(5.0.0-alpha5) Elasticsearch plugin is red; plugin:[email protected] Service Unavailable

[naisanza]

Kibana version: Kibana 5.0.0-alpha5

Elasticsearch version: Elasticsearch 5.0.0-alpha5

Server OS version: Proxmox Ubuntu 15.10 LXC

Browser version: Chrome/Firefox

Browser OS version:

Original install method (e.g. download page, yum, from source, etc.):

• https://download.elastic.co/kibana/kibana/kibana-5.0.0-alpha5-linux-x86_64.tar.gz
• https://download.elastic.co/elasticsearch/release/org/elasticsearch/distribution/tar/elasticsearch/5.0.0-alpha5/elasticsearch-5.0.0-alpha5.tar.gz
• http://javadl.oracle.com/webapps/download/AutoDL?BundleId=211989

Description of the problem including expected versus actual behavior:

Steps to reproduce:

1. Create Ubuntu 15.10 container; apt update; apt upgrade
2. Download java, kibana, elasticsearch
3. No changes to kibana.yml
4. Make changes to elasticsearch.yml (see below)
5. Start Elasticserach su elastic -c "ES_JAVA_OPTS='-Xms2048m -Xmx2048m' elasticsearch"
6. Start Kibana
7. Receive error (see screenshot)

root@Kibana-500-alpha5:~# grep -v "#" /opt/elasticsearch-5.0.0-alpha5/config/elasticsearch.yml                                                                                            cluster.name: cluster-mk1-beta
discovery.zen.minimum_master_nodes: 1
node.master: false
node.data: false

This error does not occur if elasticsearch.yml is set to:

network.host: _site_ 

and kibana.yml is set to:

elasticsearch.url: "http://10.1.10.213:9200"

Errors in browser console (if relevant):

Provide logs and/or server output (if relevant):

[jasontedor]

The problem is that you have node.data and node.master set to false; in this case, Kibana can not find the master node, and even if node.master were true but node.data is still false, after it creates the .kibana index, the master will not have a data node on which to allocate a primary shard for that index.

In the future, can you please not cross post? We are more than happy to help you even if that includes directing you from the elasticsearch repository to here, or vice-versa.

[naisanza]

@jasontedor that is what the documentation said to do for running kibana with a local elasticsearch client running to find the other nodes through it's discovery ability without requiring hardcoded addresses in the configuration

https://www.elastic.co/guide/en/kibana/current/production.html

[jasontedor]

Let's keep the discussion on elastic/elasticsearch#19968.

[jasontedor]

After further discussion with @naisanza on elastic/elasticsearch#19968, I'm reopening this as I think the documentation needs improvement. You can reference the discussion on elastic/elasticsearch#19968, but the crux of the matter is that the documentation currently makes no mention of configuring the client node to point to an Elasticsearch cluster.

[Bargs]

@jasontedor Is there any existing ES documentation that describes how to connect a node to a remote cluster? I expected there would be, but I couldn't find anything. If there is, it might be better to link the user to that rather than duplicate the instructions on this page. Since each user's network is going to be unique I think it might be tough to distill your comment in the ES ticket into a couple one-size-fits-all bullet points.

[ror6ax]

I have this issue while running Kibana on Mesos. We're using Mesos-DNS and I pass elasticsearch URL which is provided by it. Can you please advice what might be done to fix it?

ui settings Elasticsearch plugin is red
plugin:[email protected] Ready
plugin:[email protected] Not Found

[pitops]

i have the same problem as @ror6ax

Did you have any luck @ror6ax ?

Version 5.4.0 (logstash/kibana)

[thomasneirynck]

Seems this went stale. I will close this for now. We can always reopen if necessary.