Adjust log storage size

Big picture

Adjust the size of the Calico Enterprise log storage during or after installation.

Value

By default, Calico Enterprise creates the log storage with a single node. This makes it easy to get started using Calico Enterprise. Generally, a single node for logs is fine for test or development purposes. Before going to production, you should scale the number of nodes, replicas, CPU, and memory to reflect a production environment.

Concepts

Log storage terms

Term	Description
node	A running instance of the log storage.
cluster	A collection of nodes. Multiple nodes protect the cluster from any single node failing, and lets you scale resources (CPU, memory, storage space) .
replica	A copy of data. Replicas protect against data loss if a node fails. The number of replicas must be less than the number of nodes.

Before you begin...

Review log storage recommendations

Review Log storage recommendations for guidance on the number of nodes and resources to configure for your environment.

How to

• Adjusting LogStorage

caution

If you are not using a dynamic provisioner, make sure there is an available persistent volume before updating the resource requirements (cpu, memory, storage) in this section. To check that a persistent volume has the status of Available, run this command: kubectl get pv | grep tigera-elasticsearch

Adjusting LogStorage

In the following example, Calico Enterprise is configured to install 3 nodes that have 200Gi of storage each with 1 replica. Whenever the storage size is modified, resourceRequirements must be revisited respectively to support these changes.

apiVersion: operator.tigera.io/v1
kind: LogStorage
metadata:
  name: tigera-secure
spec:
  indices:
    replicas: 1
  nodes:
    count: 3
      # This section sets the resource requirements for each individual Elasticsearch node.
    resourceRequirements:
      limits:
        cpu: 1000m
        memory: 16Gi
      requests:
        cpu: 1000m
        memory: 16Gi
        storage: 200Gi
  componentResources:
    - componentName: ECKOperator
        # This section sets the resource requirements for the operator that bootstraps the Elasticsearch cluster.
      resourceRequirements:
        limits:
          memory: 512Mi
        requests:
          memory: 512Mi