3. Host Interface

The host URI contains information about the current operating system devices, such as the CPU, memory, disk and network. In addition, a host’s network hostname and system timestamp time are also exposed through the monitoring interface.

cpu

The cpu resource allows an administrator to view the CPU devices attached to the host. Selection of the cpu link in the host resource generates a list of all CPUs attached to the host, and also an aggregate all option.

Notice that AMPS also records CPU statistics for the instance process itself.

Element Description Type
idle_percent Percent of CPU time that the system didn’t spend waiting on an I/O request to complete. snapshot
iowait_percent Percent of CPU time spent waiting for I/O requests to complete. snapshot
system_percent Percent of CPU utilization time which occurred while executing kernel processes. snapshot
user_percent Percent of CPU utilization time which occurred while running at the application level. snapshot

Table 3.1: CPU Statistics

Statistics Database Tables: HCPU_STATIC, HCPU_DYNAMIC

Admin path: /amps/host/cpu/<id>/<metric>

disks

The disks resource lists each of the disk devices attached to the host and permits the inspection of disk usage statistics. This information is a readily consumable version of the file /proc/diskstats. Statistics reported are based on the statistics monitoring update frequency (see the AMPS Configuration Guide for details). This means, for example, that reads is the number of disk-reads for the given statistics update interval.

Element Description Type
filesystem_free_percent Percentage of the filesystem currently free. snapshot
mount_point The mount point for this filesystem. fixed
in_progress Number of I/O requests waiting to be processed. snapshot
read_await Average read time completion in milliseconds. interval average
write_await Average write time completion in milliseconds. interval average

Table 3.2: Disks Statistics

Statistics Database Tables: HDISKS_STATIC, HDISKS_DYNAMIC

Admin path: /amps/host/disks/<id>/<metric>

memory

The memory resource gives details about the system memory statistics. All statistics reported are based on the current system statistics reported by examining the file /proc/meminfo. These statistics are updated based on the statistics monitoring update frequency (see Interval in AMPS Configuration Reference Guide).

Element Description Type
available The total amount of memory available. Calculated as the sum of free, buffers and cached. snapshot
buffers The amount of physical memory available for file buffers. snapshot
cached The amount of physical memory used as cache memory. snapshot
free The amount of physical memory left unused by the system. snapshot
in_use The amount of memory currently in use. Provided in the admin console, calculated as total - (free + buffers + cached). snapshot
swap_free The amount of swap memory which is unused. snapshot
swap_total The total amount of physical swap memory. fixed
total Total amount of RAM. fixed

Table 3.3: Memory Statistics

Statistics Database Tables: HMEMORY_STATIC, HMEMORY_DYNAMIC

Admin path: /amps/host/memory/<metric>

name

The name resource displays the network DNS name for the host.

Admin path: /amps/host/name

network

The network resource allows an administrator to examine networking interface statistics on the host. Selecting the network resource displays a list of the network interfaces attached to the host. Selecting one of the interfaces will list the available properties. Reported statistics are rate-based and based on the statistics monitoring update frequency (see the AMPS Configuration Guide for details). For example, collisions for interface eth0 would report the total number of collisions since the last statistics update interval.

Element Description Type
bytes_in Number of bytes received by the interface. interval snapshot
bytes_out Number of bytes transmitted by the interface. interval snapshot
errors Total errors both incoming and outgoing. This number includes packets dropped, collisions, fifo, frame, and carrier errors. interval snapshot
packets_in The total number of packets received by the interface. interval snapshot
packets_out The total number of packets sent by the interface. interval snapshot

Table 3.4: Network Statistics

Statistics Database Tables: HNET_STATIC, HNET_DYNAMIC

Admin path: /amps/host/network/<id>/<metric>

operating_system

The operating_system resource displays the operating system reported by the host.

Admin path: /amps/host/operating_system

utc_time

The utc_time resource displays the system time on the host. Note: the utc_time time reflects the time on the host that the HTTP GET was processed. This differs from all other resources in the host interface as their update frequencies are determined by the Interval tag in the configuration for the Admin interface. For more information on configuring the Administrative Interface and the Interval tag, see the AMPS Configuration Guide.

Admin path: /amps/host/utc_time