Metrics Collected

Every interval (60 seconds by default), the host agent samples the metrics below and posts them to NoDisrupt in a single payload, along with the server's hostname, IP address and any tags you configured. This page lists every metric it collects, grouped by category, so you know exactly what to expect when you build a dashboard or alert on them.

CPU

• cpu_percent - total CPU utilisation across all cores, as a percentage. Sampled over a one-second window each collection cycle.

Memory

• mem_total_bytes - total physical memory installed, in bytes.
• mem_used_bytes - physical memory currently in use, in bytes.
• mem_used_percent - memory usage as a percentage of total.

Swap

• swap_total_bytes - total swap space configured, in bytes.
• swap_used_bytes - swap space currently in use, in bytes.
• swap_used_percent - swap usage as a percentage of total.

Load average

• load1 - system load average over the last 1 minute.
• load5 - system load average over the last 5 minutes.
• load15 - system load average over the last 15 minutes.

Disk

The agent reports usage per physical disk, not per partition. Partitions on the same physical disk (for example /dev/sda1 and /dev/sda2, or /dev/nvme0n1p1) are aggregated together under one disk name before being reported, so you get one pair of metrics per physical device rather than one per mount point.

The <name> part of each metric is derived from the disk's full device path: lowercased, with every run of characters that is not a letter or digit replaced by a single underscore, and leading or trailing underscores trimmed. So /dev/sda becomes dev_sda and reports as disk_dev_sda_used_percentage, an NVMe drive /dev/nvme0n1 reports as disk_dev_nvme0n1_used_percentage, and an LVM volume /dev/mapper/vg-root reports as disk_dev_mapper_vg_root_used_percentage.

• disk_<name>_used_percentage - used disk space as a percentage of total, for the physical disk <name>.
• disk_<name>_inodes_remaining - the number of free inodes remaining on that physical disk.

Network

These counters are cumulative across all network interfaces since the host last booted, not deltas since the last collection.

• net_bytes_sent - total bytes sent.
• net_bytes_recv - total bytes received.
• net_packets_sent - total packets sent.
• net_packets_recv - total packets received.