Integrating DataDog with Sisense Linux

TL;DR

This article covers how to quickly integrate Datadog for monitoring Sisense deployed on Linux.

 

What You’ll Learn:

1. How to configure Datadog monitoring for Sisense Linux

 

Use Cases For Applying What You Will Learn:

1. Check Sisense cluster health
2. Access logs for troubleshooting
3. Assess resource consumption and resource needs

 

Prerequisites:

Basic knowledge of Linux; basic knowledge of Kubernetes; access to Datadog instance

 

General Overview:

By default, Sisense will log all system information locally. This can be accessed by logging into the server and viewing the logs under /var/log/{namespace}/sisense. System information can be viewed in a visual manner by logging into Sisense as an administrator and accessing Grafana.  However, if you wish to centralize data and log collection using Datadog, the following steps will explain how to configure it.

 

Integration Instructions

Step 1

Log into Datadog and ensure you have your API key readily available.  This can be found in multiple places within your Datadog instance.

Step 2

If Sisense is deployed on a single node, log in directly to the server.  If Sisense has been deployed into a multi-node cluster, log in to either the master node or bastion server that was used for installing Sisense.

Step 3

If you wish to install the Datadog agent under a specific namespace, please create that namespace using kubectl create ns <namespace>.  Otherwise, the agent will be installed into the default namespace.

 

Step 4

Install the Datadog Helm repository:

helm repo add datadog https://docs.datadoghq.com/getting_started/containers/datadog_operator/

Followed by:

helm repo update

Step 5

Copy the italicized text below to a yaml file on your server, such as 

datadog-values.yaml.  Please update the site, clusterName, and apiKey values appropriately where the site is the Datadog host (e.g. us1.datadoghq.com). 

targetSystem: "linux"
datadog:
  site: <Datadog host> 
  clusterName: <cluster name>
  apiKey: <api key>
  logs:
    enabled: true
    containerCollectAll: true
  apm:
    portEnabled: true
    socketPath: /var/run/datadog/apm.socket
    hostSocketPath: /var/run/datadog/
  processAgent:
    enabled: true
    processCollection: true
  systemProbe:
    enableTCPQueueLength: false
    enableOOMKill: true
    collectDNSStats: false

Step 6

To install the Datadog agent, execute helm install datadog-agent –namespace datadog -f datadog-values.yaml datadog/datadog, where “datadog-agent” is a chosen release name.  This can be modified to a value of your choice.

Step 7

Check the installation using helm ls -A to ensure the installation is successful.

Finalization

Wait a few minutes after the installation completes before proceeding.

Execute kubectl get pods -n <name space> | grep datadog

If all pods are in a full running state, congratulations, Datadog monitoring is now integrated with Sisense.  If not, try the steps in the Troubleshooting section below.

The Sisense logs can now be accessed directly from Datadog as shown in the screenshot below. 

Troubleshooting

Check the Datadog agent logs to see if any errors were captured:

kubectl logs -n <datadog namespace> datadog-agent-xxxx.  You will need to find the correct pod name, but it should look similar to the example.

If you see 403 unauthorized errors, the issue is likely related to your API key or site value.  Ensure both of these values are correct.

If this does not seem to be the issue, double-check that the text is in the appropriate YAML format.  This can be double-checked with a text editor.

When logged into Datadog, further installation steps can be followed here:

https://<datadog host>/account/settings#agent/kubernetes.

Conclusion

If you need additional help, please contact Sisense Support.