I just had some quick questions on Elasticsearch that have to do with network configuration.
My first question is this, can you send data to Elasticsearch using the default .yml settings? For example if Elasticsearch is only accessible to the local host can I still send data to it through the port from another computer? (Example IP I would configure in my beats: 10.0.0.1:9200). My previous understanding led me to believe that Elasticsearch has to be set to a public IP to receive data from other computers Ex. (0.0.0.0) but now I don't think this is the case.
Another question I had was on Discovery Seed Hosts. I know that if you are planning on using an IP other than the default IP then you have to configure 1 of 3 things with discovery seed hosts being one of them. What are the other options to configure and why does this need to be done?
Detailed explanations or links to detailed explanations would be greatly appreciated.
For example if Elasticsearch is only accessible to the local host can I still send data to it through the port from another computer?
How exactly are you imagining this working from a network standpoint?
Like my gut reaction is "No, that defeats the entire purpose of using localhost and a loopback interface" but I guess technically yea, you could setup some kind of rule to forward port 9200 on your actual address to the loopback address. But that would be weird and pointlessly confusing.
As for discovery, elastic needs a list of masters so it can create/join a cluster. Every node needs to know who to tell it would like to join the cluster. If you have a cluster of 1
discovery.type: single-node
will let elasticsearch know to not bother with that.
The answer for the second question makes sense. Also I figured out the second question I was just wording it wrong. Thanks for the response. I really appreciate the help!
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.