System Metrics
Subscribing to system metrics will display a graphical visualization (like a graph with metrics data plotted along the x-axis and time along the y-axis) of the metrics.
Read the metrics reference to know the data fields included in each system metric.
Network IO Interface
You can monitor and analyze real-time network performance metrics. For instance, you can collect upload and download rates per second for all interfaces for the last hour. The following metrics reference includes the kind of data measured in a network IO graph.
Some of the metrics data measured in a network IO interface graph are:
- bytes_sent is the total number of bytes sent by the network interface
- bytes_recv is the total number of bytes received by the network interface
- packets_sent is the total number of packets sent by the network interface
- packets_recv is the total number of packets received by the network interface
- error_in is the total number of receive errors detected by the network interface
- error_out is the total number of transmit errors detected by the network interface
- drop_in is the total number of received packets dropped by the network interface
- drop_out is the total number of transmitted packets dropped by the network interface
Disk usage
Subscribing to disk usage metrics displays a graph that shows the information on disk usage metrics like available disk space, percentage of disk space used, etc.
Some of the metrics data measured in a disk usage graph are:
- free is the amount of disk space that is freely available. It is an integer value shown as a byte.
- total is the total amount of disk space. It is an integer value shown as a byte.
used is the amount of disk space that is in use. It is an integer value shown as a byte.
- used_percent is the percentage of disk space used. It is a floating-point value shown as a percentage.
- inodes_free is the number of free inodes. It is an integer value shown as files.
- inodes_used is the number of used inodes. It is an integer value shown as files.
- inodes_total is the total number of inodes. It is an integer value shown as files.
Disk IO
You can gather metrics about disk traffic and timing by subscribing to the disk IO. Read more to know the kind of information collected by disk IO metric.
Some of the metrics data measured in a disk IO interface graph are:
- reads is a counter that increments when a read request completes. It is an integer value.
- writes is a counter that increments when a write request completes. It is an integer value.
- read_bytes is the count of the number of bytes read from the device. It is an integer value shown as a byte.
- write_bytes is the count of the number of bytes written to the device. It is an integer value shown as a byte.
- read_time is the count of the number of milliseconds that read requests have waited on the device. It is an integer value shown as milliseconds.
- write_time is the count of the number of milliseconds that write requests have waited on the device. It is an integer value shown as milliseconds.
- io_time is the count of the number of milliseconds during which the device has had IO requests queued. It is an integer value shown as milliseconds.
- weighted_io_time is the count of the number of milliseconds that IO requests have waited on the device.
- iops_in_progress is the count of the number of IO requests that have been issued to the device but have not yet been completed. It does not include IO requests that are in the queue but not yet issued to the device.
Memory usage
You can collect and visualize device memory metrics by subscribing to the memory usage system metric. To know more about the data fields associated with memory usage, read here.
Some of the metrics data measured in a memory usage graph are:
- available is the amount of memory that is available. It is an integer value shown as a byte.
- available_percent is the percentage of memory that is available. It is a floating-point value shown as a percentage
- buffered is the amount of physical RAM used as cache memory. It is an integer value shown as a byte
- cached is the amount of physical RAM used as cache memory. It is an integer value shown as a byte
- free is the amount of free RAM. It is an integer value shown as byte
inactive is the amount of memory that hasn’t been used in some way. It is an integer value shown as a byte
- slab is the amount of memory used by the kernel to cache data structures for its use. It is an integer value shown as a byte
- total is the total amount of physical RAM. It is an integer value shown as a byte.
- used is the amount of RAM in use. It is an integer value shown as a byte.
- used_percent is the percentage of memory currently in use. It is a floating-point value shown as a percentage
- active is the amount of memory that has been used in some way. It is an integer value shown as a byte.
- wired is the memory where the kernel and other low-level components like device drivers and virtual memory objects are stored.
CPU load average
You can determine the percentage of CPU used by a user, process, or system by subscribing to the CPU load average metric. Read the CPU load average reference to understand the several data measurements plotted in the graph.
Some of the metrics data measured in a CPU load average graph are:
- usage_guest is the percentage of time that the CPU is running a virtual CPU for a guest operating system
- usage_guest_nice is the percentage of time that the CPU is running a virtual CPU for a guest operating system, which is a low priority and can be interrupted by other processes.
- usage_idle is the percentage of time that the CPU is idle
usage_iowait is the percentage of time that the CPU is waiting for IO operations to complete
- usage_irq is the percentage of time that the CPU is servicing interrupts
- usage_nice is the percentage of time that the CPU is in user mode with a low priority process, which a higher priority process can interrupt
- usage_softirq is the percentage of time that the CPU is servicing software interrupts
- usage_steal is the percentage of time that the CPU is in stolen time or time spent in other operating systems in a virtualized environment
- usage_system is the percentage of time that the CPU is in system mode
- usage_user is the percentage of time that the CPU is in user mode
Wireless
You can monitor and analyze the signal strength of a wifi connection by subscribing to the wireless metric. This information is extracted from /proc/net/wireless when a device is connected to a wifi connection. Read the wireless reference to understand what kind of data will be collected when it is subscribed. For more information, click here