Here are some of the questions that I have from a customer:
1 ) Is there a formula for determing the value for -Dopensearch.xcontent.string.length.max setting?
2) Is there a formula for determing the value for -Dopensearch.xcontent.name.length.max setting?
3) Is there a formula for determing the value for -Dopensearch.xcontent.depth.max setting?
4) Is there a formula for determing the value for -Dopensearch.xcontent.codepoint.max setting?
5) Should the client set -Dopensearch.xcontent.fast_double_writer to true or accept the default of false?
6) Does each node require private IP addresses?
7) Should these directory specified to store the indexes and always be used?
# Path to directory where to store the data (separate multiple locations by comma):
#
#path.data: /data/opensearch
path.data: /var/lib/opensearch
9) Is the default opensearch user required to have ssh connectivity?
10) Should OpenSearch be configured using the created user or as root user?
Hi Randy,
the questions from 1 to 5 are not clear to me. A specific formula for determining the value of these settings doesn’t exist as far as I am aware, but maybe if you queried GET /_nodes/jvm in the Rest API, you would find your answer.
Regarding the other questions :
6. The important factor is ensuring nodes can communicate effectively, whether via private or public IPs.
7. The choice of directory should be on fast storage (e.g., SSD) for better performance. /var/lib/opensearch is a common default path, but you can change it based on your environment.
8. The default OpenSearch user does not require SSH connectivity to function. OpenSearch nodes and users generally interact via HTTP/HTTPS (for RESTful APIs) and not directly over SSH.
9. It’s best practice to create a dedicated user for running OpenSearch, with the necessary privileges for accessing the required directories and performing tasks specific to OpenSearch.