0.2.4 ✔️

Summary

This release enhances the flow engine with additional statistics about the amount of data being processed by and within flows. Event nodes have also been improved with a custom retention period feature.

Tip

To see the new flow statistics in action, run any flow via the flow editor.

Breaking changes

None.

Fixes

• 🐞 fix that the HTTP node could panic when the JS code for response post-processing created invalid JSON output data

Features

• ⚡ the flow engine now tracks byte-size statistics of input and output data
• ⚡ ui: the flow editor now displays total node error counts when testing flows setting
• ⚡💠 the event node now supports a custom, event-specific retention period between 15m and 365 days. If configured, it overrides the server’s event retention

Context

Previously, the server’s SIRVEO_RETENTION_PERIOD_EVENTS setting would apply to all events. It is now possible to specify custom retention periods for specific events generated with the event node, and on a per-event basis.

Custom events without a specified retention period are treated as per the server’s SIRVEO_RETENTION_PERIOD_EVENTS setting.

Note that when a custom event was created with a retention period, the server will delete the custom event after that period, even if the SIRVEO_RETENTION_PERIOD_EVENTS is disabled.

Enhancements

• 🔼 ui: revamp the flow editor’s display of execution stats, and add new statistics
• 🔼 ui: the dashboard’s what’s new section fetches content online, with a 30-minute refresh period
• 🔼 ui: the flow editor’s display of execution statistics has been revamped, and includes new byte size metrics

Internal

• ⚙️ flow engine: track byte size of flow input
• ⚙️ flow engine: track byte size of flow outputs
• ⚙️ flow engine: track byte size of node inputs
• ⚙️ flow engine: track byte size of node outputs
• ⚙️ flow engine: track total byte size of internal data traversal
• ⚙️ flow engine: track node error channel count
• ⚙️ server: database cleanup task runs on startup, irrespective of whether retention periods are configured

Context

Previously, database cleanup would only be performed at server startup if either SIRVEO_RETENTION_PERIOD_SCHEDULES or SIRVEO_RETENTION_PERIOD_EVENTS was configured.

The server now always performs database cleanup at startup, and then periodically every 15m15s thereafter.

API

• api: /api/admin/flow/run/* provides complete execution details when flow runtime errors occur

Database

• ⚙️ database: the event table has a new cleanup_at column to support the per-event retention period feature