Bogus blocking error: "This version of Kibana requires at least Elasticsearch 1.4.0.Beta1"

[konste]

I have ES cluster upgraded to 1.4.0.Beta1 specifically to try Kibana 4.0.0 Beta1, but unfortunately all I get when navigate to http://localhost:5601 is bogus error message: "This version of Kibana requires at least Elasticsearch 1.4.0.Beta1" Whatever I do I cannot pass beyond that state. Please help!

Konstantin

[spalger]

Kibana uses the nodes info api to read the version of all nodes in the cluster in order to verify that the entire cluster is compatible with the requests it will be making. Would you mind requesting this info from your cluster and pasting the output here?

curl -XGET 'http://localhost:9200/_nodes'

[konste]

Aha! Spencer, your answer actually shed some light on this mystery! Other than three "real" nodes (all 1.4.0.Beta1) nodes info API shows two logstash nodes, which report their version as 1.1.1
Logstash nodes are not real ES nodes, it is merely the way Logstash connects to cluster in order to send documents for indexing.
So on one hand Kibana probably should ignore those as they are not going to execute any query anyways. But as a short term solution I can switch Logstash to some other way of talking to ES and avoid creation of those fake nodes, although it is the default for Logstash.

Please confirm my conclusions.

Thanks a lot!
Konstantin

[spalger]

Yep! That's the advised solution for now. Logstash's HTTP elasticsearch output should do the trick.

[mohameden]

I'm a newbie to the ELK stack, and I'm trying to understand why I have a logstash node in my ES cluster. I have no running agent of logstash. So, I tried to delete the logstash node from my cluster (to be able to use kibana 4), but It still appear ..
The command :
curl -XGET 'http://localhost:9200/_nodes' | grep logstash
is still showing results
I have the following in my ES logs:
[2014-10-20 19:01:04,133][INFO ][http ] [Fabian Cortez] bound_address {inet[/0:0:0:0:0:0:0:0:9200]}, publish_address {inet[/184.7.127.234:9200]}
[2014-10-20 19:01:04,133][INFO ][node ] [Fabian Cortez] started
[2014-10-20 19:01:04,141][INFO ][gateway ] [Fabian Cortez] recovered [0] indices into cluster_state
[2014-10-20 19:01:05,911][INFO ][cluster.service ] [Fabian Cortez] added {[logstash-charmy-29733-4108][Lwi0n1HTQWyxoSUbYdP13w][charmy][inet[/184.7.127.234:9301]]{client=true, data=false},}, reason: zen-disco-receive(join from node[[logstash-charmy-29733-4108][Lwi0n1HTQWyxoSUbYdP13w][charmy][inet[/184.7.127.234:9301]]{client=true, data=false}])


[konste]

Are you saying that you do not have Logstash running as a service (or console app) on the machine 184.7.127.234? Because that's exactly what your output implies! Or you do have Logstash running there, but wonder how come it creates ES node? That one I can explain.

[mohameden]

My bad, I found a running logstash app with ps -aux | grep logstash. Everything is OK now. Thanks a lot :)

[deviantony]

Got the same issue, with logstash reporting as being a ES node in version 1.1.1. Isn't that weird though? I'm using logstash 1.4.2.

Also, I'd like to keep the elasticsearch output in logstash (node mode, without specifying a host to use discovery) and not the elasticsearch_http one, is that integration planned with Kibana4?

[Seidr]

FYI, I managed to resolve this by upgrading the version of ElasticSearch which my version of logstash uses.

1. Download ElasticSearch 1.4.0.Beta1 from wget https://download.elasticsearch.org/elasticsearch/elasticsearch/elasticsearch-1.4.0.Beta1.zip
2. Stop logstash
3. Unzip, and place into the directory <logstash installation directory>/vendor/jar/
4. Remove (backup) the directory named <logstash installation directory>/vendor/jar/elasticsearch-1.1.1
5. Start logstash

So far, so good. No errors in the logstash console, messages from my pre-existing shippers are being received and Kibana 4 now loads fine (as does my Kibana 3 instance).

[rudijs]

@spenceralger +1 Use output elasticsearch_http and not elasticsearch fixes this for me.

[mingderwang]

ROOT CAUSE: http://awstaiwan.blogspot.tw/2015/01/kibana-this-version-of-kibana-requires.html

[dc-afinnell]

It's sort of confusing how this problem ever made it into the beta branch to begin with. Wouldn't all the developers encounter this problem? Or do all the developers have logstash connect using the _http output?

[bjolivot]

@Seidr Thank you

[i-ask-too-many-questions]

I had the same problem and this is how I fixed it:
in Logstash-1.4.2 go to vendor/jar and there you will find a copy of elasticsearch-1.1.1. Deleting it alone won't solve the problem. I replaced it with elasticsearch-1.4.2 and everything seems to work just fine after that!!

[tonygaetani]

@rudijs changing my logstash output config from elasticsearch to elasticsearch_http worked for me too! Thanks!

[argent-smith]

This doesn’t work for me. ES 1.4.4, LS 1.4.2, output elasticsearch_http

[ghost]

Hello,

I have below versions:

elasticsearch-1.4.2.tar.gz
kibana-4.0.2-linux-x86.tar.gz
logstash-1.4.2.tar.gz ==> First tried on this as it's stable.
logstash-1.5.0.rc2.tar.gz ==> later on this but no luck.

I'm trying ESK stack on Ubuntu headless VM box. It's stuck at kibana UI.

http://blog.keithkim.com/2015/01/kibana-elasticsearch-logstash-issue.html doesn't say exactly which command to hit and kill which ES node out of http://www.elastic.co/guide/en/elasticsearch/reference/current/cluster-nodes-shutdown.html ?

Where do you change exactly ? Is it in ES config or kibana config file. Which file, path? and what to what so ES will use http

[ghost]

I got it resolved after upgrading elastic search to elasticsearch-1.5.1.tar.gz

I used elasticsearch_http as output in conf file of logstat. Moving to config.

[FeirouzBenGuirat]

Hello,
I have the same problem, my screen is stuck with Kibana is loading. Give me a moment here. I'm loading a whole bunch of code. Don't worry, all this good stuff will be cached up for next time!
I have already installed>
elasticsearch-1.4.4
kibana-4.0.1
logstash-1.5.0.rc2
xampp-win32-5.6.8
and I am working on windows 7

But now I dont know what to do to make it work.
If I run Kibana Server and then http://localhost:9200/_plugin/head/ I get a blank page
and when I run Elasticsearch and then http://localhost:5601/ I am stuck with Kibana is loading.

I read all the comments but I didnt understand anything !
Could you please write me a clear solution????