View etcd Database Status

Time updated：2025-12-22 09:26:17

This section describes how to view the etcd database status.

Prerequisites

You need to join a cluster and have the Monitoring View permission in the cluster. For more information, see Cluster Members and Cluster Roles.

The WizTelemetry Monitoring extension must be installed and enabled on the KubeSphere platform.

Steps

Log in to the KubeSphere web console with a user who has Monitoring View permissions and enter your cluster.
Click Monitoring and Alerting > Cluster Monitoring in the left navigation pane.

On the Cluster Monitoring page, click the etcd Monitoring tab to view the running status of the etcd database.

The etcd Monitoring tab provides the following information:

Parameter	Description
Service Status	Displays the Leader node of the etcd cluster, the IP address of each node, and the number of Leader changes within the last hour.
Database Size	The size of the etcd database within the specified time range.
Client Traffic	Displays the data traffic sent to and received from gRPC clients.
gRPC Stream Messages	Displays the number of gRPC stream messages received and sent per second on the server side.
WAL Fsync Duration	Displays the latency of WAL calling fsync. Before applying log entries, etcd calls wal_fsync when persisting log entries to disk.
DB Fsync Duration	Displays the distribution of backend commit latency. When etcd commits its latest incremental snapshot to disk, it calls backend_commit. High disk operation latency (long WAL fsync duration or DB fsync duration) usually indicates a disk problem, which may lead to high request latency or cluster instability.
Raft Proposals	Displays the current number of Raft proposals per second for etcd. Hover over the line chart to view the proposal data at a specific point in time. Proposal Commit Rate: The commit rate of consensus proposals. If the cluster is healthy, this metric should increase over time. Several healthy members of an etcd cluster can have different general proposals simultaneously. A sustained large lag between a single member and its Leader indicates that the member is slow or unhealthy. Proposal Apply Rate: The total application rate of consensus proposals. The etcd server applies each committed proposal asynchronously. The difference between the proposal commit rate and the proposal apply rate should be small (only a few thousand even under high load). If the difference between them continues to increase, it indicates that the etcd server is overloaded. This may occur when using large-scale queries such as large range queries or a large number of txn operations. Proposal Failure Rate: The total rate of proposal failures. This value is affected by two factors: temporary failures related to Leader elections; and prolonged downtime due to the number of cluster members falling below the required number. Proposals Pending: The current number of pending proposals. An increase in pending proposals indicates high client load or that members are unable to commit proposals.

Parameter

Description

Service Status

Displays the Leader node of the etcd cluster, the IP address of each node, and the number of Leader changes within the last hour.

Database Size

The size of the etcd database within the specified time range.

Client Traffic

Displays the data traffic sent to and received from gRPC clients.

gRPC Stream Messages

Displays the number of gRPC stream messages received and sent per second on the server side.

WAL Fsync Duration

Displays the latency of WAL calling fsync. Before applying log entries, etcd calls wal_fsync when persisting log entries to disk.

DB Fsync Duration

Displays the distribution of backend commit latency. When etcd commits its latest incremental snapshot to disk, it calls backend_commit. High disk operation latency (long WAL fsync duration or DB fsync duration) usually indicates a disk problem, which may lead to high request latency or cluster instability.

Raft Proposals

Displays the current number of Raft proposals per second for etcd. Hover over the line chart to view the proposal data at a specific point in time.

Proposal Commit Rate: The commit rate of consensus proposals. If the cluster is healthy, this metric should increase over time. Several healthy members of an etcd cluster can have different general proposals simultaneously. A sustained large lag between a single member and its Leader indicates that the member is slow or unhealthy.
Proposal Apply Rate: The total application rate of consensus proposals. The etcd server applies each committed proposal asynchronously. The difference between the proposal commit rate and the proposal apply rate should be small (only a few thousand even under high load). If the difference between them continues to increase, it indicates that the etcd server is overloaded. This may occur when using large-scale queries such as large range queries or a large number of txn operations.
Proposal Failure Rate: The total rate of proposal failures. This value is affected by two factors: temporary failures related to Leader elections; and prolonged downtime due to the number of cluster members falling below the required number.
Proposals Pending: The current number of pending proposals. An increase in pending proposals indicates high client load or that members are unable to commit proposals.

For more information about etcd database parameters, refer to the etcd official documentation.

Click in the upper right corner to set the time range.

Click / in the upper right corner to enable/disable real-time data refresh.

Click in the upper right corner to refresh data manually.