| The time value will generally be a multiple of the requested sample period |
| |
| sequence of snapshots indicate a stretch of time where the process's memory |
| usage did not change. |
Domain Id | System identification number of the logical domain. |
| On ccNUMA systems, cell local memory domains are numbered starting |
| at 1, and the interleaved memory domain Id is |
| On SMP systems, the only domain is numbered 0. |
Shared Pages | Number of shared resident memory pages currently in use by this process. |
| This is typically executable code for the program, shared libraries, and |
| mmap'd (shared) regions. These are pages that could be shared with another |
| process but are not necessarily currently being shared. |
Private Pages | Number of private resident memory pages currently in use by this process. |
| These are typically static data and dynamic (malloc'd heap) regions. |
| These memory pages cannot be shared with other processes. |
Weighted Pages | Number of resident pages for this process, weighted by the number of |
| processes actually sharing each page. Private pages count as one page |
| and shared pages count as the page divided by the number of processes |
| sharing that page. Therefore, it will be at least as large as the number of |
| private pages but less than or equal to the total number of pages. It indicates |
| the “load” that this process is putting on system memory. |
Total Pages | Total number of resident pages that are currently allocated to this process. |
| It is the total of Shared Pages and Private Pages. |
Measuring System Usage Concurrently with Other Measurements
You can request a measurement of your program's system usage in addition to, and at the same time as, any other HP Caliper measurement.
To do this, use the
For example:
$ caliper cpu
You can use this option to get four kinds of system measurement data:
•Run status: how much time each process or thread spent running, eligible to run but not running, and waiting (runstatus parameter)
•cpu: how much of each process or thread running time was spent on each CPU and how often the process or thread was moved to another CPU (cpu parameter).
•io: gives a breakdown of all I/O a process or thread does, by category (10 parameter).
•System calls: the count and time spent in every system call called by a process or thread (syscalls parameter)
The all parameter causes all four kinds of system data to be measured.
As with all HP Caliper
Example Report Output
Figure 29 (page 159) shows some example output for the
158 HP Caliper Features Specific to