Skip to content

Latest commit

 

History

History
137 lines (89 loc) · 6.62 KB

README.md

File metadata and controls

137 lines (89 loc) · 6.62 KB
Raw

Confluent Cloud Monitoring with OpenTelemetry Collector and New Relic Cloud

Prerequisites

New Relic

You need to create a New Relic account.

License Key

Once you have logged in to your account, you need to collect your Ingest Key which will be needed in the next steps. You can find it on your New Relic UI, from your Profile - API Keys menu. Find your INGEST - LICENSE | License Key and then Copy Key from the 3 dots options.

Confluent Cloud

Assuming you have access to Confluent Cloud and have a Kafka Cluster running, you need to create some API Keys needed for this configuration.

Cloud API Key

You create a Cloud API Key in different ways:

  • Terraform provider: It is recommended to create Service Accounts and API Keys in an automated way and the Confluent Terraform provider allows us to do so:
  • Confluent CLI: You can find all the steps to create a Service Account, a Cloud API Key and assign a MetricsViewer role from the CLI in this documentation.
  • Cloud GUI: From the hamburger menu on the upper right corner of your browser select Cloud API Keys - from here select Add Key, then select Granular Access and you can Create a new Service Account if you do not already have one. Once you create the Service Account, you’ll get an API Key pair. Make sure you download it and store it somewhere secure (e.g. a Key Management System). Going back to the hamburger menu, from the Accounts & Access management you can assign the new Service Account a role, in this case the MetricsViewer role.

Cluster API Key

Similarly to the Cloud API Key, you can create a Cluster API Key in different ways:

Login:

confluent login

Select your environment and Kafka Cluster:

confluent environment list 
confluent environment use <env-id>
confluent kafka cluster list
confluent kafka cluster use <cluster-id>

Create a new Service Account:

confluent iam service-account create "new-relic-sa" --description "Service account for New Relic Collector"

Assign a role to the Service Account:

confluent iam rbac role-binding create --principal User:sa-123456 --role CloudClusterAdmin

Create API Key:

confluent api-key create --service-account $serviceAccountId --resource $clusterId
  • Cloud GUI: From the hamburger menu on the upper right corner of your browser select Accounts & Access - from here you need to create a new Service Account. Once you have created the Service Account, you can assign it a role by selecting the Access tab - you can assign it a role of CloudClusterAdmin for your Kafka cluster.

Virtual Machine

For this example I am working on Azure Cloud.
Create a Linux Virtual Machine on your VNet (I’m using Ubuntu v20.04), with access to Confluent Cloud. Make sure it has GoLang installed (version should be 1.19.x) and make (sudo apt install make) is also installed to build the collector.

Open Telemetry Collector

You need to clone the following GitHub repository and build the Collector using make.

git clone https://github.com/4demos/opentelemetry-collector-contrib.git

cd opentelemetry-collector-contrib

make otelcontribcol

If the make command returns an error like the following:

build github.com/Azure/azure-sdk-for-go/sdk/azcore: cannot load github.com/Azure/azure-sdk-for-go/sdk/azcore: no Go source files

Build the AzCore SDK and then run the make command again:

go get -u github.com/Azure/azure-sdk-for-go/sdk/azcore

make otelcontribcol

At the end of the build it’ll create a binary in the ./bin directory.

Configure Collector

Once your OpenTelemetry Collector is built, you can work on the configuration. Create a config.yaml file containing the following:

jmx-monitoring-stacks/ccloud-opentelemetry-newrelic/assets/newrelic/config.yaml

Lines 1 to 59 in 57c57f7

receivers:
kafkametrics:
brokers:
- $CLUSTER_BOOTSTRAP_SERVER
protocol_version: 2.0.0
scrapers:
- brokers
- topics
- consumers
auth:
sasl:
username: $CLUSTER_API_KEY
password: $CLUSTER_API_SECRET
mechanism: PLAIN
tls:
insecure_skip_verify: false
collection_interval: 30s
prometheus:
config:
scrape_configs:
- job_name: "confluent"
scrape_interval: 60s # Do not go any lower than this or you'll hit rate limits
static_configs:
- targets: ["api.telemetry.confluent.cloud"]
scheme: https
basic_auth:
username: $CONFLUENT_API_ID
password: $CONFLUENT_API_SECRET
metrics_path: /v2/metrics/cloud/export
params:
"resource.kafka.id":
- CLUSTER_ID
exporters:
otlp:
endpoint: $OTLP_ENDPOINT
headers:
api-key: $NEW_RELIC_LICENSE_KEY
processors:
batch:
memory_limiter:
limit_mib: 400
spike_limit_mib: 100
check_interval: 5s
service:
telemetry:
logs:
pipelines:
metrics:
receivers: [prometheus]
processors: [batch]
exporters: [otlp]
metrics/kafka:
receivers: [kafkametrics]
processors: [batch]
exporters: [otlp]

You need to replace the highlighted placeholders with the API Key pairs created previously, your Confluent Cloud cluster information (bootstrap url and cluster id), the OTLP url for New Relic and the New Relic license key.

OTLP Endpoint

Make sure you use the correct url for the OTLP endpoint in your configuration file, based on your New Relic region:

New Relic Dashboard

You can find here an example of Confluent Cloud Dashboard that you can import to your New Relic account.
You need to update the dashboard JSON file with your accountId information as it is set to 0 in multiple places. You will also find an accountIds array at the bottom of the dashboard which also needs to include your accountId.

Import Dashboard with Terraform

Create a new dir for your Terraform configuration. Save your dashboard to dashboards/cc-dashboard.json.
Create the Terraform configuration main.tf which will contain your configuration as below:

jmx-monitoring-stacks/ccloud-opentelemetry-newrelic/assets/terraform/main.tf

Lines 1 to 33 in 57c57f7

terraform {
# Require Terraform version 1.0 (recommended)
required_version = "~> 1.0"
# Require the latest 2.x version of the New Relic provider
required_providers {
newrelic = {
source = "newrelic/newrelic"
}
}
}
provider "newrelic" {
account_id = vars.account_id # Your New Relic account ID
api_key = vars.user_key # Your New Relic user key
region = vars.region # US or EU (defaults to US)
}
resource "newrelic_one_dashboard_json" "cc_dashboard" {
json = file("${path.module}/dashboards/cc-dashboard.json")
}
resource "newrelic_entity_tags" "cc_dashboard" {
guid = newrelic_one_dashboard_json.cc_dashboard.guid
tag {
key = "terraform"
values = [true]
}
}
output "cc_dashboard" {
value = newrelic_one_dashboard_json.cc_dashboard.permalink
}

Example Dashboard