Dashboard Guide

Server list

The main page shows all registered servers in a grid. Each card displays:

• Server name and location tag.
• Status indicator: green (healthy), yellow (warning alerts active), red (critical alerts active), gray (no data received in the last 5 minutes).
• Quick stats: current CPU usage, RAM usage, and disk usage as compact bars.
• Last seen: time since the last metrics push.

Click any server card to open the detail view.

Filtering and sorting

Use the controls at the top of the server list:

• Search: filter by server name or location.
• Tags: click a tag to show only servers with that tag. Multiple tags can be combined (AND logic).
• Status: filter by health status (healthy, warning, critical, offline).
• Sort: sort by name, status (worst first), CPU, RAM, or disk usage.

Server detail

The detail view shows comprehensive metrics and alerts for a single server. It is divided into tabs.

Overview tab

The default view with time-series charts for the most important metrics. All charts are expandable; click the expand icon to see a larger view with Y-axis labels.

• CPU: stacked area chart showing user, system, iowait, and steal percentages. Y-axis labeled in CPU %. On servers running Crucible 0.3.0+ with per-core monitoring enabled, the expanded view shows a per-core CPU chart so you can identify uneven load distribution or single-core bottlenecks.
• Memory: area chart showing used, buffers, cache, and available. Swap usage is shown as a separate line if active. Y-axis labeled in GB.
• Disk usage: bar chart for each mount point showing percentage used.
• Network: throughput chart showing RX (inbound) and TX (outbound) traffic per interface. Y-axis labeled in Mbps.

Use the time range selector in the top right to view data for the last hour, 6 hours, 24 hours, 7 days, or 30 days. Click and drag on any chart to zoom into a specific time range.

Disk tab

Detailed disk information:

• Filesystem table: mount point, filesystem type, total size, used, available, and percentage for each mount.
• I/O charts: read/write throughput and IOPS per block device over time.
• Latency chart: average I/O latency per device.
• SMART status: health summary for each physical disk, including temperature, power-on hours, reallocated sectors, and NVMe wear level.

Network tab

Detailed network information:

• Interface table: name, state (up/down), link speed, MAC address, IP addresses.
• Throughput charts: per-interface inbound and outbound bytes over time.
• Packet charts: packets per second, errors, and drops.
• Saturation: current utilization as a percentage of link speed.

Hardware tab

Physical hardware status (available when IPMI or hwmon data is present):

• Temperatures: CPU package, per-core, inlet, exhaust, and drive temperatures over time.
• Fans: RPM readings for each fan sensor.
• Power supplies: status, input voltage, and wattage for each PSU.
• ECC memory: correctable and uncorrectable error counts per DIMM.
• RAID: array status, member disks, and rebuild progress.

Alerts tab

Shows all alerts for this server with three filter tabs:

• Active: currently firing alerts that need attention.
• Acknowledged: alerts that have been acknowledged but not yet resolved. Notifications are silenced.
• All: complete alert history including resolved alerts, filterable by severity, rule, and date range.

Each alert is displayed as a structured card with:

• Priority badge: P1 through P4, indicating urgency.
• One-line summary: what happened and the current value vs. threshold.
• Evidence links: direct links to relevant charts for the time period around the alert.
• Fix commands: copy-pasteable shell commands you can run on the server to investigate or resolve the issue.

Click an active alert to see details and the Acknowledge button. Acknowledging an alert silences notifications for the current occurrence but does not disable the rule.

Muted rules

The alerts tab also shows a list of muted rules for the server. Muted rules do not fire or send notifications. This is useful during maintenance windows or when a known condition is expected. Click the mute/unmute icon next to any rule to toggle it. See the Alert Muting documentation for details.

Alerts view

Accessible from the main navigation, the Alerts view shows all active alerts across all servers in one place. This is your triage screen.

• Alerts are sorted by severity (critical first), then by time (newest first).
• Each row shows the server name, alert rule, current value, threshold, severity, and duration.
• Click a row to jump to the server detail view for that alert.
• Use the bulk acknowledge button to silence multiple alerts at once.

Settings

The Settings page is divided into sections:

API Tokens

Create and manage API tokens for Crucible agents and API access. Each token has a name, creation date, and last-used timestamp. Tokens can be revoked individually.

Channels

Configure notification channels (Email, Telegram, Slack) with per-channel priority filtering. See the Channels page for setup instructions.

Alert Defaults

Set global default thresholds for all alert rules. These apply to any server that does not have a per-server override in its collector.yaml.

Team

Invite team members by email. Each member gets their own login and can view all servers. Roles:

• Owner: full access, can manage billing and delete the account.
• Admin: can manage servers, channels, tokens, and team members.
• Viewer: read-only access to the dashboard and alerts.

Account

Update your email, password, and notification preferences. Enable two-factor authentication (TOTP) for additional security.

Keyboard shortcuts

The dashboard supports keyboard navigation for common actions:

/	Focus the search bar
g then s	Go to server list
g then a	Go to alerts view
g then t	Go to settings
j / k	Navigate up/down in lists
Enter	Open selected item
Esc	Close modal or go back
?	Show shortcut help