Deploying the LogDNA Agent With Helm
Logging your Kubernetes clusters to LogDNA is already a breeze, and now the LogDNA Kubernetes Agent Helm chart makes it even easier.
Helm is the official package manager for Kubernetes. With Helm, deploying and managing Kubernetes applications is as simple as typing a single command. This makes deploying the LogDNA Agent across your cluster absolutely effortless. Best of all, you get the full benefits of the LogDNA Agent including host logging, additional Kubernetes metadata, and automatic updates.
This article will walk you through the process of deploying the LogDNA chart. Before getting started, make sure you have Helm installed and that your Kubernetes cluster is running version 1.13 or later.
Installing the LogDNA Kubernetes Agent Chart
To install the chart, simply run the following commands on your workstation. Replace LOGDNA_INGESTION_KEY with your actual LogDNA ingestion key, which you can find in the Add a Log Source page in the LogDNA web app. You can specify a release name by changing the RELEASE_NAME parameter, or remove it to let Helm generate a custom one.
$ helm repo add logdna https://assets.logdna.com/charts
$ helm install --set logdna.key=$LOGDNA_INGESTION_KEY my-release logdna/agent
Running this command creates a DaemonSet based on the LogDNA Kubernetes DaemonSet, which deploys the LogDNA Agent to each node in your cluster. The LogDNA Agent automatically monitors log files on the host, updates itself, and adds relevant Kubernetes metadata to each log. You can add custom tags to your logs by adding --set logdna.tags=$TAG to the helm install command. You can also set additional configuration parameters such as the LogDNA Agent image tag to use, resource limits on memory usage, and tolerations.
Lastly, if you need to uninstall the LogDNA Agent, you can do so by running:
$ helm delete RELEASE_NAME
To learn more, check out the chart on the documentation.
When you’re ready to get started, log into your LogDNA account, copy your ingestion key, and start logging your Kubernetes cluster in a matter of seconds. If you have any questions, please contact us.