Upgrade Milvus to 2.5.16+ for FastGPT

Upgrade Milvus to 2.5.16+ Overview

This documented upgrade procedure preserves all existing Milvus data volumes and the legacy modeldata collection, which acts as the primary vector source for any subsequent migration workflows tied to FastGPT. No pre-existing dataset stored in Milvus will be modified during the initial upgrade steps, ensuring continuity of existing vector search functionality prior to validation.

Step-by-Step Upgrade Steps

This section outlines the minimal required changes to upgrade your Milvus deployment for compatibility with FastGPT:

1. Access your existing docker-compose configuration file that defines the Milvus service.
2. Update the Milvus container image tag to the mandatory version:

# Milvus service in docker-compose
image: milvusdb/milvus:v2.5.16

1. Restart the Milvus service to pull the new image and apply the configuration changes.

Post-Upgrade Validation

Following the service restart, it is critical to validate the integrity of the modeldata collection, as this dataset is required for ongoing migration operations:

1. Confirm that the modeldata collection is present in your Milvus instance.
2. Verify that the collection contains non-empty vector data, as an empty or missing collection will break subsequent migration workflows.

If the modeldata collection is either missing or entirely empty, stop all migration operations immediately and restore your Milvus data from a pre-upgrade backup to recover the required vector source data.

Source: FastGPT official documentation and source