Enabling the nodes for data lake components

When you have a multi-node setup, you can dedicate a node to run data lake components only like MinIO, PostgreSQL, Redis, Kafka, and so on.

To enable the node for data lake components

1. Identify the node to run only infra components.
2. Run the following command to enable the node for infra components.
   $ kubectl taint node <nodename> helix_dedicated=true:NoSchedule
3. Run the respective commands to assign or label the node for each infra component that you want to run on the dedicated node.
   ​
   MINIO
   $ kubectl label node <nodename> dedicated_node_minio=true
   POSTGRES
   $ kubectl label node <dedicated node names> dedicated_node_pg=true
       $ kubectl label node <dedicated node names>dedicated_node_pgpool=true
   KAFKA
   $ kubectl label node <nodename> dedicated_node_kafka=true
   REDIS
   $ kubectl label node <nodename> dedicated_node_redis=true
   VM
   $ kubectl label node <nodename> dedicated_node_vm=true
   ELASTICSEARCH
   $ kubectl label node <nodename> dedicated_node_es=true
   # Postgres-ha
      $ kubectl label node <dedicated node names> dedicated_node_pg=true
   # pgpool
      $ kubectl label node <dedicated node names> dedicated_node_pgpool=true
   
   # kafka
      $ kubectl label node <dedicated node names> dedicated_node_kafka=true
   # zookeeper
      $ kubectl label node <dedicated node names> dedicated_node_zookeeper=true
   
   # redis-ha
      $ kubectl label node <dedicated node names> dedicated_node_redis_ha=true
   # redis-haproxy
      $ kubectl label node <dedicated node names> dedicated_node_redis_haproxy=true
   
   # victoriametrics-insert
      $ kubectl label node <dedicated node names> dedicated_node_vminsert=true
   # victoriametrics-select
      $ kubectl label node <dedicated node names> dedicated_node_vmselect=true
   # victoriametrics-storage
      $ kubectl label node <dedicated node names> dedicated_node_vmstorage=true
   # victoriametrics-insert-agg
      $ kubectl label node <dedicated node names> dedicated_node_vminsert_agg=true
   # victoriametrics-select-agg
      $ kubectl label node <dedicated node names> dedicated_node_vmselect_agg=true
   # victoriametrics-storage-agg
      $ kubectl label node <dedicated node names> dedicated_node_vmstorage_agg=true
   
   # opensearch-events-master
      $ kubectl label node <dedicated node names> dedicated_node_opensearch_events_master=true
   # opensearch-events-data
      $ kubectl label node <dedicated node names> dedicated_node_opensearch_events_data=true
   # opensearch-logs-master
      $ kubectl label node <dedicated node names> dedicated_node_opensearch_logs_master=true
   # opensearch-logs-data
      $ kubectl label node <dedicated node names> dedicated_node_opensearch_logs_data=true
   
   # minio
      $ kubectl label node <dedicated node names> dedicated_node_minio=true
4. Open infra.config file and edit the ENABLE_DEDICATED_NODES_FOR_INFRA parameter value to yes.
5. For more details, See Configuration-file-settings.
6. Open dedicated-nodes-support.config file and set the value to yes for the component that you want to enable the dedicated node.
   For example, to enable dedicated node for only redis, set the redis parameter values to yes as below:
   # Postgres
   DN_PTPOSTGRES_HA=no
   DN_PTPGPOOL=no
   
   # Kafka
   DN_KAFKA=no
   DN_ZOOKEEPER=no
   
   # Redis
   DN_REDIS_HA=yes
   DN_REDIS_HAPROXY=yes
   
   # VicotriaMetrics
   DN_VICTORIAMETRICS_INSERT=no
   DN_VICTORIAMETRICS_SELECT=no
   DN_VICTORIAMETRICS_STORAGE=no
   DN_VICTORIAMETRICS_AGG_INSERT=no
   DN_VICTORIAMETRICS_AGG_SELECT=no
   DN_VICTORIAMETRICS_AGG_STORAGE=no
   
   # OpenSearch
   DN_ELASTICSEARCH_EVENTS_MASTER=no
   DN_ELASTICSEARCH_EVENTS_DATA=no
   DN_ELASTICSEARCH_LOGS_MASTER=no
   DN_ELASTICSEARCH_LOGS_DATA=no
   
   # Minio
   DN_MINIO=no