Dashboard

Available in VPC

In the Dashboard menu, you can view the detailed status (Health, Status) of the cluster created by the Search Engine Service through the detailed cluster dashboards, and information such as indexes, Documents, and storage capacity being used. Depending on the Cluster Health, you can use Kibana or OpenSearch Dashboards to determine the cause of the cluster's problems.

Dashboard page

The basic description of the Dashboard menu for using the Search Engine Service is as follows:
ses-dashboard-vpc_screen_ko

Field	Description
① Menu name	Name of the menu currently being viewed
② Basic features	Create clusters (refer to Create cluster), view Search Engine Service details, refresh page
③ List	list of clusters held

Detailed cluster dashboard

If you click a cluster in the cluster list, you can check the detailed dashboard of that cluster. The following describes the cluster's dashboard details.
ses-dashboard-vpc_details_ko

Field	Description
① List of nodes	List of nodes in the selected cluster
② Cluster information	Cluster Health: cluster status (refer to Cluster status for a description of each status) Number of Indices: number of indices in the cluster Number of Shards: number of shards in the cluster Number of Document: number of documents in the cluster Number of Segments: number of segments in the cluster Disk Used: disk usage
③ Index information	Health: display index status with colors (green, yellow, red) Status: index status Index: index name Primary: number of primary shards storing indexes Replica: number of replica shards storing indexes Document: number of documents in index Deleted Document: number of documents deleted Store size: data capacity of the index

Cluster status

The Search Engine Service expresses the Cluster Health in 4 ways: running, warning, error, unknown. For each status, check the description of the status and how to solve the problem.

Note

Use Kibana or OpenSearch Dashboards' Console to identify and resolve the cause of cluster problems. For more information on how to use Kibana and OpenSearch Dashboards, see Use Kibana and OpenSearch Dashboards.

Running

The cluster is functioning normally without any problems.

Warning

The Primary Shard is correctly allocated to the data node, but the Replica Shard is not allocated correctly.

In most cases, the Replica Shard is assigned to the data node after a certain period, and the cluster status changes to Running.
If the status of the cluster does not change to Running after some time, check the following:
- You can run the following command to determine the index of the yellow status, the unallocated shard, and the cause of the unallocated shard.
```
GET _cat/indices
GET _cat/shards
GET _cluster/allocation/explain
```
- If the disk usage of all data nodes is 85% or above, then shards can't be allocated.
  - You can check the disk usage of a node in the Monitoring menu or by running the following command:
```
GET _cat/allocation?v
```

Note

You can receive a notification when the disk usage exceeds a set threshold through the Cloud Insight's event rule settings (refer to Monitoring conditions and notification settings).

Error

This happens when neither the Primary Shard nor the Replica Shard is allocated to a data node.

You can find the index in the red status and the reason for unallocated shards by running the following command:
```
GET _cat/indices
GET _cluster/allocation/explain
```
You can try to restore the index in the red status by running the following command:
- Restoration may fail.
- If the restoration fails, then you must delete the index before you can restore the cluster to its normal operating status.
```
POST _cluster/reroute?retry_failed=true
```

Unknown

This is what happens when the Search Engine Process is terminated or there is a temporary network error.

This is usually due to Out Of Memory (OOM).
If you can access Kibana or OpenSearch Dashboards, run the following command in the Console to check whether the manager node and data node are output normally.
```
GET _cat/nodes
```
If you cannot access Kibana or OpenSearch Dashboards, restart the cluster (refer to Restart cluster).