Virtual memory

ECK

By default, Elasticsearch uses memory mapping (mmap) to efficiently access indices. Default values for virtual address space on Linux distributions can be too low for Elasticsearch to work properly, which may result in out-of-memory exceptions. This is why the quickstart example disables mmap through the node.store.allow_mmap: false setting. For production workloads, we recommended you increase the kernel setting vm.max_map_count to 1048576 and leave node.store.allow_mmap unset.

The kernel setting vm.max_map_count=1048576 can be set on the host directly, by a dedicated init container which must be privileged, or a dedicated Daemonset.

Important

For Elasticsearch version 8.16 and later, set the vm.max_map_count kernel setting to 1048576; for Elasticsearch version 8.15 and earlier, set vm.max_map_count to 262144.

For more information, check the Elasticsearch documentation on Virtual memory.

Optionally, you can select a different type of file system implementation for the storage. For possible options, check the store module documentation.

		spec:
  nodeSets:
  - name: default
    count: 3
    config:
      index.store.type: niofs
		
	

Using an Init Container to set virtual memory

To add an init container that changes the host kernel setting before your Elasticsearch container starts, you can use the following example Elasticsearch spec:

		cat <<EOF | kubectl apply -f -
apiVersion: elasticsearch.k8s.elastic.co/v1
kind: Elasticsearch
metadata:
  name: quickstart
spec:
  version: 9.2.2
  nodeSets:
  - name: default
    count: 3
    podTemplate:
      spec:
        initContainers:
        - name: sysctl
          securityContext:
            privileged: true
            runAsUser: 0
          command: ['sh', '-c', 'sysctl -w vm.max_map_count=1048576']
EOF
		
	

Note that this requires the ability to run privileged containers, which is likely not the case on many secure clusters.

Using a Daemonset to set virtual memory

To use a Daemonset that changes the host kernel setting on all nodes:

		cat <<EOF | kubectl apply -n elastic-system -f -
apiVersion: apps/v1
kind: DaemonSet
metadata:
  name: max-map-count-setter
  labels:
    k8s-app: max-map-count-setter
spec:
  selector:
    matchLabels:
      name: max-map-count-setter
  template:
    metadata:
      labels:
        name: max-map-count-setter
    spec:
      initContainers:
        - name: max-map-count-setter
          image: docker.io/bash:5.2.21
          resources:
            limits:
              cpu: 100m
              memory: 32Mi
          securityContext:
            privileged: true
            runAsUser: 0
          command: ['/usr/local/bin/bash', '-e', '-c', 'echo 262144 > /proc/sys/vm/max_map_count']
      containers:
        - name: sleep
          image: docker.io/bash:5.2.21
          command: ['sleep', 'infinity']
EOF
		
	

To run an Elasticsearch instance that waits for the kernel setting to be in place:

		cat <<'EOF' | kubectl apply -f -
apiVersion: elasticsearch.k8s.elastic.co/v1
kind: Elasticsearch
metadata:
  name: elasticsearch-sample
spec:
  version: 9.2.2
  nodeSets:
  - name: default
    count: 1
    # Only uncomment the below section if you are not using the previous Daemonset to set max_map_count.
    # config:
    #  node.store.allow_mmap: false
    podTemplate:
      spec:
        # This init container ensures that the `max_map_count` setting has been applied before starting Elasticsearch.
        # This is not required, but is encouraged when using the previous Daemonset to set max_map_count.
        # Do not use this if setting config.node.store.allow_mmap: false
        initContainers:
        - name: max-map-count-check
          command: ['sh', '-c', "while true; do mmc=$(cat /proc/sys/vm/max_map_count); if [ ${mmc} -eq 262144 ]; then exit 0; fi; sleep 1; done"]
EOF
		
	