How to View Server Metrics

Monitor CPU, memory, disk, load, processes, network interfaces, and storage health for any registered server.

Opening the Server Detail Page

From the Servers list, click the server's name or the chart icon on its row.

Header Summary Card

The top card shows essential information about the server:

Field	Description
Server Name	The name you assigned in Pinguzo
Status	Online or Offline
Hostname	System hostname reported by the agent
IP Address	Primary network interface IP
OS	Linux distribution and version (e.g., Ubuntu 22.04 LTS)
Kernel	Kernel version and architecture (e.g., 5.15.0 x86_64)
SMART Health	Disk health: Passed, Failed, or N/A
Last Seen	Timestamp of the most recent agent check-in
Agent Version	Version string of the installed Pinguzo Agent

Time Range Selector

All charts respond to the time range picker at the top of the page:

1 hour — minute-by-minute resolution, ideal for real-time troubleshooting
24 hours — default view; good daily overview
7 days — weekly trend for capacity planning
30 days — monthly view; useful for identifying growth patterns
Custom — specify any date range

Page Tabs

The server detail page is organized into five tabs:

Overview

Processes

Network

Disk

Overview Tab (default)

Contains the primary resource utilization charts. This is the first place to look when investigating a performance issue or an alert.

CPU Usage (%)

A line chart of overall CPU utilization across all cores. Values above 80% sustained for several minutes typically indicate a problem worth investigating in the Processes tab.

CPU Steal (%)

On cloud virtual machines (AWS, GCP, DigitalOcean, etc.), steal represents the percentage of time the hypervisor was busy with other tenants when your VM needed CPU time. High steal (above 10%) indicates your instance may be on an oversubscribed host.

Memory Usage (%)

RAM utilization as a percentage of total installed memory. This reflects actual used memory; buffer/cache memory used by the OS is not counted as "used." Values near 100% with heavy swapping indicate memory pressure.

Disk Usage (%)

The highest disk usage percentage across all mounted partitions. If you have multiple disks, the chart shows the worst case. Switch to the Storage tab to see each partition individually.

Load Average

A combo chart showing the 1-minute, 5-minute, and 15-minute load averages. A load average equal to the number of CPU cores means the CPU is fully utilized. Significantly higher values indicate queuing. The rule of thumb:

Load / cores < 1.0 → healthy
Load / cores 1.0–2.0 → moderate pressure
Load / cores > 2.0 → potential bottleneck

Load per core metric When configuring alert policies for load, Pinguzo uses load per core (5-minute average ÷ CPU core count) as the threshold metric. This makes the same threshold usable across servers with different core counts.

Processes Tab

Shows a snapshot of running processes, updated each time the agent checks in (every minute).

Top Processes by CPU

Lists the top processes sorted by CPU usage percentage at the time of the selected snapshot. Columns:

Column	Description
PID	Process ID
Process Name	Command name (e.g., `nginx`, `mysqld`, `python3`)
CPU %	CPU usage percentage at snapshot time
Memory %	Memory usage percentage at snapshot time

Top Processes by Memory

Same table sorted by memory usage — useful for identifying memory leaks or runaway processes.

Tip: Use time range + processes together If a CPU spike appears in the Overview tab, navigate to the Processes tab and select that time range to see which process was responsible at that moment.

Network Tab

Per-interface network statistics collected by the agent:

Column	Description
Interface	Network interface name (e.g., `eth0`, `ens3`)
Bytes In	Total bytes received
Bytes Out	Total bytes transmitted
Packets In	Total packets received
Packets Out	Total packets transmitted
Errors In	Receive errors (indicates hardware or driver issues)
Errors Out	Transmit errors
Dropped In	Packets dropped on receive (may indicate buffer overflow)

Disk Tab

Detailed per-partition disk usage and SMART health data:

Partition Usage Table

Column	Description
Device	Block device name (e.g., `/dev/sda1`)
Mount Point	Where the partition is mounted (e.g., `/`, `/var`)
Filesystem	Filesystem type (ext4, xfs, btrfs, etc.)
Total Size	Partition capacity in GiB
Used	Used space in GiB
Available	Free space in GiB
Usage %	Percentage of space used; color-coded green/yellow/red
Inodes Used	Inodes usage
Inodes %	Percentage of inodes used

SMART Health

If the smartctl utility is installed on the monitored server, the Storage tab shows the SMART self-assessment for each physical disk:

SMART: Passed — disk reports no errors
SMART: Failed — disk has encountered errors; back up data immediately
N/A — smartctl not installed or disk does not support SMART

Install smartctl for SMART monitoring On Debian/Ubuntu: apt install smartmontools
On RHEL/CentOS/Rocky: dnf install smartmontools
After installation, the agent will automatically pick it up on the next check-in.

Next Steps

Configure Alerts — get notified when CPU, memory, or disk exceeds thresholds
View Incidents — see all server incidents including offline events
Install Pinguzo Agent — troubleshoot agent installation or update to the latest version