Install Birdwatcher
This page demonstrates how to install Birdwatcher.
Local install
If you have installed Milvus Standalone using docker, you’d better download and install the built binary, install Birdwatcher as a common Go module, or build Birdwatcher from the source.
Install it as a common Go module.
git clone https://github.com/milvus-io/birdwatcher.git cd birdwatcher go install github.com/milvus-io/birdwatcher
Then you can run Birdwatcher as follows:
go run main.go
Build it from the source.
git clone https://github.com/milvus-io/birdwatcher.git cd birdwatcher go build -o birdwatcher main.go
Then you can run Birdwatcher as follows:
./birdwatcher
Download the already-built binary
First, open the latest release page, and find the prepared binaries.
wget -O birdwatcher.tar.gz \ https://github.com/milvus-io/birdwatcher/releases/download/latest/birdwatcher_<os>_<arch>.tar.gz
Then you can decompress the tarball and use Birdwatcher as follows:
tar -xvzf birdwatcher.tar.gz ./birdwatcher
Install as a Kubernetes pod
If you have installed either Milvus Standalone using the Helm charts or the Milvus Operator or Milvus Cluster using the Helm charts or the Milvus Operator, you are advised to install Birdwatcher as a Kubernetes pod.
Prepare deployment.yml
apiVersion: apps/v1
kind: Deployment
metadata:
name: birdwatcher
spec:
selector:
matchLabels:
app: birdwatcher
template:
metadata:
labels:
app: birdwatcher
spec:
containers:
- name: birdwatcher
image: milvusdb/birdwatcher
resources:
limits:
memory: "128Mi"
cpu: "500m"
If the image available on DockerHub is not the latest, you can build an image of Birdwatcher using the Dockerfile provided with the source code as follows:
git clone https://github.com/milvus-io/birdwatcher.git
cd birdwatcher
docker build -t milvusdb/birdwatcher .
To deploy a locally built image, you need to add imagePullPolicy
to the above specs and set it to Never
.
...
- name: birdwatcher
image: milvusdb/birdwatcher
imagePullPolicy: Never
...
Apply deployment.yml
Save the above YAML in a file and name it deployment.yml
, and run the following command
kubectl apply -f deployment.yml