List of available server metrics

Note

See the dashboards documentation for example charts of the metrics described below.

Metric name	Description
CPU usage (avg)	The average usage of a server’s total CPU capacity across all cores.
CPU usage per core (avg)	The average usage of each CPU core.
CPU usage (max)	The maximum usage of a server’s total CPU capacity across all cores. For data points spanning more than a one-minute range, the value is the maximum usage over any one-minute period within the time span of the data point.
CPU usage per core (max)	The maximum usage of each CPU core. For data points spanning more than a one-minute range, the value is the maximum usage over any one-minute period within the time span of the data point.
CPU iowait (avg)	The average percent of time waiting on disk I/O across all CPU cores. When a process is waiting on disk I/O, the kernel will normally schedule another process to run so CPU time is not wasted. However, if there are no runnable processes and so the CPU must wait for disk I/O to complete, that time is considered “iowait”.
CPU iowait per core (max)	The maximum percent of time waiting on disk I/O of each CPU core. When a process is waiting on disk I/O, the kernel will normally schedule another process to run so CPU time is not wasted. However, if there are no runnable processes and so the CPU must wait for disk I/O to complete, that time is considered “iowait”.
Memory usage - effective (max)	The maximum percentage of a server’s memory unavailable for use by processes. Memory that is unavailable for use includes memory already in use by processes as well as memory the kernel reserves for low-level operations. When all memory is unavailable, a server will become slow as it must use swap space. When swap is full, the kernel must kill processes.
Memory usage - including disk cache (max)	The maximum percentage of a server’s memory in use for any purpose, including the kernel’s disk cache. It is not bad for a server to use nearly all of its memory for some purpose as unused memory is essentially wasted memory.
Swap usage (max)	The maximum percentage of a server’s swap space in use. Swap space is disk space the kernel uses as secondary memory. As disk access is extremely slow compared to memory access, a server will become slow if the kernel is frequently needing to move data in and out of swap. However, some swap usage is normal even when a server is not low on memory.
OOM kills	The number of processes killed by the kernel’s Out of Memory (OOM) manager. When a server runs out of both memory and swap space, the kernel must kill processes. Note: OOM Kills chart data is not available on Ubuntu 16.04.
Disk usage	The percentage of storage space in use on a filesystem.
Inode usage	The percentage of inodes in use on a filesystem. Inodes are data structures used by filesystems to store information about every file and directory. High inode usage usually indicates there are a large number of small files.
Network traffic (avg)	The average rate of traffic sent and received by a server.