Health Monitoring

Health Monitoring gives operations and platform teams continuous visibility into server CPU, memory, disk, disk I/O, and agent status so they can catch resource bottlenecks before they become outages.

When To Use It

• Use Health Monitoring when a server, virtual machine, or supported host needs resource-level monitoring.
• Use Endpoint Monitoring when the user-facing risk is whether a URL or application endpoint responds successfully.
• Use Windows Event Monitoring when the risk is a Windows event log, event ID, event source, or severity pattern.
• Use Certificate, DNS, and Domain Monitoring when the risk is certificate expiration, DNS behavior, or domain status rather than host-resource pressure.

Supported Systems

During private beta, Health Monitoring installs on supported Linux and Windows Server hosts. macOS is supported for CLI workflows that manage Health Monitoring, but it is not a monitored-host target.

Prerequisites

• A Bitaic workspace role that can register and manage Health Monitoring hosts.
• A supported Linux or Windows Server host where the Health Monitoring agent can be installed.
• Local administrator rights for installation or upgrades, plus a dedicated runtime service identity for ongoing collection.
• Outbound HTTPS access from the host to Bitaic API and ingestion endpoints.
• Notification routing for CPU, memory, disk-used, or disk I/O utilization threshold alerts.

What Bitaic Checks

• CPU utilization as health.cpu.utilization_percent.
• Memory utilization as health.memory.utilization_percent.
• Disk usage as health.disk.used_percent.
• Disk I/O throughput as health.disk.read_bytes_per_second and health.disk.write_bytes_per_second.
• Disk I/O utilization as health.disk.io_utilization_percent.
• Agent heartbeat age as health.agent.heartbeat_age_seconds.
• Threshold state for CPU utilization, memory utilization, disk used, and disk I/O utilization.
• Host identity, telemetry freshness, and agent status.
• Real-time and historical health metrics in the Bitaic dashboard.

Agent Runtime And Freshness

Threshold Baseline

Launch Health thresholds use whole percentage values for the percentage metrics collected by the agent. Disk read and write byte-throughput metrics are available for analysis, but they do not have default launch thresholds.

Each threshold object has warning and critical fields. Custom values must be whole percentages from 1 through 100, supplied as a pair, with critical greater than warning. Threshold state is clear, warning, critical, or unknown; stale or missing telemetry evaluates as unknown. The dashboard is the primary threshold-editing surface, with API updates and config-file seeding available for automation.

Configuration Example

health-agent-config.yml
agent:
  collection_interval_seconds: 60
  local_buffer_hours: 24
metrics:
  cpu_usage: true
  memory_usage: true
  disk_usage: true
  disk_io: true
thresholds:
  cpu_utilization_percent:
    warning: 80
    critical: 95
  memory_utilization_percent:
    warning: 85
    critical: 95
  disk_used_percent:
    warning: 85
    critical: 95
  disk_io_utilization_percent:
    warning: 80
    critical: 95

Setup Flow

1. Confirm the host is inside the supported Linux or Windows Server OS matrix.
2. Confirm local administrator access for installation and outbound HTTPS access to Bitaic.
3. Install the Health Monitoring agent on the host.
4. Enable the required CPU, memory, disk usage, and disk I/O metrics.
5. Configure safe threshold values in the dashboard, through the API, or by seeding launch values in health-agent-config.yml.
6. Start the agent and confirm the dashboard shows host identity, agent status, recent metrics, and alert state.

Example Workflow: Monitor a Production Host

Goal: install Health Monitoring on a host, collect core resource metrics, and verify alerts before the host becomes a reliability risk.

1. Choose the host to monitor and confirm it is in the supported systems matrix with the required operating-system permissions. Use macOS only as a CLI workstation for Health Monitoring workflows.
2. Install the Health Monitoring agent with local administrator rights and authenticate it to the correct Bitaic workspace.
3. Set the collection interval and enable CPU, memory, disk usage, and disk I/O metrics in health-agent-config.yml.
4. Use the default percentage thresholds or tune them for the host role, then connect those thresholds to the alert route used by the operations team.
5. Start the agent and confirm the dashboard shows fresh host identity, agent status, metrics, and alert state.
6. During normal operations, compare resource trends with deployments, scheduled jobs, capacity changes, and application latency.

Data, Alerts, And Dashboard Output

Bitaic alerts when CPU, memory, disk-used, or disk I/O utilization thresholds reach warning or critical. Alert routing uses the same notification and incident integrations as other Bitaic monitoring products.

Dashboard Views

• Current host status with the latest CPU, memory, disk, disk I/O, and agent-state values, plus telemetry freshness.
• Historical metric trends that help teams compare current resource pressure with prior behavior.
• Threshold and alert-state views that show which metric crossed a safe limit, the configured warning and critical values, and when the alert began.

CLI And API

CLI support includes health install, start, agent status, and health status commands. API support includes retrieving health metrics and managing health thresholds.

• Create or list monitored hosts with /v1/workspaces/{workspace_id}/targets and type: "host".
• Attach host health checks with /v1/workspaces/{workspace_id}/targets/{target_id}/checks and type: "host_health".
• Retrieve health metric data with /v1/workspaces/{workspace_id}/metrics and metrics such as health.cpu.utilization_percent, health.memory.utilization_percent, health.disk.used_percent, and health.agent.heartbeat_age_seconds.
• Update health thresholds with /v1/workspaces/{workspace_id}/checks/{check_id} and the nested thresholds fields used by host_health checks.

bitaic install agent health
bitaic agent start health
bitaic agent status health
bitaic health status

$ bitaic health status
host: app-server-01
os_family: linux
os_version: ubuntu-22.04
architecture: x86_64
agent_status: running
telemetry_state: fresh
agent_freshness_seconds: 18
collection_interval_seconds: 60
cpu_usage_percent: 42
memory_usage_percent: 68
disk_usage_percent: 71
disk_read_bytes_per_second: 1048576
disk_write_bytes_per_second: 524288
disk_io_utilization_percent: 19
alert_state: clear
checked_at: 2026-08-18T14:30:00Z

Troubleshooting