Changelog #49 - Faster logs, Improved initial deployment lifecycle, Fixed public data transfer metric, and more

Announcements
, , , ,
1

Hello, and welcome to this week’s changelog update! Let’s jump into what’s new:

  1. Faster build and runtime logs
    Build and runtime logs are now displayed faster to track what is happening and ease troubleshooting your Services. Previously it could take a few seconds before appearing.

    Faster logs
    Faster logs2304×1296 248 KB

    Runtime logs are retained for 72 hours. To go over this limitation, you can stream your logs to specialized services using Koyeb’s logs exporter.

  1. Improved initial deployment lifecycle
    Previously, at runtime, an Instance event “Instance is unhealthy” was displayed as part of the initial deployment process. This behavior was confusing and is now fixed.

  2. Fixed public data transfer metric
    Thanks to Raging_Code, we identified an issue when rendering the public data transfer metric in the control panel. This is now resolved.

  3. Build a Retrieval-Augmented Generation Chatbot using pgvector
    Rishi recently published a guide to build a retrieval-augmented generation chatbot using pgvector. It is a good occasion to leverage our serverless Postgres databases. You can even deploy it with one click!