Skip to content

Latest commit

 

History

History
63 lines (44 loc) · 1.91 KB

README.md

File metadata and controls

63 lines (44 loc) · 1.91 KB
Raw

Elasticsearch Tools

This is a set of tasks we use to manage an Elasticsearch cluster.

Available tasks are:

  • create an index,
  • optimize a whole cluster, starting with the index having the greatest number of deleted documents,
  • optimize an index,
  • enable / disable rack awareness,
  • perform a rolling restart,
  • include / exclude a rack, zone or a set of hosts,
  • update an index or a cluster refresh time,
  • update an index or a cluster number of replicas.

We use a set of variables for all our clusters:

  • rack_id: the rack name for rack awareness.
  • cluster_addr: by default, the host you'll target with the role
  • es_version: Elasticsearch version, 1, 2 or 5

Create an index

ansible-playbook -i inventory -l host -e task=create_index -e index_name [ -e shards=number_of shards -e replicas=number_of_replicas] playbooks/ops_es_tools.yml

Optimize a cluster

ansible-playbook -i inventory -l host -e task=optimize_cluster -e es_version={1,2,5} playbooks/ops_es_tools.yml

Optimize an index

ansible-playbook -i inventory -l host -e task=optimize_index -e es_version={1,2,5} -e index=index_name playbooks/ops_es_tools.yml

Rolling restart

ansible-playbook -i inventory -l cluster -e task=rolling_restart -e rack=foo playbooks/ops_es_tools.yml
ansible-playbook -i inventory -l cluster -e task=rolling_restart -e rack=bar playbooks/ops_es_tools.yml

Routing allocation

ansible-playbook -i inventory -l host -e task=routing_allocation -e index=index_name -e action={enable,disable} -e what={rack,zone,_ip} -e value="some,thing,or,nothing" playbooks/ops_es_tools.yml

Update refresh time

ansible-playbook -i inventory -l host -e task=update_refresh_time -e refresh_time=60s playbooks/ops_es_tools.yml

Update the number of replicas

ansible-playbook -i inventory -l host -e task=update_replicas -e replicas=42 [-e index=index_name] playbooks/ops_es_tools.yml