Nagios Support Forum

Hi

May I check with you, is there a way to allocate centralized storage(SSD) for the NLS cluster, rather than giving same size of storage for the each nodes in the cluster?
For example;
In case if we have to allocated 2 TB SSD for NLS cluster, then currently we have to allocate 2 TB SSD for each nls nodes in the cluster which is really cost effective and it will causes fail the project.

by configuring
in /etc/sysconfig/elasticsearch

and assign mounted SSD partition for all the nodes ?

DATA_DIR="/new/path/data"

Regards
Luke.

This is not how the backend elasticsearch database is designed. The whole point is that every node that exists in the cluster runs on separate equipment, if you loose a node then the data is distributed across the other nodes and can be recovered when a replacement node is added.

This design is completely different to historic solutions and is oriented at being more robust and disaster proof, the existing practices you are used to do not apply to this solution.

If you wanted to use SSD's on each node then this is simply going to be the associated cost for that type of setup.