github-mirrors/kibana
1
0
Fork 0
mirror of https://github.com/elastic/kibana.git synced 2025-04-24 01:38:56 -04:00
Code Issues Projects Releases Packages Wiki Activity

[Stack Monitoring] Document multi-node ES setup from source (#137548)

Browse source
This commit is contained in:
Mat Schaffer 2022-07-29 19:26:51 +09:00 • committed by GitHub
parent 07f532e926
commit ef5a9f1378
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
1 changed files with 171 additions and 11 deletions
Download patch file Download diff file Expand all files Collapse all files

182
x-pack/plugins/monitoring/dev_docs/how_to/running_components_from_source.md
View file

@ -40,7 +40,7 @@ For multi-cluster tests its best to create a package first:
Then move into the distro path:
```shell
cd build/distribution/local/*
cd "$(ls -1dt build/distribution/local/elasticsearch-* | head -n1)"
```
Then start the server (for example with internal collection enabled):
@ -68,7 +68,7 @@ http://localhost:9200/_security/user/kibana_system/_password -d'{"password": "ch
To start the second server (in another terminal from the same directory), first copy the config and export the new location as `ES_PATH_CONF`
```shell
export ES_PATH_CONF=config2
export ES_PATH_CONF=config-secondary
mkdir -p "${ES_PATH_CONF}"
cp -r config/* "${ES_PATH_CONF}"
```
@ -82,7 +82,7 @@ echo changeme | ./bin/elasticsearch-keystore add xpack.monitoring.exporters.id0.
And finally start the server
```shell
./bin/elasticsearch -E cluster.name=secondary -E http.port=9201 -E transport.port=9301 -E path.data=data2 -E xpack.license.self_generated.type=trial \
./bin/elasticsearch -E cluster.name=secondary -E http.port=9210 -E transport.port=9310 -E path.data=data-secondary -E xpack.license.self_generated.type=trial \
-E xpack.monitoring.collection.enabled=true \
-E xpack.monitoring.exporters.id0.type=http -E xpack.monitoring.exporters.id0.host=http://localhost:9200 \
-E xpack.monitoring.exporters.id0.auth.username=elastic \
@ -93,9 +93,9 @@ You'll likely want to reset the passwords for the secondary cluster as well:
```shell
curl -k -u elastic:'PASSWORD' -H 'Content-Type: application/json' \
http://localhost:9201/_security/user/elastic/_password -d'{"password": "changeme"}'
http://localhost:9210/_security/user/elastic/_password -d'{"password": "changeme"}'
curl -k -u elastic:changeme -H 'Content-Type: application/json' \
http://localhost:9201/_security/user/kibana_system/_password -d'{"password": "changeme"}'
http://localhost:9210/_security/user/kibana_system/_password -d'{"password": "changeme"}'
```
For metricbeat collection, omit the monitoring settings, provide both cluster hosts to the elasticsearch metricbeat module config (see [metricbeat](#metricbeat) section below), and remove the exporter password from the keychain:
@ -110,14 +110,14 @@ Once you have two clusters going you can use something like this to configure th
```
curl -u elastic:changeme -H 'Content-Type: application/json' \
-XPUT -d'{"persistent": {"cluster": {"remote": {"secondary": {"seeds": ["127.0.0.1:9301"]}}}}}' \
-XPUT -d'{"persistent": {"cluster": {"remote": {"secondary": {"seeds": ["127.0.0.1:9310"]}}}}}' \
http://localhost:9200/_cluster/settings
```
Create an index on the secondary cluster:
```
curl -XPOST -H'Content-Type: application/json' -d'{"some": "stuff"}' -u elastic:changeme http://localhost:9201/stuff/_doc
curl -XPOST -H'Content-Type: application/json' -d'{"some": "stuff"}' -u elastic:changeme http://localhost:9210/stuff/_doc
```
Then use the "Cross-Cluster Replication" kibana UI to set up a follower index (`stuff-replica`) in the main cluster.
@ -136,6 +136,166 @@ You can create job using the machine learning UI in kibana. Select (or create) a
Once the job is created push the "Start job running in real time". This will help exercise the active-job state in Stack Monitoring UI.
### Multi-node configuration
This is a little trickier, but possible.
As noted [in the public docs](https://www.elastic.co/guide/en/elasticsearch/reference/master/modules-discovery-bootstrap-cluster.html) be careful about the presence of any old `data*` or `config*` directories. The commands below should work on the first run, but if you have previously started any other clusters with the same paths, they might fail. Elasticsearch won't start if it detects the possibility of data loss due to incorrect cluster bootstrapping.
Here's an example setting up a 3-node `main` cluster across 3 terminal windows, assuming you've already started an initial cluster and set `xpack.security.http.ssl` set to false in the config as above.
Note that the first two nodes will pause after initial setup, waiting for the last one to become available.
In the first terminal run this for node main-0:
```
export ES_PATH_CONF=config-main-0
mkdir -p "${ES_PATH_CONF}"
cp -r config/* "${ES_PATH_CONF}"
./bin/elasticsearch -E cluster.name=main -Enode.name=main-0 \
-E http.port=9200 -E transport.port=9300 -E path.data=data-main-0 \
-Ediscovery.seed_hosts=127.0.0.1:9300,127.0.0.1:9301,127.0.0.1:9302 \
-Ecluster.initial_master_nodes=127.0.0.1:9300,127.0.0.1:9301,127.0.0.1:9302
```
This will hold waiting for a second node.
In the second terminal run this for node main-1:
```
export ES_PATH_CONF=config-main-1
mkdir -p "${ES_PATH_CONF}"
cp -r config/* "${ES_PATH_CONF}"
./bin/elasticsearch -E cluster.name=main -Enode.name=main-1 \
-E http.port=9201 -E transport.port=9301 -E path.data=data-main-1 \
-Ediscovery.seed_hosts=127.0.0.1:9300,127.0.0.1:9301,127.0.0.1:9302 \
-Ecluster.initial_master_nodes=127.0.0.1:9300,127.0.0.1:9301,127.0.0.1:9302
```
At this point, the 2 nodes will elect a master which will print out the password for the `elastic` user. You can use this to then reset the password on any of the nodes, the same as above, like so:
```
curl -u elastic:'PASSWORD' -H 'Content-Type: application/json' \
http://localhost:9200/_security/user/elastic/_password -d'{"password": "changeme"}'
curl -u elastic:changeme -H 'Content-Type: application/json' \
http://localhost:9200/_security/user/kibana_system/_password -d'{"password": "changeme"}'
```
Then in the third terminal run this for node main-2:
```
export ES_PATH_CONF=config-main-2
mkdir -p "${ES_PATH_CONF}"
cp -r config/* "${ES_PATH_CONF}"
./bin/elasticsearch -E cluster.name=main -Enode.name=main-2 \
-E http.port=9202 -E transport.port=9302 -E path.data=data-main-2 \
-Ediscovery.seed_hosts=127.0.0.1:9300,127.0.0.1:9301,127.0.0.1:9302 \
-Ecluster.initial_master_nodes=127.0.0.1:9300,127.0.0.1:9301,127.0.0.1:9302
```
One the third node is booted you should be able to see all 3 in the nodes list of any instance. You can confirm this with:
```
curl -u elastic:changeme localhost:9200/_cat/nodes
```
Starting a 3 node secondary cluster is a similar process, included here for easy copy-paste.
In a terminal for secondary-0 run:
```
export ES_PATH_CONF=config-secondary-0
mkdir -p "${ES_PATH_CONF}"
cp -r config/* "${ES_PATH_CONF}"
./bin/elasticsearch -E cluster.name=secondary -Enode.name=secondary-0 \
-E http.port=9210 -E transport.port=9310 -E path.data=data-secondary-0 \
-Ediscovery.seed_hosts=127.0.0.1:9310,127.0.0.1:9311,127.0.0.1:9312 \
-Ecluster.initial_master_nodes=127.0.0.1:9310,127.0.0.1:9311,127.0.0.1:9312
```
For secondary-1 run:
```
export ES_PATH_CONF=config-secondary-1
mkdir -p "${ES_PATH_CONF}"
cp -r config/* "${ES_PATH_CONF}"
./bin/elasticsearch -E cluster.name=secondary -Enode.name=secondary-1 \
-E http.port=9211 -E transport.port=9311 -E path.data=data-secondary-1 \
-Ediscovery.seed_hosts=127.0.0.1:9310,127.0.0.1:9311,127.0.0.1:9312 \
-Ecluster.initial_master_nodes=127.0.0.1:9310,127.0.0.1:9311,127.0.0.1:9312
```
At this point you can reset the password via port 9210:
```
curl -u elastic:'PASSWORD' -H 'Content-Type: application/json' \
http://localhost:9210/_security/user/elastic/_password -d'{"password": "changeme"}'
curl -u elastic:changeme -H 'Content-Type: application/json' \
http://localhost:9210/_security/user/kibana_system/_password -d'{"password": "changeme"}'
```
Then for secondary-2 run:
```
export ES_PATH_CONF=config-secondary-2
mkdir -p "${ES_PATH_CONF}"
cp -r config/* "${ES_PATH_CONF}"
./bin/elasticsearch -E cluster.name=secondary -Enode.name=secondary-2 \
-E http.port=9212 -E transport.port=9312 -E path.data=data-secondary-2 \
-Ediscovery.seed_hosts=127.0.0.1:9310,127.0.0.1:9311,127.0.0.1:9312 \
-Ecluster.initial_master_nodes=127.0.0.1:9310,127.0.0.1:9311,127.0.0.1:9312
```
Note that all 6 notes will need to be in the metricbeat config if you want to run the Stack Monitoring UI as well. Here's an example `metricbeat.multinode.yaml` you can use as a starting point:
```yaml
http.enabled: true
metricbeat.modules:
- module: system
- module: elasticsearch
xpack.enabled: true
period: 10s
hosts:
# main
- "localhost:9200"
- "localhost:9201"
- "localhost:9202"
# secondary
- "localhost:9210"
- "localhost:9211"
- "localhost:9212"
username: "elastic"
password: "changeme"
- module: kibana
xpack.enabled: true
basepath: "/ftw"
period: 10s
hosts: [ "localhost:5601" ]
username: "elastic"
password: "changeme"
- module: beat
xpack.enabled: true
period: 10s
hosts:
# filebeat
- "http://localhost:5068"
output.elasticsearch:
hosts: [ "localhost:9200" ]
username: "elastic"
password: "changeme"
```
## Kibana
See the [local setup](local_setup.md) guide for running from source.
@ -170,7 +330,7 @@ metricbeat.modules:
# main
- "localhost:9200"
# secondary
- "localhost:9201"
- "localhost:9210"
username: "elastic"
password: "changeme"
@ -265,8 +425,8 @@ filebeat.modules:
server:
enabled: true
var.paths:
- ../../elasticsearch/build/distribution/local/elasticsearch-8.1.0-SNAPSHOT/logs/*.log
- ../../elasticsearch/build/distribution/local/elasticsearch-8.1.0-SNAPSHOT/logs/*_server.json
- ../../elasticsearch/build/distribution/local/elasticsearch-*-SNAPSHOT/logs/*.log
- ../../elasticsearch/build/distribution/local/elasticsearch-*-SNAPSHOT/logs/*_server.json
output.elasticsearch:
hosts: [ "http://localhost:9200" ]
@ -424,4 +584,4 @@ Once you open the enterprise search UI, you can create the demo engine to get th
### External metricbeat
WIP - this should be possible now that https://github.com/elastic/kibana/issues/121975 is resolved, but steps haven't been written yet.
WIP - this should be possible now that https://github.com/elastic/kibana/issues/121975 is resolved, but steps haven't been written yet.