Health always yellow status

Elastic Stack Elasticsearch
1

Hi,

I start the ES with bellow config.

node.master: true
node.data: true
index.number_of_shards: 5
index.number_of_replicas: 1

index:
analysis:
analyzer:
default:
tokenizer: lowercase
filter: [asciifolding,standard]
bootstrap.mlockall: true
http.port: 9200
http.max_content_length: 100mb

Disable HTTP completely:

http.enabled: true

First, when I start the ES with no data, the status is green.
But after a while, when I index my data, it go to yellow.

I have try to restart .... etc, but it's still yellow.

What should I do to make it better ?

The Server is: 24 core, 32Gb Ram. 25 Tb Free Disk Space.

Thanks.

--

2

Yellow status means : everything is OK but you don't satisfy replica rules. That
means that if one of your nodes disappear, you will lost some documents.

You asked for 1 replica (default). But you run on a single node, so shards are
not replicated.

If this status bother you, you can:

  • set replica to 0 (you can do it live)
  • start a new node

David

Le 6 octobre 2012 à 09:52, kidkid zkidkid@gmail.com a écrit :

Hi,

I start the ES with bellow config.

node.master: true
node.data: true
index.number_of_shards: 5
index.number_of_replicas: 1

index:
analysis:
analyzer:
default:
tokenizer: lowercase
filter: [asciifolding,standard]
bootstrap.mlockall: true
http.port: 9200
http.max_content_length: 100mb

Disable HTTP completely:

http.enabled: true

First, when I start the ES with no data, the status is green.
But after a while, when I index my data, it go to yellow.

I have try to restart .... etc, but it's still yellow.

What should I do to make it better ?

The Server is: 24 core, 32Gb Ram. 25 Tb Free Disk Space.

Thanks.

--

--
David Pilato
http://www.scrutmydocs.org/
http://dev.david.pilato.fr/
Twitter : @dadoonet / @elasticsearchfr / @scrutmydocs

--

3

Hi,

I read the document and see that:

Replicas is a copy of shard.
In my setting, it's 1, mean I will have 1 replica for each of 5 shard.

The more replicas I have, the fast my ES could search.

it don't relate to node.

On Saturday, October 6, 2012 4:10:45 PM UTC+7, David Pilato wrote:

Yellow status means : everything is OK but you don't satisfy replica
rules. That means that if one of your nodes disappear, you will lost some
documents.

You asked for 1 replica (default). But you run on a single node, so
shards are not replicated.

If this status bother you, you can:

  • set replica to 0 (you can do it live)
  • start a new node

David

Le 6 octobre 2012 à 09:52, kidkid <zki...@gmail.com <javascript:>> a
écrit :

Hi,

I start the ES with bellow config.

node.master: true
node.data: true
index.number_of_shards: 5
index.number_of_replicas: 1

index:
analysis:
analyzer:
default:
tokenizer: lowercase
filter: [asciifolding,standard]
bootstrap.mlockall: true
http.port: 9200
http.max_content_length: 100mb

Disable HTTP completely:

http.enabled: true

First, when I start the ES with no data, the status is green.
But after a while, when I index my data, it go to yellow.

I have try to restart .... etc, but it's still yellow.

What should I do to make it better ?

The Server is: 24 core, 32Gb Ram. 25 Tb Free Disk Space.

Thanks.

--

--
David Pilato
http://www.scrutmydocs.org/
http://dev.david.pilato.fr/
Twitter : @dadoonet / @elasticsearchfr / @scrutmydocs

--

4

I don't understand your answer.

May be you can have look at my devoxx slides: Elasticsearch - Devoxx France 2012 - English version | PPT

At the end, I explained that.
There are also nice talks here Elasticsearch Platform — Find real-time answers at scale | Elastic

--
David :wink:
Twitter : @dadoonet / @elasticsearchfr / @scrutmydocs

Le 7 oct. 2012 à 09:48, kidkid zkidkid@gmail.com a écrit :

Hi,

I read the document and see that:

Replicas is a copy of shard.
In my setting, it's 1, mean I will have 1 replica for each of 5 shard.

The more replicas I have, the fast my ES could search.

it don't relate to node.

On Saturday, October 6, 2012 4:10:45 PM UTC+7, David Pilato wrote:

Yellow status means : everything is OK but you don't satisfy replica rules. That means that if one of your nodes disappear, you will lost some documents.

You asked for 1 replica (default). But you run on a single node, so shards are not replicated.

If this status bother you, you can:

  • set replica to 0 (you can do it live)
  • start a new node

David

Le 6 octobre 2012 à 09:52, kidkid zki...@gmail.com a écrit :

Hi,

I start the ES with bellow config.

node.master: true
node.data: true
index.number_of_shards: 5
index.number_of_replicas: 1

index:
analysis:
analyzer:
default:
tokenizer: lowercase
filter: [asciifolding,standard]
bootstrap.mlockall: true
http.port: 9200
http.max_content_length: 100mb

Disable HTTP completely:

http.enabled: true

First, when I start the ES with no data, the status is green.
But after a while, when I index my data, it go to yellow.

I have try to restart .... etc, but it's still yellow.

What should I do to make it better ?

The Server is: 24 core, 32Gb Ram. 25 Tb Free Disk Space.

Thanks.

--

--
David Pilato
http://www.scrutmydocs.org/
http://dev.david.pilato.fr/
Twitter : @dadoonet / @elasticsearchfr / @scrutmydocs

--

--