Download
RU EN DE FR ES PT 中文

Servers — User Guide | YoBench

How to use the Servers module in YoBench: SSH connection, CPU/RAM/disk/network metrics collection, charts, alert rules, commands and terminal.

What the Servers module does

The module turns YoBench into a lightweight Grafana/Zabbix alternative: connect to Linux servers over SSH and the app collects CPU, memory, disk and network metrics without installing any agent on the box. You see historical charts, can configure alerts on thresholds and run preset shell commands or open a terminal right inside YoBench.

What you get:

  • Single overview screen — every Linux server in one table with current metrics and connection status.
  • Agentless metrics — CPU, load average, memory, disk, disk I/O and network are collected with one SSH call in about a second.
  • 1h / 24h / 7d / 30d charts — data is downsampled automatically per period.
  • Alert rules — threshold + condition (single tick / delta between samples / N consecutive) and YoBench sends a notification.
  • Terminal and commands — an embedded SSH terminal plus a library of preset shell commands.
  • Encrypted secrets — passwords and private SSH keys are stored encrypted in the local database.

Connection parameters

When creating a server you specify:

  • SSH host — IP or DNS name.
  • SSH port — default 22.
  • Username — e.g. root or deploy.
  • Auth type — password or key (PEM private SSH key).
  • Password or Private key — depending on the auth type. Stored encrypted.
  • Polling interval (sec) — check_interval. Range 5–86400; default 60 sec.
  • Enabled — automatic collection flag. Untick to pause without deleting.

Per-server proxy is not supported. The secure connection uses SSH2 with a curated safe-algorithm list and TOFU host verification (the fingerprint is remembered on the first connection).

Collected metrics

A single SSH call runs a built-in shell script without sysstat/iostat dependencies:

  • CPU — cpu_usage_pct (delta of /proc/stat over 1 sec).
  • Load average — 1-, 5-, 15-minute.
  • Memory — total, used (MB), free (%).
  • Disk for / — total, used (GB), free (%).
  • Disk I/O — disk_read_mbps, disk_write_mbps (from /proc/diskstats).
  • Network — net_in_mbps, net_out_mbps (from /sys/class/net).

Charts and periods

The metrics modal exposes these buttons:

  • 1 hour — second-level granularity.
  • 24 hours — minute granularity (1440 points).
  • 7 days — 15-minute buckets (672 points).
  • 30 days — hourly buckets (720 points).

Data is kept for 30 days; older points are removed automatically by a cron job at 3:00 AM.

Alert rules

Every rule is bound to a server and a metric:

  • Metric — one of 12 (CPU, load, memory, disk, network).
  • Condition:
    • single_tick — the current value crosses the threshold.
    • delta — the difference between adjacent samples is over the threshold.
    • consecutive — N samples in a row above/below the threshold.
  • Operator — gt (>) or lt (<).
  • Threshold — numeric value.
  • Debounce — 10 minutes between notifications of the same rule (anti-spam).

Notifications fire only if Server notifications is enabled in settings.

Commands and terminal

  • Commands — a library of shell scripts attached to a server. Define a command once, run it later with one click.
  • Terminal — an interactive SSH session right inside YoBench.

Global module settings

Under Settings → Servers:

  • Enable server monitoring (serversEnabled) — master switch.
  • Server notifications (serversNotificationsEnabled) — turns alert-rule firing on/off.

Notifications go out through YoBench's central notification service (Telegram / email — the app-wide channels).

Workflow

1. Add a server

  1. Open the Servers module from the left sidebar.
  2. Click Add server.
  3. Fill in host, port, user; pick the auth method and provide password or key.
  4. Set the polling interval.
  5. Click Test to verify the connection, then save.

2. Collect metrics manually

The Collect now button on a server row triggers a one-off collection without waiting for the next tick.

3. Explore charts

Clicking a server opens the metrics modal. Switch between 1h / 24h / 7d / 30d and analyze CPU, memory, disk and network trends.

4. Configure alerts

Open Alert rules, set a threshold and condition type. Once Server notifications is on in settings, you'll receive messages when rules fire.

5. Use commands and terminal

  • In Commands, add the scripts you run often (e.g. systemctl status).
  • The Terminal button opens an interactive SSH session for ad-hoc troubleshooting.

Next steps

  • Run Health Check for parallel web availability monitoring.
  • For your local machine, use System Monitor.
  • Notifications are configured in the central notifications section (Telegram, email).

Help and feedback

Want Windows servers, proxy support or other metrics? Contact us via the feedback form.