When configured correctly, Elasticsearch clusters are highly available and fault tolerant. That doesn’t necessarily mean they are impervious to failure. Human error and hardware failure are always a possibility. Troubleshooting data availability issues on a distributed system can be a challenge so let’s take some time to demystify some basic troubleshooting routines when you notice yellow or even red indexes in your Elasticsearch cluster. In this hands-on lab, you are given the opportunity to:
* Use the `_cat` APIs to quickly get vital information about your cluster
* Use the `_cluster/allocation/explain` API to find out why and index is or is not allocated
* Use the `_settings` API to quickly update index settings
Learning Objectives
Successfully complete this lab by achieving the following learning objectives:
- Troubleshoot the “accounts-1” index.
Using the Secure Shell (SSH), log in to the
data-2anddata-3nodes ascloud_uservia the public IP address.Become the
elasticuser with:sudo su - elasticStart the elasticsearch node with:
/home/elastic/elasticsearch/bin/elasticsearch -d -p pidUse the Kibana console tool to execute the following:
PUT accounts-1/_settings { "number_of_replicas": 3 }- Troubleshoot the “accounts-2” index.
Use the Kibana console tool to execute the following:
PUT accounts-2/_settings { "number_of_replicas": 3 }- Troubleshoot the “accounts-3” index.
Use the Kibana console tool to execute the following:
PUT accounts-3/_settings { "index.routing.allocation.exclude._name": null }
