How to use this image

All examples in this guide use the public image. If you've mirrored the repository for your own use (for example, to your Docker Hub namespace), update your commands to reference the mirrored image instead of the public one.

For example:

• Public image: dhi.io/milvus:<tag>
• Mirrored image: <your-namespace>/dhi-milvus:<tag>

For the examples, you must first use docker login dhi.io to authenticate to the registry to pull the images.

What's included in this milvus image

Milvus⁠ is a high-performance, cloud-native vector database built for large-scale approximate nearest neighbor (ANN) search, commonly used to power retrieval-augmented generation (RAG), semantic search, and recommendation systems.

The image ships a single milvus server binary that runs any Milvus role from one entrypoint — for example, milvus run standalone for a self-contained instance, or milvus run querynode, milvus run datanode, and similar roles when deploying a distributed cluster. Matching upstream, the image sets ENTRYPOINT ["/usr/bin/tini", "--"] with no default CMD, so you always supply the role you want milvus to run.

Start a milvus image

The milvus CLI has no --version flag; running it without arguments prints the available commands and roles, and the server version is printed in the startup banner of milvus run:

$ docker run --rm dhi.io/milvus:<tag> milvus

standalone is the only role that requires no external dependencies: it bundles an embedded etcd, local object storage, and an embedded message queue in a single process, and is the fastest way to get a working instance. The embedded etcd reads its configuration from the file referenced by ETCD_CONFIG_PATH, which is not shipped in the image (upstream's standalone recipe generates and mounts it the same way), so create and mount it:

$ cat > embedEtcd.yaml <<'EOF'
listen-client-urls: http://127.0.0.1:2379
advertise-client-urls: http://127.0.0.1:2379
quota-backend-bytes: 4294967296
auto-compaction-mode: revision
auto-compaction-retention: '1000'
EOF

$ docker run -d --name milvus-standalone \
  -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 \
  -e DEPLOY_MODE=STANDALONE \
  -v "$(pwd)/embedEtcd.yaml:/milvus/configs/embedEtcd.yaml:ro" \
  -p 127.0.0.1:19530:19530 \
  -p 127.0.0.1:9091:9091 \
  dhi.io/milvus:<tag> \
  milvus run standalone

Milvus ships with authentication disabled (common.security.authorizationEnabled is false), so the gRPC API on port 19530 and the metrics/health endpoint on port 9091 accept any caller. The examples publish both ports on the loopback interface only; publish them on other interfaces once you have enabled authentication and TLS through user.yaml (see Custom configuration⁠ and the upstream authentication⁠ and TLS⁠ guides).

Milvus is ready once the gRPC API accepts connections and the health endpoint reports healthy. The runtime image has no shell and no curl, so check the health endpoint from the host, or from a sidecar sharing the container's network namespace:

$ docker run --rm --network container:milvus-standalone dhi.io/busybox:1 wget -qO- http://localhost:9091/healthz

Common milvus use cases

Standalone with persistent storage

The command above keeps no state across container restarts. To persist data, mount a host directory over /var/lib/milvus — the directory Milvus uses for its embedded etcd data, local object storage, and rocksmq. Because this image runs as the nonroot user 65532 (unlike the upstream image, which runs as root), the host directory must be owned by, or writable by, that user before you start the container:

$ mkdir -p volumes/milvus
$ sudo chown 65532:65532 volumes/milvus

The embedded etcd also needs a config file at the path referenced by ETCD_CONFIG_PATH:

$ cat > embedEtcd.yaml <<'EOF'
listen-client-urls: http://127.0.0.1:2379
advertise-client-urls: http://127.0.0.1:2379
quota-backend-bytes: 4294967296
auto-compaction-mode: revision
auto-compaction-retention: '1000'
EOF

$ docker run -d --name milvus-standalone \
  -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 \
  -e DEPLOY_MODE=STANDALONE \
  -v "$(pwd)/volumes/milvus:/var/lib/milvus" \
  -v "$(pwd)/embedEtcd.yaml:/milvus/configs/embedEtcd.yaml:ro" \
  -p 127.0.0.1:19530:19530 \
  -p 127.0.0.1:9091:9091 \
  dhi.io/milvus:<tag> \
  milvus run standalone

Custom configuration

Milvus resolves its configuration directory as $CWD/configs unless the MILVUSCONF environment variable overrides it. Because the image's WORKDIR is /milvus and /milvus/configs is a symlink to the packaged configuration directory, mount a user.yaml override (Milvus merges it over the packaged milvus.yaml) at /milvus/configs/user.yaml. Keep the standalone environment from the sections above — the user.yaml mount is additive:

$ docker run -d --name milvus-standalone \
  -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 \
  -e DEPLOY_MODE=STANDALONE \
  -v "$(pwd)/embedEtcd.yaml:/milvus/configs/embedEtcd.yaml:ro" \
  -v "$(pwd)/user.yaml:/milvus/configs/user.yaml:ro" \
  -p 127.0.0.1:19530:19530 \
  -p 127.0.0.1:9091:9091 \
  dhi.io/milvus:<tag> \
  milvus run standalone

FIPS-validated cryptography

The fips and fips-dev variants ship a Milvus binary linked with the Go Cryptographic Module (GOFIPS140), which runs in FIPS 140-3 mode by default, and configure the system OpenSSL used by the C++ core with the image's FIPS-validated provider (openssl-provider-fips). The startup log reports Milvus FIPS in OpenSSL: enabled once the provider is active. The line before it, Milvus FIPS in Go: BoringCrypto false, is expected: the Go side's FIPS mode comes from the Go Cryptographic Module, not from upstream's BoringCrypto build. No extra configuration is needed beyond selecting a -fips tag. Two Rust components compiled into the C++ core carry their own TLS and hashing crates, which are outside both validated modules: the full-text index binding (rustls with ring, both lines) and, in Milvus 3.0, the storage v2 bridge that reads and writes Lance and Vortex files in object storage (rustls with aws-lc-rs, native-tls over the system OpenSSL, and RustCrypto digests). Object storage traffic that goes through that bridge negotiates TLS with rustls rather than with the validated OpenSSL provider.

Cluster deployments

The standalone role above is self-contained and suitable for development, testing, and many production workloads. Distributed cluster deployments additionally require an external etcd cluster, S3-compatible object storage (for example MinIO), and a Pulsar or Kafka message queue, typically coordinated with the Milvus Operator or the official Helm chart:

• Install Milvus in cluster mode⁠
• Milvus Helm charts⁠

When following the upstream Operator manifests or Helm chart, override the image reference to dhi.io/milvus:<tag> in place of the default milvusdb/milvus image.

Non-hardened images vs. Docker Hardened Images

Unlike the upstream milvusdb/milvus image, which runs as root, this image's runtime and fips variants run as the nonroot user 65532. If you bind-mount a host directory over /var/lib/milvus, make sure it's owned by or writable by 65532:65532 — the in-image directory already is, but host paths are not until you chown them (see Standalone with persistent storage⁠). For the same reason, embedded-etcd deployments must set ETCD_DATA_DIR to a writable location such as /var/lib/milvus/etcd: the upstream default (default.etcd, relative to the root-owned working directory /milvus) is not writable by the nonroot user.

The runtime image also has no shell and no curl, so upstream's own healthcheck (curl -f http://localhost:9091/healthz) doesn't work unmodified; see Start a milvus image⁠ for a working alternative.

Upstream's sample TLS key pairs under configs/cert are not shipped. When you enable TLS (common.security.tlsMode), mount your own certificates and point tls.serverPemPath, tls.serverKeyPath and tls.caPemPath at them in user.yaml.

Image variants

Docker Hardened Images come in different variants depending on their intended use. Image variants are identified by their tag.

• Runtime variants are designed to run your application in production. These images are intended to be used either directly or as the FROM image in the final stage of a multi-stage build. These images typically:

  • Run as a nonroot user
  • Do not include a shell or a package manager
  • Contain only the minimal set of libraries needed to run the app

• Build-time variants typically include dev in the tag name and are intended for use in the first stage of a multi-stage Dockerfile. These images typically:

  • Run as the root user
  • Include a shell and package manager
  • Are used to build or compile applications

• FIPS variants include fips in the variant name and tag. They come in both runtime and build-time variants. These variants use cryptographic modules that have been validated under FIPS 140, a U.S. government standard for secure cryptographic operations. For example, usage of MD5 fails in FIPS variants.

To view the image variants and get more information about them, select the Tags tab for this repository, and then select a tag.

Migrate to a Docker Hardened Image

To migrate your application to a Docker Hardened Image, you must update your Dockerfile. At minimum, you must update the base image in your existing Dockerfile to a Docker Hardened Image. This and a few other common changes are listed in the following table of migration notes.

The following steps outline the general migration process.

1. Find hardened images for your app.

   A hardened image may have several variants. Inspect the image tags and find the image variant that meets your needs.

2. Update the base image in your Dockerfile.

   Update the base image in your application's Dockerfile to the hardened image you found in the previous step. For framework images, this is typically going to be an image tagged as dev because it has the tools needed to install packages and dependencies.

3. For multi-stage Dockerfiles, update the runtime image in your Dockerfile.

   To ensure that your final image is as minimal as possible, you should use a multi-stage build. All stages in your Dockerfile should use a hardened image. While intermediary stages will typically use images tagged as dev, your final runtime stage should use a non-dev image variant.

4. Install additional packages

   Docker Hardened Images contain minimal packages in order to reduce the potential attack surface. You may need to install additional packages in your Dockerfile. Inspect the image variants to identify which packages are already installed.

   Only images tagged as dev typically have package managers. You should use a multi-stage Dockerfile to install the packages. Install the packages in the build stage that uses a dev image. Then, if needed, copy any necessary artifacts to the runtime stage that uses a non-dev image.

   For Alpine-based images, you can use apk to install packages. For Debian-based images, you can use apt-get to install packages.

Troubleshooting migration

The following are common issues that you may encounter during migration.

General debugging

The hardened images intended for runtime don't contain a shell nor any tools for debugging. The recommended method for debugging applications built with Docker Hardened Images is to use Docker Debug⁠ to attach to these containers. Docker Debug provides a shell, common debugging tools, and lets you install other tools in an ephemeral, writable layer that only exists during the debugging session.

Permissions

By default image variants intended for runtime, run as the nonroot user. Ensure that necessary files and directories are accessible to the nonroot user. You may need to copy files to different directories or change permissions so your application running as the nonroot user can access them.

Privileged ports

Non-dev hardened images run as a nonroot user by default. As a result, applications in these images can't bind to privileged ports (below 1024) when running in Kubernetes or in Docker Engine versions older than 20.10. To avoid issues, configure your application to listen on port 1025 or higher inside the container, even if you map it to a lower port on the host. For example, docker run -p 80:8080 my-image will work because the port inside the container is 8080, and docker run -p 80:81 my-image won't work because the port inside the container is 81.

No shell

By default, image variants intended for runtime don't contain a shell. Use dev images in build stages to run shell commands and then copy any necessary artifacts into the runtime stage. In addition, use Docker Debug to debug containers with no shell.

Entry point

Docker Hardened Images may have different entry points than images such as Docker Official Images. Use docker inspect to inspect entry points for Docker Hardened Images and update your Dockerfile if necessary.