Elasticsearch on Kommander

Elasticsearch overview

Elasticsearch is a distributed, RESTful search and analytics engine capable of addressing a growing number of use cases. As the core of the Elastic Stack, Elasticsearch centrally stores your data for lightning fast search, fine‑tuned relevancy, and powerful analytics that scale with ease.

Kommander catalog

Kommander catalog adds integration for Elasticsearch in Helm-based drivers

To access the catalog:

Create a Workspace
Attach a Cluster to the Workspace
Create a Project in the Workspace
Select the created Project
View the available Platform Services in the Project Catalog
Select the version you’d like to deploy from the version drop-down, and then select Deploy.

Install

From the Project Catalog select the desired version of Elasticsearch, and select Deploy.

The Kommander UI should resemble the following image. The dialog is populated with appropriate defaults:

Elasticsearch Service Install Configuration

The clusterName field above is used by Kibana and other components to refer to Elasticsearch.

With the included defaults, an Elasticsearch cluster with the following topology is created:

1 - Client Node
3 - Master Nodes
3 - Data Nodes

You will see the following pods under the project namespace on the Kubernetes cluster:

kubectl get pods

NAME                         READY   STATUS    RESTARTS   AGE
elasticsearch-oss-client-0   1/1     Running   0          69m
elasticsearch-oss-data-0     1/1     Running   0          69m
elasticsearch-oss-data-1     1/1     Running   0          69m
elasticsearch-oss-data-2     1/1     Running   0          69m
elasticsearch-oss-master-0   1/1     Running   0          69m
elasticsearch-oss-master-1   1/1     Running   0          69m
elasticsearch-oss-master-2   1/1     Running   0          69m

Each of the above nodes correspond to the following Elastic Node Roles

Parameters

The Kommander Catalog Elasticsearch Platform Service creates an ensemble of the upstream Elastic Helm Chart

Full list of Configuration Parameters that can be applied to any of the current node roles:

The current set of default parameters applied can be found in this file.

Update parameters

Update parameters by directly modifying them in the Kommander Catalog UI:

Example: Increasing Data Node Replicas from the default to 4.

Select Edit Service for the selected Elasticsearch instance.

Modify elasticsearch-data to include the updated replicas count, as seen below:

elasticsearch-data:
# It is suggested that this match the catalog addon name.
clusterName: "elasticsearch-oss"
# This must equal <elasticsearch-master.clusterName>-master
masterService: "elasticsearch-oss-master"
replicas: 4

Select Save

The output below shows that we now have 4 data nodes.

kubectl get pods

NAME                         READY   STATUS    RESTARTS   AGE
elasticsearch-oss-client-0   1/1     Running   0          70m
elasticsearch-oss-data-0     1/1     Running   0          70m
elasticsearch-oss-data-1     1/1     Running   0          70m
elasticsearch-oss-data-2     1/1     Running   0          70m
elasticsearch-oss-data-3     1/1     Running   0          2m8s
elasticsearch-oss-master-0   1/1     Running   0          70m
elasticsearch-oss-master-1   1/1     Running   0          70m
elasticsearch-oss-master-2   1/1     Running   0          70m

Monitoring

Kommander includes Prometheus and Grafana as part of the federated Workspace Platform Services along with centralized monitoring.

The Kommander Catalog includes Elasticsearch-Exporter to be used in conjunction with Elasticsearch to export metrics to Prometheus and Grafana.

External access

The Kommander Catalog includes Kibana which provides a graphical way to view data stored in Elasticsearch

The following services are exposed by Elasticsearch:

kubectl get services

NAME                                TYPE        CLUSTER-IP    EXTERNAL-IP   PORT(S)             AGE
elasticsearch-oss-client            ClusterIP   10.0.30.140   <none>        9200/TCP,9300/TCP   91m
elasticsearch-oss-client-headless   ClusterIP   None          <none>        9200/TCP,9300/TCP   91m
elasticsearch-oss-data              ClusterIP   10.0.5.4      <none>        9200/TCP,9300/TCP   91m
elasticsearch-oss-data-headless     ClusterIP   None          <none>        9200/TCP,9300/TCP   91m
elasticsearch-oss-ingest            ClusterIP   10.0.29.173   <none>        9200/TCP,9300/TCP   91m
elasticsearch-oss-ingest-headless   ClusterIP   None          <none>        9200/TCP,9300/TCP   91m
elasticsearch-oss-master            ClusterIP   10.0.48.163   <none>        9200/TCP,9300/TCP   91m
elasticsearch-oss-master-headless   ClusterIP   None          <none>        9200/TCP,9300/TCP   91m

Elasticsearch provides comprehensive REST API

The following example shows how to query the Elasticsearch REST API:

Port-forward the elasticsearch-client service:

kubectl port-forward service/elasticsearch-oss-client 9200:9200 &

Forwarding from 127.0.0.1:9200 -> 9200
Forwarding from [::1]:9200 -> 9200
Handling connection for 9200

Elasticsearch should be ready to receive requests:

$ curl -s http://127.0.0.1:9200 | jq
{
"name": "elasticsearch-oss-client-0",
"cluster_name": "elasticsearch-oss",
"cluster_uuid": "TLAr4CMwSBKb2e45dHUwgQ",
"version": {
    "number": "7.10.0",
    "build_flavor": "oss",
    "build_type": "docker",
    "build_hash": "51e9d6f22758d0374a0f3f5c6e8f3a7997850f96",
    "build_date": "2020-11-09T21:30:33.964949Z",
    "build_snapshot": false,
    "lucene_version": "8.7.0",
    "minimum_wire_compatibility_version": "6.8.0",
    "minimum_index_compatibility_version": "6.0.0-beta1"
},
"tagline": "You Know, for Search"
}

As an example, query the Cluster REST API

curl -s http://127.0.0.1:9200/_nodes/_master | jq
{
"_nodes": {
    "total": 1,
    "successful": 1,
    "failed": 0
},
"cluster_name": "elasticsearch-oss",
"nodes": {
    "7HxKd320QnGnimu9ldny7A": {
    "name": "elasticsearch-oss-master-0",
    "transport_address": "192.168.110.153:9300",
    "host": "192.168.110.153",
    "ip": "192.168.110.153",
    "version": "7.10.0",
    "build_flavor": "oss",
    [...output omitted...]
        }
    }
}

Troubleshooting

To troubleshoot deployments, look for issues in the following:

Ensure clusterName and masterService fields are consistent across all node-roles.
Look for crash-looping pods and inspect their logs.
```
watch kubectl get pods
```
Monitor all the events occurring in the namespace, this can help detect common issues such as insufficient resources on the cluster to start various pods.
```
kubectl get events -w -n <namespace>
```