HP Matrix Operating Environment Software - Manual (Page 145)

metric: One or more of the following metrics: CPU_UTIL, MEM_UTIL, NET_UTIL, DISK_UTIL,
CPU_ALLOC, MEM_ALLOC, PHYS_CPUS, PHYS_MEM, CPU_QUEUE, PAGES_PER_SEC,
IOS_PER_SEC, PKT_PER_SEC, DISK_USED.
The command only outputs a result when there is data for the given metric during the provided
period of time. If the metric data covers some of the time range, but does not encompass all of
the time range, the column is padded with NaN values. You cannot import a file containing only
metrics — you must include the UTIS with any metrics that you select.
The fields that appear on this line are defined as:
YYYYMMDDhhmm: Timestamp in local time, given in units of YYYY (year), MM (month, as 01
to 12), DD (day, as 01 to 31), hh (hours, as 00 to 23), and mm (minutes, as 00 to 59). This format
is also used to specify the begin and end times for the capprofile(1m) command. This field
is optional when importing data.
UTIS: Universal Time (GMT) in seconds (standard UNIX time in seconds since 1 January 1970)
(Required)
CPU_UTIL: CPU utilization expressed as the number of CPUs used (5 minute average).
MEM_UTIL: Memory utilization or amount of memory used in gigabytes (at end of interval
reading).
NET_UTIL: Network bandwidth utilization, in Mb/s (10^6 bits, megabits per second) (5 minute
average).
DISK_UTIL: Disk bandwidth utilization in MB/s (10^6 bytes, megabytes per second) (5 minute
average).
CPU_ALLOC: Number of CPUs allocated (active CPUs only; cores or hardware threads on
multicore). Hyperthreaded processors are not counted in CPU_ALLOC.
MEM_ALLOC: Memory allocation in GB (gigabytes).
PHYS_CPUS: Number of physical CPUs (at end of interval reading).
PHYS_MEM: Physical memory available, in GB (gigabytes).
CPU_QUEUE: The average depth of the CPU run queue.
PAGES_PER_SEC: The peak pages input and output per second.
IOS_PER_SEC: The number of disk I/O operations executed per second.
PKT_PER_SEC: The number of network packets either sent or received per second.
DISK_USED: The disk space used on local disks, expressed in 1024-byte blocks..
The last line of the import file header must include at least one metric column label and may
have any combination of the metric column labels, in any order. It must have the UTIS timestamp
column label in position 1 or, if the YYYYMMDDhhmm column label is present, in position 2.
A metric label can appear only once in the header. UTIS timestamps must increment in a consistent
order for each row; YYYYMMDDhhmm values are provided for readability.
To import a subset of metrics, specify only those metrics in the header. For example, to import
only the CPU_UTIL metric, the last header line should contain “YYYYMMDDhhmm, UTIS,
CPU_UTIL” or “ UTIS, CPU_UTIL.”
Each subsequent line after the header contains utilization values for a specified moment in time.
Each data line, called a sample, has a timestamp followed by the values as designated in the
header. Utilization values for a sample are separated by a specified delimiter character or by a
comma (default). The UTIL seconds must be in 5-minute intervals (increments of 300 seconds).
NOTE: In some of the following examples, UTIS values are truncated to omit following zeroes.
EXAMPLES
The following example shows utilization data in compressed format created using the following
command:
# capprofile -x -b20090824 -e20090825 node05.company.com
145