Nagios Support Forum

Hi Team,
My new project is to install the actual Kibana in front of NLS using the same elasticsearch and to view the dashboard differences.
1. I have installed Kibana on the same NLS machine(I have a 4 node cluster--- Kibana installed on node1)
2. I am able to visulaze all logs collected by Elasticsearch search and view in Kibana.
3.However, When I use Kibana and NLS the performace gets impacted and causes memory issue.
4.Currently my servers are running on 64GB.

Ques:
1. Should I revisit the server metrics again?
2. Should the memory be increased?
3. Should I install Kibana on all 4 servers for better performance.

uma K wrote:My new project is to install the actual Kibana in front of NLS using the same elasticsearch and to view the dashboard differences.

This is not supported.

Your different Kibana is going to try to read and write to an altered index in a format that is different than what Nagios Log Server expects and could potentially mess up your Nagios Log Server dashboards.

Running anything outside of the stack that comes with NLS is untested territory and not something we can really support, but if you'd like to try increasing the RAM on the machine, make sure to that elasticsearch doesn't use more than 32Gigs as this will cause performance problems. By default elasticseach will get half of the total system memory, but this can be manually set by editing /etc/sysconfig/elasticsearch and setting the ES_HEAP_SIZE.

Thanks for your clear explanation.