Scale Milvus Standalone

Milvus Standalone is a single-machine server deployment. All components of Milvus Standalone are packed into a single Docker image, making deployment convenient. This topic describes how to scale a Milvus instance running in this mode.

Prerequsites

When deploying Milvus Standalone with Docker or Docker Compose, the deployment script (standalone_embed.sh) or configuration file (docker-compose.yml) creates several volumes and maps them to host directories to ensure data persistence.

To scale a Milvus instance deployed in this manner, you must stop and remove the existing container or container stack, redeploy Milvus Standalone with updated configuration settings, and reuse the persisted data on your host to launch a new instance.

The following table lists the volume mapping between the host and containers.

Before running the procedures in this guide, ensure your data persists in the above host paths.

Scale instances deployed using Docker

To scale a currently running Milvus instance, you must stop the instance, remove the container, and redeploy the instance with new settings and persisted data.

The specific procedure is as follows:

1. Run docker stats milvus-standalone to view the CPU and memory allocated to the Milvus instance. The output should be similar to the following:

   CONTAINER ID   NAME                CPU %     MEM USAGE / LIMIT     MEM %     NET I/O       BLOCK I/O         PIDS
   917da667f2ff   milvus-standalone   6.10%     171.8MiB / 3.886GiB   4.32%     1.57kB / 0B   1.01GB / 1.79MB   31
   

   In the command output, you can find the current resource usage of your Milvus instance.

2. Stop and remove the container.

   $ docker stop milvus-standalone
   $ docker rm milvus-standalone
   

3. Locate the standalone_embed.sh script file, find the docker run command, and add the resource limits.

   ...
       sudo docker run -d \
           --name milvus-standalone \
           --security-opt seccomp:unconfined \
           -e ETCD_USE_EMBED=true \
           -e ETCD_DATA_DIR=/var/lib/milvus/etcd \
           -e ETCD_CONFIG_PATH=/milvus/configs/embedEtcd.yaml \
           -e COMMON_STORAGETYPE=local \
           -v $(pwd)/volumes/milvus:/var/lib/milvus \
           -v $(pwd)/embedEtcd.yaml:/milvus/configs/embedEtcd.yaml \
           -v $(pwd)/user.yaml:/milvus/configs/user.yaml \
           -p 19530:19530 \
           -p 9091:9091 \
           -p 2379:2379 \
           --health-cmd="curl -f http://localhost:9091/healthz" \
           --health-interval=30s \
           --health-start-period=90s \
           --health-timeout=20s \
           --health-retries=3 \
           --memory="4g" \          # New memory limit
           --cpus="2.0" \           # New CPU limit
           milvusdb/milvus:v2.5.11 \
           milvus run standalone  1> /dev/null
   

4. Ensure the persisted data is in the same folder as the standalone_embed.sh script, and run the script as follows:

   sudo  bash standalone_embed.sh start
   

5. Run docker stats milvus-standalone to view the CPU and memory allocated to the Milvus instance after scaling. The output should be similar to the following:

   CONTAINER ID   NAME                CPU %     MEM USAGE / LIMIT   MEM %     NET I/O       BLOCK I/O        PIDS
   7aea450f87ce   milvus-standalone   7.52%     210.9MiB / 4GiB     5.15%     1.05kB / 0B   610kB / 8.19kB   29
   

Scale instances deployed using Docker Compose

To scale a currently running Milvus instance, you must stop the instance, remove the container stack, and redeploy the instance with new settings and persisted data.

The specific procedure is as follows:

1. Run docker stats milvus-standalone to view the CPU and memory allocated to the Milvus instance. The output should be similar to the following:

   CONTAINER ID   NAME                CPU %     MEM USAGE / LIMIT     MEM %     NET I/O       BLOCK I/O         PIDS
   917da667f2ff   milvus-standalone   6.10%     171.8MiB / 3.886GiB   4.32%     1.57kB / 0B   1.01GB / 1.79MB   31
   

   In the command output, you can find the current resource usage of your Milvus instance.

2. Stop and remove the container stack.

   $ docker compose down
   

3. Locate the docker-compose.yml configuration file, find the standalone section, and add the resource limits.

   ...
     standalone:
       container_name: milvus-standalone
       image: milvusdb/milvus:v2.5.8
       command: ["milvus", "run", "standalone"]
       deploy:
         resources:
           limits:
             cpus: "2"   # new cpu limits
             memory: 4G  # new memory limits
       security_opt:
       - seccomp:unconfined
       environment:
         ETCD_ENDPOINTS: etcd:2379
         MINIO_ADDRESS: minio:9000
       volumes:
         - ${DOCKER_VOLUME_DIRECTORY:-.}/volumes/milvus:/var/lib/milvus
       healthcheck:
         test: ["CMD", "curl", "-f", "http://localhost:9091/healthz"]
         interval: 30s
         start_period: 90s
         timeout: 20s
         retries: 3
       ports:
         - "19530:19530"
         - "9091:9091"
       depends_on:
         - "etcd"
         - "minio"
   

4. Ensure the persisted data is available, and run docker compose as follows:

   docker compose up -d
   

5. Run docker stats milvus-standalone to view the CPU and memory allocated to the Milvus instance after scaling. The output should be similar to the following:

   CONTAINER ID   NAME                CPU %     MEM USAGE / LIMIT   MEM %     NET I/O       BLOCK I/O        PIDS
   7aea450f87ce   milvus-standalone   7.52%     210.9MiB / 4GiB     5.15%     1.05kB / 0B   610kB / 8.19kB   29