I'm planning to setup ElasticSearch across two datacentres. I know it is
not recommended approach to have the cluster across two data centres, but I
wanted an easy way to replicate the data cross site. I have some doubts
about how many master nodes I should set up.
Scenario 1:
Two master eligible nodes. One in each Site.
discovery.zen.minimum_master_nodes: 2
Site A: Master active
If Site A goes down a new master cannot be elected as min master node is
set to 2.
Question 1: How does site B behave? I believe, no new indexes or documents
can be created but the searches can still be carried out.
Scenario 2:
3 master eligible nodes. Two in Site A and One in Site B.
discovery.zen.minimum_master_nodes: 3/2 +1 = 2
Site A: Master active
If Site A goes down a new master cannot be elected as min master node is
set to 2.
Question 2: What kind master configuration can help mitigate such a
scenario?
I'm planning to setup Elasticsearch across two datacentres. I know it is
not recommended approach to have the cluster across two data centres, but I
wanted an easy way to replicate the data cross site. I have some doubts
about how many master nodes I should set up.
Scenario 1:
Two master eligible nodes. One in each Site.
discovery.zen.minimum_master_nodes: 2
Site A: Master active
If Site A goes down a new master cannot be elected as min master node is
set to 2.
Question 1: How does site B behave? I believe, no new indexes or documents
can be created but the searches can still be carried out.
Scenario 2:
3 master eligible nodes. Two in Site A and One in Site B.
discovery.zen.minimum_master_nodes: 3/2 +1 = 2
Site A: Master active
If Site A goes down a new master cannot be elected as min master node is
set to 2.
Question 2: What kind master configuration can help mitigate such a
scenario?
Apache, Apache Lucene, Apache Hadoop, Hadoop, HDFS and the yellow elephant
logo are trademarks of the
Apache Software Foundation
in the United States and/or other countries.