Milvus is committed to providing the best vector databases to power AI applications and vector similarity search. However, the team is continuously working to bring in more features and the best utilities to enhance user experience. This page lists out some known limitations that the users may encounter when using Milvus.
Resource | Limit |
Collection | 255 characters |
Field | 255 characters |
Index | 255 characters |
Partition | 255 characters |
The name of a resource can contain numbers, letters, and underscores (_). A resource name must start with a letter or an underscore (_).
Resource | Limit |
Collection | 65,536 |
Connection / proxy | 65,536 |
Resource | Limit |
Partition | 4,096 |
Shard | 16 |
Field | 64 |
Index | 1 |
Entity | unlimited |
Data type | Limit |
VARCHAR | 65,535 |
Property | Limit |
Dimension | 32,768 |
Operation | Limit |
Insert | 64 MB |
Search | 64 MB |
Query | 64 MB |
In current release, data to be load must be under 90% of the total memory resources of all query nodes to reserve memory resources for execution engine.
Vectors | Limit |
topk (number of the most similar result to return) | 16,384 |
nq (number of the search requests) | 16,384 |