Elasticsearch
Overview
Elasticsearch is a distributed, open source search and analytics engine for all types of data, including textual, numerical, geospatial, structured, and unstructured. Elasticsearch is built on Apache Lucene and was first released in 2010 by Elasticsearch N.V. (now known as Elastic).
Monitoring Connector assets
Monitored objects
- Databases
- Nodes
- Shards
- Clusters
- Indices
- Documents
- Licences
Discovery rules
- Services
| Rule name | Description |
|---|---|
| App-DB-Elasticsearch-Indice-Statistics | Discover the indices on your Elasticsearch DB |
| App-DB-Elasticsearch-Node-Statistics | Discover the nodes on your Elasticsearch DB |
Use the discovery module to add the monitoring of your Elasticsearch databases, Go to Configuration > Services > Scan to perform a scan.
Collected metrics
- Cluster-statistics
- Indice-statistics
- License
- Node-statistics
| Metric name | Description | Unit |
|---|---|---|
| display | Name of the Elasticsearch cluster | |
| status | Status of the Elasticsearch cluster | |
| nodes_total | Number of nodes | |
| nodes_data | Number of data | |
| nodes_coordinating | Number of nodes coordinating | |
| nodes_master | Number of master nodes | |
| nodes_ingest | Number of ingest data | |
| indices_count | Number of indices | |
| shards_total | Total number of shards | |
| shards_active | Number of active shards | |
| shards_unassigned | Number of unassigned shards | |
| shards_relocating | Number of relocating shards | |
| active_shards_percent | Percentage of active shards | % |
| tasks_pending | Number of pending tasks | |
| docs_count | Number of documents | |
| size_in_bytes | Size of all shards assigned | B |
| Metric name | Description | Unit |
|---|---|---|
| display | Name of the Elasticsearch Indice | |
| status | Status of the Elasticsearch Indice | |
| shards_active | Number of active shards | |
| shards_unassigned | Number of unassigned shards | |
| docs_count | Number of documents | |
| size_in_bytes_primaries | Size of all primary shards | B |
| size_in_bytes_total | Total size of all shards assigned | B |
| Metric name | Description | Unit |
|---|---|---|
| type | Licence type | |
| status | Licence status | |
| issued_to | Licence owner | |
| issue_date | Licence issue date |
| Metric name | Description | Unit |
|---|---|---|
| display | Name of the Elasticsearch node | |
| indices_count | Number of indices in the node | |
| heap_used_percent | Percentage of memory currently in use by the heap | % |
| heap_used_in_bytes | Amount of memory available for use by the heap | B |
| heap_max_in_bytes | Maximum amount of memory available for use by the heap | B |
| available_in_bytes | Total number of bytes available | B |
| total_in_bytes | Total size of all file stores | B |
| docs_count | Number of documents inside the indice | |
| size_in_bytes | Total size of all shards assigned to the node | B |
Prerequisites
In order to monitor an Elasticsearch cluster, it must be prepared acccording to Elasticsearch's official documentation.
In order to be able to communicate with the Elasticsearch node's API, the Centreon Poller should access to the port 9200 with the http protocol on the Elasticsearch node.
Installation
- Online Licence Business & IT Editions
- Offline Licenses
- Install the Plugin on all Centreon pollers monitoring Elasticsearch ressources:
yum install centreon-plugin-Applications-Databases-Elasticsearch
- Install the Monitoring Connector 'Elasticsearch' in the Configuration > Monitoring Connectors Manager page of the Web Centreon interface
- Install the Plugin on all Centreon pollers monitoring Elasticsearch ressources:
yum install centreon-plugin-Applications-Databases-Elasticsearch
- Install the Monitoring Connector RPM on the Centreon server Central:
yum install centreon-pack-applications-databases-elasticsearch
- In the Configuration > Monitoring Connectors Manager page of the Web Centreon interface, install the Monitoring Connector 'Elasticsearch'
Host Configuration
Apply the "App-DB-Elasticsearch-custom" template to your newly created host. Then fill the macros value fileds marked as mandatory below:
| Mandatory | Macro | Description | Default value | Example |
|---|---|---|---|---|
| X | ELASTICSEARCHPORT | The Elasticsearch instance port | 9200 | 1234 |
| X | ELASTICSEARCHPROTO | The Elasticsearch instance protocol | http | https |
| ELASTICSEARCHUSERNAME | The Elasticsearch instance username | centreon | ||
| ELASTICSEARCHPASSWORD | The Elasticsearch instance password | centreon |
FAQ
How to check in the CLI that the configuration is OK and what are the main options for ?
One the Plugin is installed, you can test it directly in the command line from your Centreon poller with the user centreon-engine:
/usr/lib/centreon/plugins/centreon_elasticsearch.pl \
--hostname=168.253.16.125 \
--port=9200 \
--proto=http \
--plugin=database::elasticsearch::restapi::plugin \
--mode=node-statistics \
--filter-name='Node 1' \
--username='Elasticsearch_username' \
--warning-disk-free='20%' \
--critical-disk-free='5%'
Output:
OK: Node 'i-Vertix Node 1' JVM Heap: 26%, Free Disk Space: 1.56TB, Documents: 4362761044, Data: 1.26TB | 'i-Vertix Node 1#node.jvm.heap.usage.percentage'=26%;;;0;100 'i-Vertix Node 1#node.jvm.heap.usage.bytes'=36380302240B;;;0;137151119360 'i-Vertix Node 1#node.disk.free.bytes'=1710072680448B;;;0;3113589145600 'i-Vertix Node 1#node.documents.total.count'=4362761044;;;0; 'i-Vertix Node 1#node.data.size.bytes'=1386278479651B;;;0;
The command request statistic to the Elasticsearch node named 'Node 1' (--mode=node-statistics --filter-name='Node 1) with the IP/FQDN address 168.253.16.125 (--hostname=168.253.16.125).
We will use the port 9200 (--port=9200) and the http protocol (proto=http'').
The username of the datebase is Elasticsearch_username (--username='Elasticsearch_username') and its pasword is Elasticsearch_password(--password='Elasticsearch_password').
This command will trigger a WARNING alert if there is less than 20% of free space on the disk (--warning-disk-free='20%'') and a CRITICAL alert if less than 5% (--critical-disk-free='5%').
Some thresholds can also be set on metrics with options "--warning-" and "--critical-".
All the available modes can be listed with the command line:
/usr/lib/centreon/plugins/centreon_elasticsearch.pl \
--list-mode
And the different mode's options can be displayed with the --help parameter:
/usr/lib/centreon/plugins/centreon_elasticsearch.pl \
--mode=node-statistics \
--help
Why do I get the following error:
UNKNOWN: 500 Can't connect
This error message means that the Centreon Plugin couldn't successfully connect to the host API. Check that no third party device (such as a firewall) is blocking the request. A proxy connection may also be necessary to connect to the API. This can be done by using the --proxyurl option in the command.
UNKNOWN: 501 Protocol scheme 'connect' is not supported
When using a proxy to connect to the host, this error message means that the Centreon Plugin library does not support the proxy connection protocol.
In order to prevent this issue, use the curl HTTP backend by adding the following option to the command: --http-backend='curl'.