timestamps. Fix #620.
• [astarte_appengine_api] Don't crash when removing an alias with non-existing tag. Fix 495.
• [astarte_trigger_engine] Correctly serialize events containing datetime and array values.
• [astarte_data_updater_plant] Don't crash when receiving binaryblobarray and datetimearray values.
• Update Cyanide BSON library, in order to fix crash when handling ill-formed BSON arrays.

[1.0.0] - 2021-06-30

Added

• Add support for volatile triggers on interfaces with object aggregation.

Changed

• Document future removal of Astarte Operator's support for Cassandra.
• Log application version when starting.

Fixed

• [astarte_appengine_api] Fix the support for null values in interfaces, the fix contained in 1.0.0-rc.0 was incomplete.

[1.0.0-rc.0] - 2021-05-10

Added

• [astarte_appengine_api] Add /v1/<realm>/version endpoint, returning the API application version.
• [astarte_realm_management_api] Add /v1/<realm>/version endpoint, returning the API application version.
• [astarte_pairing_api] Add /v1/<realm>/version endpoint, returning the API application version.
• [astarte_housekeeping_api] Add /v1/version endpoint, returning the API application version.

Changed

• [astarte_realm_management] Make amqp_routing_key mandatory in AMQP actions.
• Update documentation for backing up and restoring Astarte.
• Update documentation for Operator's uninstall procedure.

Fixed

• [astarte_appengine_api] Don't crash when an interface contains null values, just show them as null in the resulting JSON.
• [astarte_realm_management] Fix log noise due to Cassandra warnings when checking health (see #420).

[1.0.0-beta.2] - 2021-03-24

Fixed

• [astarte_e2e] Fix alerting mechanism preventing "unknown" failures to be raised or linked.
• [astarte_appengine_api] Allow retrieving data from interfaces with parametric endpoint and object aggregation (see #480).
• [astarte_appengine_api] Encode binaryblob values with Base64 even if they are contained in an aggregate value.
• [astarte_trigger_engine] Encode binaryblob values with Base64 even if they are contained in an aggregate value.

Changed

• [astarte_e2e] Client disconnections are responsible for triggering a mail alert.
• Run tests against RabbitMQ 3.8.14 and ScyllaDB 4.4-rc.4 / Cassandra 3.11.10.
• Update dependencies to latest available versions (see mix.lock files).
• Update Elixir to 1.11.4 and OTP to 23.2.
• Rename device metadata to attributes. This requires a manual intervention on the database, see the Schema Changes documentation for additional information.

[1.0.0-beta.1] - 2021-02-16

Fixed

• [astarte_appengine_api] Fix regression that made it impossible to use Astarte Channels.
• [astarte_appengine_api] Fix bug that prevented data publishing in object aggregated interfaces.
• [astarte_appengine_api] Fix regression that prevented properties to be set before the first connection of a device.
• [astarte_realm_management] Fix a bug that prevented AMQP triggers to be correctly installed.
• [astarte_data_updater_plant] Mark device as offline and send device_disconnected event when forcing a device disconnection.
• [astarte_data_updater_plant] Fix bug that blocked queues when trying to disconnect an already disconnected device.

Added

• [astarte_housekeeping] Allow deleting a realm. The feature can be enabled with an environment variable (defaults to disabled).
• [astarte_data_updater_plant] Declare custom exchanges when an AMQP trigger is loaded.

Changed

• [astarte_housekeeping_api] Remove format check on Cassandra datacenter name when a realm is created, the datacenter is just verified against the one present in the database.
• [housekeeping] Increase the delay between connection attempts to 1000 ms, for an overall number of 60 attempts.
• [data_updater_plant] Default the total queue count to 128, de facto exploiting multiqueue support.
• [data_updater_plant] Default the queue range end to 127.
• Update Phoenix to version 1.5.
• Rework metrics to reduce the clutter while monitoring astarte services.
• [realm_management] Allow updating doc, description and explicit_timestamp within mappings when bumping an interface minor.
• Remove postgresql dependency in docker-compose, make CFSSL stateless.
• Update Operator's documentation for install/upgrade/uninstall procedures.

[1.0.0-alpha.1] - 2020-06-19

Fixed

• Make sure devices are eventually marked as disconnected even if they disconnect while VerneMQ is temporarily down (see #305).

Changed

• [appengine_api] Always return an object when GETting on object aggregated interfaces.
• Replace Conform and Distillery with Elixir native releases.
• Remove the ASTARTE_ prefix from all env variables.
• [realm_management_api] Triggers http actions are now validated.
• [realm_management_api] It is now possible to omit the device_id in a device_trigger. This is equivalent to passing * as device_id. The old behaviour is still supported.

Added

• [appengine_api] Add metadata to device
• [trigger_engine] Allow configuring preferred http method (such as PUT or GET) (see #128).
• [trigger_egnine] Add optional support to custom http headers, such as Authorization: Bearer ... (see #129).
• [data_updater_plant] Handle device heartbeat sent by VerneMQ plugin.
• [data_updater_plant] Deactivate Data Updaters when they don't receive messages for some time, freeing up resources.
• [appengine_api] Support SSL connections to RabbitMQ.
• [data_updater_plant] Support SSL connections to RabbitMQ.
• [trigger_engine] Support SSL connections to RabbitMQ.
• Default max certificate chain length to 10.
• AMQP trigger actions (publish to custom exchanges) as an alternative to http triggers actions.
• Ensure data pushed towards the device is correctly delivered when using QoS > 0.
• [realm_management_api] Allow installing device-specific and group-specific triggers. To do so, pass the device_id or group_name key inside the simple_trigger.
• [data_updater_plant] Add support for device-specific and group-specific triggers.
• Add support for device error triggers.

Removed

• [appengine_api] Remove deprecated not versioned socket route.

[0.11.5] - Unreleased

Fixed

• [realm_management] Avoid deleting all interfaces sharing the same name by mistake, only the v0 interface can be deleted.
• [data_updater_plant] Use a reasonable backoff time (at most around 5 minutes) when publishing to RabbitMQ.

[0.11.4] - 2021-01-26

Fixed

• Avoid creating an housekeeping_public.pem directory if docker-compose up doesn't find the housekeeping keypair.
• [trigger_engine] Correctly handle triggers on binaryblob interfaces, serializing value with base64 like appengine does.
• [data_updater_plant] Consider database_retention_ttl when inserting data on device owned aggregate interfaces.
• [realm_management] Do not allow value_change, value_change_applied and path_removed triggers on datastreams.
• [realm_management] Do not allow /* as match path when using value_change and value_change_applied. (workaround to astarte-platform#513).
• [trigger_engine] Update certifi to 2.5.3 (includes 2020-11-13 mkcert.org full CA bundle).

[0.11.3] - 2020-09-24

[0.11.2] - 2020-08-14

Added

• [trigger_engine] Add ignore_ssl_errors key in trigger actions, allowing to ignore SSL actions when delivering an HTTP trigger action.
• [trigger_engine] Update certifi to 2.5.2
• Update Elixir to 1.8.2

Changed

• [appengine_api] Remove topic from channel metrics.

[0.11.1] - 2020-05-18

Added

• [data_updater_plant] Add DATA_UPDATER_PLANT_AMQP_DATA_QUEUE_TOTAL_COUNT environment variable, this must be equal to the total number of queues in the Astarte instance.
• [trigger_engine] Add TRIGGER_ENGINE_AMQP_PREFETCH_COUNT environment variable to set the prefetech count of AMQPEventsConsumer, avoiding excessive memory usage.

Fixed

• Wait for schema_version agreement before applying any schema change (such as creating tables or a new realm). (see #312.
• [appengine_api] Fix the metric counting discarded channel events, it was not correctly increased.
• [data_update_plant] Validate UTF8 strings coming from the broker (i.e. interface and path) to avoid passing invalid strings to the database.
• [data_updater_plant] Fix a bug that was sometimes stalling a data updater queue process (see #375.

[0.11.0] - 2020-04-13

Fixed

• [appengine_api] Handle server owned datetime values correctly
• [housekeeping] Fix a bug preventing the public key of newly created realms to be correctly inserted to the realm (see #294).
• [data_updater_plant] Fix a bug that was preventing volatile triggers (specifically, the ones targeting the * interface) to be loaded immediately.

[0.11.0-rc.1] - 2020-03-26

Fixed

• [data_updater_plant] Discard unexpected object aggregated values on individual interfaces.
• [trigger_engine] 500 was not included in the range of HTTP server errors, causing a crash.

[0.11.0-rc.0] - 2020-02-26

Added

• [pairing_api] Add health endpoint.
• [realm_management_api] Add health endpoint.
• [housekeeping] Add Prometheus instrumenters and exporters.
• [trigger_engine] Add health endpoint.
• [housekeeping] Add health endpoint.
• [realm_management] Add health endpoint.
• [pairing] Add health endpoint.
• [data_updater_plant] Add health endpoint.
• [data_updater_plant] Export specific metrics with telemetry.
• [trigger_engine] Export specific metrics with telemetry.
• [appengine_api] Export specific metrics with telemetry.

Changed

• [realm_management] Correctly handle parametric endpoints regardless of the ordering, so that overlapping endpoints are always refused.
• [all] Make Elixir logger handle OTP requests: print stack traces only when needed.
• [appengine_api] Handle aggregated server owned interfaces.
• [appengine-api] Handle TTL for server owned interfaces.

[0.11.0-beta.2] - 2020-01-24

Added

• [pairing] Add Prometheus instrumenters and exporters.
• [realm_management] Add Prometheus instrumenters and exporters.
• [housekeeping_api] Add pretty_log.
• [trigger_engine] Add pretty_log.
• [pairing] Add pretty_log.
• [pairing_api] Add Prometheus instrumenters and exporters
• [realm_management_api] Add Prometheus instrumenters and exporters
• [housekeeping_api] Add Prometheus instrumenters and exporters
• Add standard interfaces for generic sensors.
• [trigger_engine] Add Prometheus instrumenters and exporters
• [pairing] Expose registration count and get_credentials count metrics.

Changed

• [realm_management] Handle hyphens in interface_name. (#96)
• [realm_management] Restrict the use of * as interface_name only to incoming_data data triggers.

Fixed

• [data_updater_plant] Load incoming_data triggers targeting any_interface. (#139)
• [housekeeping] Remove extra column in realm migration, preventing the correct upgrade to 0.11.
• [appengine_api] Fix crash that was happening when Channels received an event with an empty BSON as value (e.g. an IncomingDataEvent generated by an unset property).

[0.11.0-beta.1] - 2019-12-26

Added

• Add astarte_import tool, which allows users to import devices and data using XML files.
• [appegnine_api] Add new /v1/socket route for Astarte Channels. The /socket route is deprecated and will be removed in a future release.
• [appengine_api] Add groups support, allowing to group devices and access them inside a group hierarchy.
• [appengine_api] Add Prometheus metrics.
• [appengine_api] Show interface stats (exchanged messages and bytes) in device introspection.
• [appengine_api] Add previous_interfaces field to device details.
• [appengine_api] Allow installing group triggers in Astarte Channels.
• [data_updater_plant] Add support to multiple queues with consistent hashing
• [data_updater_plant] Save exchanged bytes and messages for all interfaces.
• [housekeeping] Add groups related columns and tables (schema has been changed).
• [housekeeping] Add interface stats related columns (schema has been changed).
• [housekeeping] Add database retention ttl and policy related columns (schema has been changed).
• Allow specifying initial introspection when registering a device.
• [realm_management] Trigger validation, checks that the interface is existing and performs validation on object aggregation triggers.

Changed

• Use separate docker images with docker-compose
• Use Scylla instead of Cassandra with docker-compose
• Authorization regular expressions must not have delimiters: they are implicit.
• [appengine_api] Change logs format to logfmt.
• [data_updater_plant] Changed logs format to logfmt.
• [realm_management] Changed logs format to logfmt.
• [realm_management_api] Changed logs format to logfmt.
• [housekeeping] Change database driver, start using Xandra.
• [housekeeping_api] Move health check API from /v1/health to /health to be consistent with all Astarte components.

[0.10.2] - 2019-12-09

Added

• [appengine_api] Add timestamp field to channel events.
• Add device unregister API, allowing to reset the registration of a device.
• [trigger_engine] Trigger timestamp is now extracted from SimpleEvent and not generated. This means that all triggers generated from the same event will have the same timestamp.

Fixed

• [appengine_api] Fix invalid dates handling, they should not cause an internal server error.
• [appengine_api] Gracefully handle existing aliases instead of returning an internal server error.
• [appengine_api] Fix querying object aggregated interface with explicit timestamp, use value_timestamp to avoid an internal server error.
• [appengine_api] Device details now show false in the connected field for never connected devices (null was returned before)
• [appengine_api] Handle out-of-band RPC errors gracefully instead of crashing.
• [data_updater_plant] Do not accept invalid paths that have consecutive slashes.
• [data_updater_plant] Do not accept invalid paths in object aggregated interfaces.
• [data_updater_plant] Do not delete all congruent triggers when deleting a volatile trigger.
• [data_updater_plant] Load volatile device triggers as soon as they're installed.
• [realm_management_api] Handle trigger not found reply from RPC, return 404 instead of 500.

Changed

• Use the timestamp sent by VerneMQ (or explicit timestamp if available) to populate SimpleEvent timestamp.
• [data_updater_plant] Update suggested RabbitMQ version to 3.7.15, older versions can be still used.

[0.10.1] - 2019-10-02

Added

• Support both SimpleStrategy and NetworkTopologyStrategy replications when creating a realm.
• Add sanity checks on the replication factor during realm creation.

Fixed

• Auth was refusing any POST method, a workaround has been added, however this will not work with regex.
• Fix reversed order when sending binaryblobarray and datetimearray.
• [data_updater_plant] Fix a bug that was causing a crash-loop in some corner cases when a message was sent on an outdated interface.
• [data_updater_plant] Send consumer properties correctly when handling emptyCache control message.
• Use updated interface validation: object aggregated properties interfaces are not valid.
• Use updated interface validation: server owned object aggregated interfaces are not yet supported, hence not valid.
• [realm_management_api] Trying to create a trigger with an already taken name now fails gracefully with an error instead of crashing.
• [trigger_engine] Fix datetime type handling, now it is properly serialized.

[0.10.0] - 2019-04-16

[0.10.0-rc.0] - 2019-04-03

Added

• [data_updater_plant] Add missing support to incoming object aggregated data with explicit_timestamp.

[0.10.0-beta.3] - 2018-12-19

Added

• [appengine_api] Binary blobs and date time values handling when PUTing and POSTing on a server owned interface.

Fixed

• docker-compose: Ensure CFSSL persists the CA when no external CA is provided.
• [data_updater_plant] Correctly handle Bson.UTC and Bson.Bin incoming data.
• [data_updater_plant] Fix crash when an interface that has been previously removed from the device introspection expires from cache.
• [data_updater_plant] Undecodable BSON payloads handling (handle Bson.Decoder.Error struct).
• [data_updater_plant] Discard invalid introspection payloads instead of crashing the data updater process.
• [realm_management_api] Correctly serialize triggers on the special "*" interface and device.

[0.10.0-beta.2] - 2018-10-19

Added

• Automatically add begin and end delimiters to authorization regular expressions.
• [appengine_api] Value type and size validation.
• [appengine_api] Option to enable HTTP compression.
• [data_updater_plant] Allow to expire old data using Cassandra TTL.
• [data_updater_plant] Publish set properties list to /control/consumer/properties.

Fixed

• [appengine_api] Path was added twice in authorization path, resulting in failures in authorization.
• [appengine_api] POST to a datastream endpoint doesn't crash anymore.
• [data_updater_plant] Validate all incoming values before performing any further computation on them, to avoid crash loops.
• [data_updater_plant] Fix a bug preventing data triggers to be correctly loaded
• [realm_management] Interface update, it was applying a broken update.
• [realm_management_api] Do not reply "Internal Server Error" when trying to delete a non existing interface.

Changed

• [appengine_api] "data" key is used instead of "value" when PUT/POSTing a value to an interface.
• [appengine_api] APPENGINE_MAX_RESULTS_LIMIT env var was renamed to APPENGINE_API_MAX_RESULTS_LIMIT.

[0.10.0-beta.1] - 2018-08-10

Added

• First Astarte release.