Updated April 18, 2023
Definition of Kubernetes Daemonset
Kubernateshas has a wide variety of options and deployment methods to execute the containers. Daemonset is one among them. The kubernetes Daemonset is a tool for containers to check that all nodes and their subset are executed on one copy of a pod. Daemonsets can develop the pod on new nodes which can be added to the cluster. When implying the Kubernetes, most of them don’t think about the execution of pods. But at times, the user wants to execute a single pod on all the nodes. If the user wants to execute fluentd on all the nodes to gather logs, he uses Daemonset is implied to be in contact with kubernetes to ensure that the instances of the pods are in the nodes of the cluster.
What is Kubernetes Daemonset?
The Daemonset has usually explained with a YAML file and the components in the YAML files give the user the additional control over the deployment process of the pod. Pods are the simple objects deployed in Kubernetes which signifies the single instances of an executable process in the cluster. Pod comprises one or multiple containers which are managed as single resources.
A Daemonset is a dynamic object in kubernetes which is managed by a controller. The user can set the desired state that represents the specific pods which need to exist on every node. The compromise in the control loop can compare the current practical state with the desired state. If the practical node, doesn’t fit the matching pod, then the controller of the Daemonset creates a new one automatically. This automated process has all recently created nodes and existing nodes. The pods developed by the controller of Daemonset are neglected by the Kubernetes scheduler and present as the same node itself.
What are the Use Case for Daemonset?
The use case for Daemonset are as follows,
To execute a daemon for storage of cluster on every node, use: -glusterd-ceph.
To execute a daemon for the collection of logs on every node, use: fluentd-logstash.
To execute a daemon for monitoring of nodes on every node, use: -collectd, -datadog agent as Prometheus node exporter.
As the use case is complex, the user can deploy many possible Daemonsets for a typical kind by a variety of flags, CPU requests, and memory units for different types of hardware.
Daemonset can enhance the performance of the cluster by pod deployment to execute maintenance activities and support services for every node. The background processes like monitoring apps in Kubernetes and other services must be present along with clusters to offer timely and relevant services.
It is mostly adapted for long-running services like collection of logs, monitoring of node resources, storage in clusters, and pods related to infrastructure. It is standard for Daemonset to run on a single daemon throughout all the nodes in the cluster. The multiple Daemonset can manage one type of Daemonset by applying different labels. This label can identify rules in deployment based on the behavior of individual nodes. Daemonsets is deployed on the background tasks which the user can execute on all the possible nodes but it doesn’t need any intervention. The user can have Daemonset for every type of daemon to execute on all nodes. The user can also execute multiple Daemonset for every single daemon type. But can use various configurations for diverse hardware models and resource requirements.
How to Create a Daemonset?
- To create a Daemonset the following steps are involved.
- The Daemonset is developed in a YAML file with few parts.
- It requires apiVersion
- It requires the type or kind of the behavior for Daemonset
- It needs the metadata for the Daemonset
- It needs spec_template for pod definition which the user needs to execute on all nodes.
- It needs spec_selector to manage the pods for Daemonset and this kind must be label specific in a template of pods. The selector name defined in the template is applied in the selector. But this name cannot be changed once the Daemonset is created without leaving the pods created prior in the Daemonset.
- The spec_template-spec-node selector is used to execute only the nodes subset which suits the selector
- Spec-template-spec-affinity is used to execute on nodes subset which has affinity match.
- Once the configuration is completed, Daemonset is created in the cluster.
Methods to communicate with pods Kubernetes Daemonset
The user can transfer data from other pods to a pod of Daemonset executed on the node. The user can transmit data metrics to pod monitoring in two ways,
The pod spec in Daemonset has specifiy host port to uncover it on the node. Then it can interact directly with the IP of node which is executed on it.
Using the same pod selector, the service is created as a Daemonset and then it is used to reach the Daemonset. The limitation of this service is that a Daemonset of a pod is random to receive the request. This implies that the request may pass over the network instead of hanging on a similar pod as the requested pod.
The possible patterns for communicating daemon pods are push, nodeIP with known port, DNS, and service.
In push, the pods placed in Daemonset are customized to transfer the updates to other services like stats database and they do have no clients.
In NodeIP with Known port, the Daemonset pods use the hostport, and so the pods are connected through node IPs. The client understands the use nodeIP list and is also aware of the port conventions.
DNS develop headless service and explore Daemonset using endpoint to retrieve reports from DNS
In-Service, it creates the service with a similar pod selector and uses it to reach the daemon of the random node.
Conclusion
Hence, the feature of Daemonset is used to check that few or all the pods are executed and scheduled on every available node. It executes the copy of the desired pod throughout all the nodes. So when a new node is joined to the cluster of Kubernetes, the new pod is added to the recently attached node.
Recommended Articles
This is a guide to Kubernetes Daemonset. Here we discuss Definition, What is Kubernetes Daemonset? How to create a Daemonset?. You may also have a look at the following articles to learn more –
