后端发布版本

关于本文

本文列出了所有 RainMaker 后端发布版本。如需获取支持，请联系 esp-rainmaker-support@espressif.com。

2025 年后端发布版本

3.5.0 (2025.10.1)

If you are upgrading from any version older than 3.0.0, please note that there would be a downtime during the upgrade

Refer RainMaker version 3.0.0 release notes for downtime considerations.

For SAR deployment method, please upgrade the Node Registration module to v28 after upgrading the backend.

Please upgrade the RainMaker Dashboard to 2.6.0 after upgrading the backend.

New Features:
- OTA:
  - OTA Job: Allow force-pushed retries for OTA jobs so that failed or aborted jobs can be resumed by nodes. Refer to the Re-trigger OTA Job API for details.
  - OTA Status: Added support for nodes to send a timestamp with status updates for accuracy and better deduplication. Also added the timestamp source: NODE/CLOUD. Refer to the OTA Status Topic for details.
- Admin notifications: The Superadmin can now add additional email addresses to receive node registration status updates. Refer to Swagger for details.
- User Data Access: Users can now export and download their data stored in RainMaker, as required by certain privacy regulations. Admin users can also do the same on behalf of end users. Refer to the Data Request API for details. Currently supports basic user details.
- Location-based Automation:
  - Added parameter max_weather_triggers_requests_per_minute for configuring location-based automation triggers. This limits the number of weather API requests per minute, helping control external API usage and costs. Refer to Swagger for details.
  - Added a user API to fetch admin-configured units for location-based weather automation triggers (Standard, Metric, Imperial).
- All Lambda functions running on Python 3.9 have been upgraded to Python 3.13.
Other changes
- Node registration: Node IDs are now generated locally by default. Please pull RainMaker Admin CLI v1.2.3.
Bug Fixes:
- OTA:
  - Secure signed OTA images failed to download over MQTT.
  - Secure signed OTA jobs failed to deliver when targeting more than 5,000 nodes.
  - Force-pushed OTA jobs were not pushed to all nodes if a node completed the job while it was being created. For non force-push jobs, some nodes lacked the 'triggered' status. Upon OTA fetch, the nodes could receive the OTA job update.
- Custom SMS template: Forgot Password and Sign-up flows were using the same SMS message template.
- Node disconnection events were processed after 20+ seconds delay.
- User node groups:
  - If a node was added to a group during user-node mapping and then added to another group, deleting the original mapping group removed the user-node mapping, but the node remained reachable through the second group.
- User node group sharing:
  - Group sharing allowed only one pending request per group at a time, the latest one overriding the previous ones, even if the requests were for different users.
  - After a user added their own nodes to a shared group, unsharing that group for the same user resulted in an error.
  - Group sharing requests were not deleted when a user was deleted.
- Node Registration:
  - Registration with --update_nodes nodes did not add nodes to the created group.
  - The number of nodes generated was intermittently less than the requested count.
  - Node certificate registration failed when registering more than 50K nodes.