Cluster Dashboard

The CLUSTER dashboard focuses on replication health and comparative performance across cluster members. To view the cluster dashboard, select a cluster name in the cluster navigator.

If you have configured AI for your Workbench, the dashboard displays the AI Overview just below the header. The overview provides a concise, AI-powered summary of your object's health, with notes about history and recommended actions to maintain system health.

Tiles below the heading of the CLUSTER dashboard provide an at-a-glance overview of the state of your cluster:

• The OK tile (green) shows the count of servers that conform to all configured alert thresholds.
• The WARNING tile (orange) shows the count of servers with one or more active threshold violations.
• The OFFLINE tile (red) shows the count of servers that the collector cannot reach.

Below the status tiles, the Event Timeline displays a timeline with indicators that show monitored events that have occurred across the monitored servers. See Event Timeline for details about using the time range selector, event type filters, and reviewing event details.

Tiles below the event timeline provide a quick glance into the performance of your selected estate, server, or cluster. Hover over a chart or graph to review detailed information about a specific point in time for the selected metric.

The Workbench displays the following tiles:

• The XID AGE tile shows the transaction ID age for the cluster.
• The CACHE HIT RATIO tile shows the buffer cache hit ratio across the cluster.
• The TRANSACTIONS tile shows the transaction rate for the cluster.
• The CHECKPOINTS tile shows the checkpoint activity for the cluster.

Reviewing Active Alerts

The Active Alerts pane shows the alerts that are currently active across the cluster.

See Using Alerts for details on reviewing, acknowledging, and analyzing alerts, and for how to find an alert's acknowledgment reason or a past alert from the Event Timeline.

Reviewing Cluster Topology

The Topology pane renders an interactive diagram showing servers as nodes with color-coded replication edges. Each edge represents a replication relationship between two servers.

A colored dot on each node indicates server status; a green dot marks an online server. Each node tile displays a label that identifies the server.

The diagram uses the following color scheme for edges:

• Physical and streaming replication edges display in blue.
• Spock replication edges display in orange.
• Logical replication edges display in green.

Edge labels display the replication type so you can distinguish between different replication methods at a glance.

Monitoring the Cluster

The Monitoring pane presents replication health and comparative performance data for the cluster.

Reviewing Replication Lag

The Replication Lag pane tracks replication lag over the selected time range for the replication relationships in the cluster. Three tiles at the top of the panel present the current lag values in milliseconds; a time-series chart below the tiles plots the same metrics over the selected time range.

The pane displays the following tiles:

• The WRITE LAG tile shows the current write lag for the cluster in milliseconds.
• The FLUSH LAG tile shows the current flush lag for the cluster in milliseconds.
• The REPLAY LAG tile shows the current replay lag for the cluster in milliseconds.

The Replication Lag Over Time chart plots the write, flush, and replay lag metrics against a time axis; a legend identifies each line by color and label.

When the Workbench detects no primary server, the graphic displays the message No primary server detected in this cluster.

Reviewing Comparative Metrics

The Comparative Metrics pane presents side-by-side metrics for all servers
in the cluster; use the pane to identify performance disparities between cluster members. The pane arranges four bar charts in a grid, and each chart plots one bar per server, labeled by server name along the x-axis.

The pane displays the following charts:

• The Transaction Rate (commits/sec) chart shows the commit rate for each server in the cluster.
• The Cache Hit Ratio (%) chart shows the buffer cache hit ratio for each server in the cluster.
• The Rollback Rate (%) chart shows the transaction rollback rate for each server in the cluster.
• The Connection Count chart shows the number of connections for each server in the cluster.

Hover over a bar to display a tooltip with the server name and the metric's value for that server. Click a server entry to navigate to the server dashboard for that server.

Reviewing Cluster Settings

Each cluster or cluster node displays a gear icon when you hover over the object name; click the gear icon to open the Cluster Settings dialog. The dialog organizes its settings into a horizontal tab bar, and Cancel and Save buttons at the bottom discard or retain your changes.

DETAILS Tab

The DETAILS tab presents a form that identifies the cluster and defines its replication behavior. The tab includes the following fields:

• The Name field displays or modifies the display name for the cluster.
• The Description field is a multi-line text area that holds optional notes about the cluster.
• The Replication Type dropdown specifies the replication technology used for the cluster.

TOPOLOGY Tab

The TOPOLOGY tab presents a visual diagram of the cluster members and lets you assign servers and define replication relationships. The diagram displays each member node as a tile with a status dot and a role badge that identifies the node's role in the cluster.

Use the ADD SERVER section to add a new node to the cluster:

• Use the Server dropdown to search for and select an unassigned server.
• Use the Role dropdown to set the role the server will hold in the cluster.
• Use the + Add button to add the selected server to the cluster.

A list of currently assigned servers appears below the ADD SERVER section. Server details display the server name, a role badge, its host and port, and a Delete (trash) icon. Select the Delete icon to remove the server from the cluster.

The RELATIONSHIPS section at the bottom of the tab shows the replication relationships the topology diagram presents. Use the section's controls to define a new relationship between two cluster members:

• Use the Source dropdown to select the source node for the relationship.
• Use the Target dropdown to select the target node for the relationship.
• Use the Type dropdown to select the replication type, such as "Replicates with (Spock)".
• Click + Add to create the relationship between the selected nodes.

ALERT OVERRIDES Tab

The ALERT OVERRIDES tab lets you tailor the alert rules for the selected cluster. A table lists the current rules and settings; each row describes one alert rule and its threshold. The table contains the following columns:

• Name identifies the alert rule.
• Metric names the metric the rule monitors.
• Condition specifies the threshold that triggers the alert.
• Severity indicates the alert level, such as "warning".
• Enabled provides a toggle that activates or deactivates the rule for the selected cluster.
• Actions provides an edit (pencil) icon that opens the rule for adjustment.

The Workbench groups the rows under category headers such as AVAILABILITY, CONNECTIONS, and LOCKS; additional categories appear as you scroll through the table. See Alert Rules for the full list of built-in rules and their default thresholds.

PROBE CONFIGURATION Tab

The PROBE CONFIGURATION tab controls the probes that collect metrics for the selected cluster. A table lists the current probes and their settings; each row describes one probe and its configuration. The table contains the following columns:

• Name identifies the probe.
• Description explains what the probe monitors.
• Enabled provides a toggle that activates or deactivates the probe for the selected cluster.
• Interval specifies how often the probe collects data, in seconds.
• Retention specifies how long the Workbench retains the probe's collected data.
• Actions provides an edit (pencil) icon that opens the probe for adjustment.

The table scrolls to reveal additional probes below the visible rows. See Probe Management for the full list of built-in probes and their scopes.

NOTIFICATION CHANNELS Tab

The NOTIFICATION CHANNELS tab manages the channels that deliver alert notifications for the selected cluster. A table lists the available channels and their settings; each row describes one channel and its current override state. The table contains the following columns:

• Name identifies the notification channel.
• Type shows the channel type, such as email, Slack, Mattermost, or webhook.
• Description shows the channel's optional description.
• Estate Default indicates whether the channel applies to all servers or clusters by default.
• Enabled provides a toggle that activates or deactivates the channel for the selected cluster, overriding the estate default.
• Actions provides controls for managing the channel's override for the selected cluster.

If you have not configured any notification channels, the tab displays the empty state "No notification channels found." See Notification Channels for the full list of supported channel types and how to configure them.