Run Milvus in Kubernetes with Helm

This page illustrates how to start a Milvus instance in Kubernetes using Milvus Helm charts.

Overview

Helm uses a packaging format called charts. A chart is a collection of files that describe a related set of Kubernetes resources. Milvus provides a set of charts to help you deploy Milvus dependencies and components.

Prerequisites

• Install Helm CLI.

• Create a K8s cluster.

• Install a StorageClass. You can check the installed StorageClass as follows.

  $ kubectl get sc
  
  NAME                  PROVISIONER                  RECLAIMPOLICY    VOLUMEBIINDINGMODE    ALLOWVOLUMEEXPANSION     AGE
  standard (default)    k8s.io/minikube-hostpath     Delete           Immediate             false 
  

• Check the hardware and software requirements before installation.

• Before installing Milvus, it is recommended to use the Milvus Sizing Tool to estimate the hardware requirements based on your data size. This helps ensure optimal performance and resource allocation for your Milvus installation.

Install Milvus Helm Chart

Before installing Milvus Helm Charts, you need to add Milvus Helm repository.

$ helm repo add milvus https://zilliztech.github.io/milvus-helm/

helm repo add zilliztech https://zilliztech.github.io/milvus-helm
helm repo update
# upgrade existing helm release
helm upgrade my-release zilliztech/milvus

Then fetch Milvus charts from the repository as follows:

$ helm repo update

You can always run this command to fetch the latest Milvus Helm charts.

Online install

1. Deploy a Milvus cluster

Once you have installed the Helm chart, you can start Milvus on Kubernetes. This section will guide you through the steps to starting Milvus.

$ helm install my-release milvus/milvus

In the above command, my-release is the release name, and milvus/milvus is the locally installed chart repository. To use a different name, replace my-release with the one you see fit.

The command above deploys a Milvus cluster with its components and dependencies using default configurations. To customize these settings, we recommend you use the Milvus Sizing Tool to adjust the configurations based on your actual data size and then download the corresponding YAML file. To learn more about configuration parameters, refer to Milvus System Configurations Checklist.

2. Check Milvus cluster status

Run the following command to check the status of all pods in your Milvus cluster.

$ kubectl get pods

Once all pods are running, the output of the above command should be similar to the following:

NAME                                             READY  STATUS   RESTARTS  AGE
my-release-etcd-0                                1/1    Running   0        3m23s
my-release-etcd-1                                1/1    Running   0        3m23s
my-release-etcd-2                                1/1    Running   0        3m23s
my-release-milvus-datanode-68cb87dcbd-4khpm      1/1    Running   0        3m23s
my-release-milvus-indexnode-5c5f7b5bd9-l8hjg     1/1    Running   0        3m24s
my-release-milvus-mixcoord-7fb9488465-dmbbj      1/1    Running   0        3m23s
my-release-milvus-proxy-6bd7f5587-ds2xv          1/1    Running   0        3m24s
my-release-milvus-querynode-5cd8fff495-k6gtg     1/1    Running   0        3m24s
my-release-minio-0                               1/1    Running   0        3m23s
my-release-minio-1                               1/1    Running   0        3m23s
my-release-minio-2                               1/1    Running   0        3m23s
my-release-minio-3                               1/1    Running   0        3m23s
my-release-pulsar-autorecovery-86f5dbdf77-lchpc  1/1    Running   0        3m24s
my-release-pulsar-bookkeeper-0                   1/1    Running   0        3m23s
my-release-pulsar-bookkeeper-1                   1/1    Running   0        98s
my-release-pulsar-broker-556ff89d4c-2m29m        1/1    Running   0        3m23s
my-release-pulsar-proxy-6fbd75db75-nhg4v         1/1    Running   0        3m23s
my-release-pulsar-zookeeper-0                    1/1    Running   0        3m23s
my-release-pulsar-zookeeper-metadata-98zbr       0/1   Completed  0        3m24s

3. Forward a local port to Milvus

Run the following command to get the port at which your Milvus cluster serves.

$ kubectl get pod my-release-milvus-proxy-6bd7f5587-ds2xv --template
='{{(index (index .spec.containers 0).ports 0).containerPort}}{{"\n"}}'
19530

The output shows that the Milvus instance serves at the default port 19530.

Then, run the following command to forward a local port to the port at which Milvus serves.

$ kubectl port-forward service/my-release-milvus 27017:19530
Forwarding from 127.0.0.1:27017 -> 19530

Optionally, you can use :19530 instead of 27017:19530 in the above command to let kubectl allocate a local port for you so that you don’t have to manage port conflicts.

By default, kubectl’s port-forwarding only listens on localhost. Use the address flag if you want Milvus to listen on the selected or all IP addresses. The following command makes port-forward listen on all IP addresses on the host machine.

$ kubectl port-forward --address 0.0.0.0 service/my-release-milvus 27017:19530
Forwarding from 0.0.0.0:27017 -> 19530

Offline install

If you are in a network-restricted environment, follow the procedure in this section to start a Milvus cluster.

1. Get Milvus manifest

Run the following command to get the Milvus manifest.

$ helm template my-release milvus/milvus > milvus_manifest.yaml

The above command renders chart templates for a Milvus cluster and saves the output to a manifest file named milvus_manifest.yaml. Using this manifest, you can install a Milvus cluster with its components and dependencies in separate pods.

2. Download image-pulling script

The image-pulling script is developed in Python. You should download the script along with its dependencies in the requirement.txt file.

$ wget https://raw.githubusercontent.com/milvus-io/milvus/master/deployments/offline/requirements.txt
$ wget https://raw.githubusercontent.com/milvus-io/milvus/master/deployments/offline/save_image.py

3. Pull and save images

Run the following command to pull and save the required images.

$ pip3 install -r requirements.txt
$ python3 save_image.py --manifest milvus_manifest.yaml

The images are pulled into a sub-folder named images in the current directory.

4. Load images

You can now load the images to the hosts in the network-restricted environment as follows:

$ for image in $(find . -type f -name "*.tar.gz") ; do gunzip -c $image | docker load; done

5. Deploy Milvus

$ kubectl apply -f milvus_manifest.yaml

Till now, you can follow steps 2 and 3 of the online install to check the cluster status and forward a local port to Milvus.

Upgrade running Milvus cluster

Run the following command to upgrade your running Milvus cluster to the latest version:

$ helm repo update
$ helm upgrade my-release zilliztech/milvus

Uninstall Milvus

Run the following command to uninstall Milvus.

$ helm uninstall my-release

What’s next

Having installed Milvus in Docker, you can:

• Check Hello Milvus to see what Milvus can do.

• Learn the basic operations of Milvus:

  • Manage Databases
  • Manage Collections
  • Manage Partitions
  • Insert, Upsert & Delete
  • Single-Vector Search
  • Hybrid Search

• Upgrade Milvus Using Helm Chart.

• Scale your Milvus cluster.

• Deploy your Milvus cluster on clouds:

  • Amazon EKS
  • Google Cloud
  • Microsoft Azure

• Explore Milvus Backup, an open-source tool for Milvus data backups.

• Explore Birdwatcher, an open-source tool for debugging Milvus and dynamic configuration updates.

• Explore Attu, an open-source GUI tool for intuitive Milvus management.

• Monitor Milvus with Prometheus.