Monitoring the state of Elasticsearch clusters and hosts

1. In the management console, go to the folder page and select Managed Service for Elasticsearch.

2. Click the cluster name and open the Monitoring tab.

3. To get started with Yandex Monitoring metrics, dashboards, or alerts, click Open in Monitoring in the top panel.

The page displays the following charts:

• Active shards: Number of active primary shards and the total number of active shards in the cluster.

• Deletion rate: Number of delete operations per second, per host.

• Disk space usage percent: Shows how much disk space is used on each host (in %).

• Flushes: Number of transaction log flush operations per host.

• Health status: Shows the cluster's health and technical condition:

  • 0 (red): Cluster is unhealthy or partially functional. At least one of the primary shards is unavailable. If the cluster responds to queries, the search results will be incomplete.
  • 1 (yellow): Cluster is healthy. There is no access to at least one of the replica shards. The search results in the cluster responses are complete; however, if there are more unavailable shards, cluster performance will be disrupted.
  • 2 (green): Cluster is healthy. All cluster shards are available.

• Indexing rate: Number of indexing operations per second, per host.

• JVM heap: The use of JVM heap memory per host (in bytes).

• JVM heap pressure: The use of a pool of long-lived JVM objects per host (%).

• JVM old collections: The number of garbage collection cycles in the pool of long-lived JVM objects per host.

• JVM young collections: The number of garbage collection cycles in the pool of new JVM objects per host.

• Merges: The number of index segment merges per host.

• Nodes: The number of hosts with the Data node role and the total number of hosts in the cluster.

• Open file descriptors: The number of open file descriptors per host.

• Other shards: Number of inactive shards in each of the following states:

  • Delayed unassigned: Host assignment is delayed.
  • Unassigned: There is no assigned host.
  • Relocating: Moving to another host.
  • Initializing: Being initialized.

• Process CPU: Usage of processor cores on each host due to the JVM Elasticsearch process.

• Query cache: The number of queries in the cache on each host.

• Read bytes: Disk read rate on each host (bytes per second).

• Read operations: Number of read operations per second, per host.

• Refreshes: The number of index segment refresh operations per host.

• Search queries: The number of search queries per second, per host.

• Segments: Number of index segments per host.

• Store size: The size of index storage on disk (in bytes).

• Write bytes: Disk write rate on each host (bytes per second).

• Write operations: Number of write operations per second, per host.

1. In the management console, go to the folder page and select Managed Service for Elasticsearch.
2. Click the cluster name and open the Hosts tab.
3. Select the Monitoring tab.
4. Select the host from the drop-down list.

This page displays charts showing the load on an individual host in the cluster:

• CPU: Load on processor cores. As the load goes up, the Idle value goes down.
• Disk bytes: Speed of disk operations (bytes per second).
• Disk IOPS: Number of disk operations per second.
• Memory: Use of RAM, in bytes. At high loads, the Free value goes down, while the other values go up.
• Network bytes: Speed of data exchange over the network, in bytes per second.
• Network packets: Number of packets exchanged over the network, per second.

1. In the management console, select the folder with the cluster you want to configure alerts for.

2. In the list of services, select  Monitoring.

3. Under Service dashboards, select:

   • Managed Service for Elasticsearch to configure cluster alerts.
   • Managed Service for Elasticsearch — Host Overview to configure host alerts.

4. In the chart you need, click and select Create alert.

5. If the chart shows multiple metrics, select a data query to generate a metric and click Continue. For more information about the query language, see the Yandex Monitoring documentation.

6. Set the Alarm and Warning thresholds to trigger an alert.

7. Click Create alert.

1. Create an alert.
2. Add a status metric.
3. In the alert parameters, set up your alert thresholds.

1. In the management console, go to the folder page and select Managed Service for Elasticsearch.
2. Hover over the indicator in the Availability column in the required cluster row.