Skip to content

Latest commit

 

History

History

helm-charts

Folders and files

NameName
Last commit message
Last commit date

parent directory

..
kafka
kafka
 
 
vektonn
vektonn
 
 
README.md
README.md
 
 

README.md

Helm Charts

To run provided charts locally you will need to install minikube and Helm.

  1. Start minkube:

    minikube start

  2. Deploy single-noded Kafka cluster:

    helm upgrade --install kafka-dummy kafka

    Note that kafka Helm chart presented here is for demonstration purposes only. It does not provide any persistent data storage!

  3. Deploy your first Vektonn release r1:

    helm upgrade --install vektonn-r1 vektonn

  4. Verify that everything is up and running:

    minikube kubectl get all
    NAME                                       READY   STATUS    RESTARTS   AGE
pod/kafka-dummy-5fb7b9d899-97pkc           1/1     Running   0          36s
pod/vektonn-index-shard-766db64645-m75zr   1/1     Running   0          7s
pod/vektonn-index-shard-766db64645-vck22   1/1     Running   0          7s
pod/vektonn-r1-api-5bfd6b5fcb-65cjp        1/1     Running   0          7s

NAME                          TYPE        CLUSTER-IP      EXTERNAL-IP   PORT(S)        AGE
service/kafka-broker          ClusterIP   None            <none>        9092/TCP       36s
service/vektonn-index-shard   ClusterIP   10.99.136.170   <none>        8082/TCP       7s
service/vektonn-r1-api        NodePort    10.104.24.141   <none>        80:30080/TCP   7s

NAME                                  READY   UP-TO-DATE   AVAILABLE   AGE
deployment.apps/kafka-dummy           1/1     1            1           36s
deployment.apps/vektonn-index-shard   2/2     2            2           7s
deployment.apps/vektonn-r1-api        1/1     1            1           7s

NAME                                             DESIRED   CURRENT   READY   AGE
replicaset.apps/kafka-dummy-5fb7b9d899           1         1         1       36s
replicaset.apps/vektonn-index-shard-766db64645   2         2         2       7s
replicaset.apps/vektonn-r1-api-5bfd6b5fcb        1         1         1       7s

  5. Provide access to Vektonn API service from host machine:

    minikube service --url vektonn-r1-api
    * Starting tunnel for service vektonn-r1-api.
|-----------|----------------|-------------|------------------------|
| NAMESPACE |      NAME      | TARGET PORT |          URL           |
|-----------|----------------|-------------|------------------------|
| default   | vektonn-r1-api |             | http://127.0.0.1:59241 |
|-----------|----------------|-------------|------------------------|

  6. Verify that Vektonn API service is respsonding:

    curl -v -X POST http://localhost:59241/api/v1/search/QuickStart.Index/1.0 \
--header 'Content-Type: application/json' \
--data @- <<BODY
{
  "k": 3,
  "queryVectors": [
    {
      "isSparse": false,
      "coordinates": [0.0, 0.0]
    }
  ]
}
BODY
    * Connected to localhost (::1) port 59241 (#0)
> POST /api/v1/search/QuickStart.Index/1.0 HTTP/1.1
> Host: localhost:59241
> User-Agent: curl/7.75.0
> Accept: */*
> Content-Type: application/json
> Content-Length: 98
>
} [98 bytes data]
< HTTP/1.1 200 OK
< Content-Type: application/json; charset=utf-8
< Transfer-Encoding: chunked
<
{ [90 bytes data]
[{"queryVector":{"isSparse":false,"coordinates":[0,0]},"nearestDataPoints":[]}]
* Connection #0 to host localhost left intact