This adds a Region configuration option to the device-profile, which lists the region configurations for the selected Region. If a region configuration is selected, then the device will only be able to work under the selected configuration. If no region configuration is selected, then the device will be able to operate under all available region configurations for the selected region.
This adds support for Java and Kotlin API SDK code generation. (#64)
descriptionconfiguration option per region configuration.
idwithin the region configuration (
namewill be used as fallback option).
- Make gateway state in UI consistent and make expected stats interval configurable. (#76)
- Add Python type information to Python API SDK code. (#68)
- Add back
UplinkRxInfomessage (the status will be reported as no CRC until the ChirpStack Gateway Bridge, ChirpStack Concentratord and / or ChirpStack MQTT Forwarder have been updated).
- Add back Class-B ping-slot parameters to the device-profile.
- Update Class-B ping-slot data-rate configuration in examples.
- Remove separate gateway topic config and move it into single
- Reset internally stored channels to default on
ADRACKRequplink to avoid out-of-sync channel configuration on device.
- Update internal dependencies.
- Fix hiding Delete device option if the user has no permissions to perform this action. (#71)
- Fix not recording device metrics if auto-detect of measurements is disabled. (#94)
- Fix Redis
key_prefixconfiguration. While this value could be configured, it was not applied to the generated keys.
- Fix header
z-indexissue in UI. This was causing the dropdowns to render partly behind the header.
- Do not wait for integrations to finish before sending downlink.
- Update JS API dependencies to latest versions.
- Replace relative paths in Rust API build to absolute. (#69)
- Fix setting the full frame-counter to the uplink frame after resolving the device-session (this can affect payload decryption).
This feature logs API requests to Redis Streams. This enables external services to monitor for example device create, update and deletes by reading from the Redis Streams. A code-example can be found here.
While the feature to log frames was already present, it was not possible
to only read uplink frames of devices that are unknown. This extends the frame
logging feature to also log uplinks for unknown devices, in which case DevEUI
0000000000000000 is used. A code-example for reading the frame log can be
A code-example to read event logs from the Redis Streams was added. It can be found here.
metadatafields in gateway messages consistent.
- Emit all fields for JSON integration messages, even if they are their default values. (#63)
- Fix Redis pipelined commands in case Redis Cluster is configured.
- Fix UI notifications z-index.
import-ttn-lorawan-devicessub-command has been renamed to
keep_alive_intervalfor MQTT configuration.
- Fix incorrect splitting of multiple URLs in HTTP integration. (#62)
- Fix missing ThingsBoard location and status telemetry.
- Send ThingsBoard telemetry (fPort, fCnt, ...) even in case there is no decoded payload.
- Fix LeafletJS controls floating over header (UI).
- Fix coding-rate for LoRaWAN 2.4GHz. (#51)
- Fix not removing of queue-item after it was sent, if payload did not fit RX2.
- Add code example for reading frame-logs from Redis Streams.
- Add missing metadata logging and add code example for reading metadata from Redis Streams.
- Make it possible to enable / disable auto-detection of metrics in device-profile. (#42)
- Add Redis config examples for username and password configuration. (#54)
- Add metadata tab to gateway configuration in UI.
ca_pathfor MQTT integration. (#47)
- Fix Cannot serialize NaN as google.protobuf.Value.number_value error.
- Fix logout URL in case OIDC is enabled.
- Fix closing and detecting if eventlog channel is closed.
- Fix metrics interval calculation on daily aggregate in case of DST to non-DST timezone change.
- A new menu option Regions was added to the web-interface, exposing per region information.
- Internal dependencies were updated.
- Fix Wi-Fi geolocation issue in LoRa Cloud integration.
- Fix missing per data-rate stats in gateway dashboard.
- Fix terminating stream loop (and releasing of Redis connection) on client disconnect. (#40)
- Fix rendering
client_cert_lifetimerendering in configuration template (mqtt integration).
After many months of development and testing, we are really excited to share ChirpStack v4.
The aim of ChirpStack v4 is to make it significantly easier to setup and use ChirpStack, compared to the previous version. One of the major changes that you will notice is that the ChirpStack Network Server and ChirpStack Application Server have been merged into a single component. Over the years we have seen many issues reported on the forum and GitHub, related to setting up and connecting both services. ChirpStack v4 also provides multi-region support out-of-the-box, including region configuration. No longer it is needed to define your own configuration file or setup multiple ChirpStack Network Server instances to serve multiple regions simultaneously.
A big thank you to the ChirpStack community for supporting and contributing to the ChirpStack project! Please find below a breakdown of all the new features and changes that v4 brings.
ChirpStack v4 adds multi-region support, removing the need to setup multiple ChirpStack Network Server instances. Configuration files are included for the common regions (as defined by the LoRa Alliance), which should help getting started with ChirpStack.
Each enabled region has its own gateway backend, making it possible to use one
or multiple MQTT brokers for the different gateway pools. In case a single
MQTT broker is used, each backend must be configured with its own MQTT topic prefix
ChirpStack v4 also supports multiple configurations of the same region, e.g. to configure a US915 for 8 channels and to configure a US915 band for 16 channels.
ChirpStack v4 adds support for importing the TTN LoRaWAN Devices repository as device-profile templates, including codec functions if these are defined in this repository.
In the device-profile template and / or device-profile, it is possible to define the measurements that are exposed by the device in the decoded payload. Once defined, ChirpStack will automatically aggregate and store this data. These metrics can be viewed in the web-interface on the device dashboard.
Instead of using a single configuration file (e.g.
ChirpStack makes use of a configuration directory such that the configuration
can be split in multiple files. By default you will find a single
configuration file, and many
region_...toml configuration files, split
ChirpStack v4 removes the TOML hierarchy to environment variable mapping.
Instead it allows you to define the variables like
will get automatically substituted when an environment variable is found with
The REST interface that was present in ChirpStack Application Server v3 has
been removed, in favor of the gRPC API interface (please see the
of the repository for the API definitions). However, a gRPC to REST interface
bridge component will be provided as a separate service. Please note that
in v3, this bridge component was embedded and REST interface calls were
internally translated to gRPC calls. Therefore, gRPC was always recommended
interface to use.
ChirpStack v4 is fully compatibility with the latest version of ChirpStack
Gateway Bridge v3. This should help migrating from v3 to v4. Please note that
the ChirpStack Gateway Bridge must be configured with the
ChirpStack v4 contains a rewrite of the ChirpStack Application Server v3 UI. The new UI aims to be more user-friendly. Under the hood the API interface has been ported to gRPC-web and all code has been ported to Typescript.
The implementation of Passive Roaming has been improved, adding support for
/fns server endpoint suffixes. The usage of this
suffix is not specified in the Backend Interfaces specification, but is
required by some other network-server implementations.
All identifiers that are exposed have been changed to UUID. Previously most
identifiers (e.g. users, applications...) were incremental integers. In case
ChirpStack is setup as multi-tenant instance, this could expose some information
about the number of clients on the network. The migration script (see below)
will migrate these integers by converting these as strings, prefixed with
zeros in the UUID format. E.g. ID
123 would be converted to the UUID string
All binary identifiers have been changed to string type in the API. While
binary fields are more efficient, these were confusing when encoded as JSON
as the Protobuf to JSON mapping uses base64 encoding for binary fields.
For example, a Gateway ID
0102030405060708 was encoded as
While the structure of API messages is roughly the same as the ChirpStack Application Server API interface, some small changes have been made.
The integration event messages have been restructured for better consistency.
Each event message has a
deviceInfo field which holds device-related
information (tenant id & name, application id & name, device-profile id &
name, device EUI & name and tags).
ChirpStack v4 will make it a lot easier to make customizations, especially when API changes are involved, as API definitions are no longer separated from the code. In v3 these definitions were moved to an external repository to avoid cross dependencies.
For ChirpStack v4, it was decided to use Rust rather than Go. This was not an easy choice and the arguments for this decision are debatable. However, as most code was touched during the ChirpStack Application Server and ChirpStack Network Server merge, it was the only moment to re-consider this. The Rust memory management prevents many memory related pitfalls and helps catching bugs at compile time rather than runtime.
The recommended way to migrate from v3 to ChirpStack v4 is to create a new PostgreSQL and Redis database and to use the ChirpStack v3 to v4 migration script. This script will copy all the data from the "old" into the "new" database. While the script does not make any modifications to the old database, it is always recommended to make a backup first.