DOKK / manpages / debian 12 / pcp / pcp-atop.1.en
PCP-ATOP(1) General Commands Manual PCP-ATOP(1)

pcp-atop - Advanced System and Process Monitor

Interactive Usage:

pcp [pcp options] atop [-aAcCdDfFgGmMnNopRsuvxyY1] [-L linelen] [-Plabel[,label]... [-Z]] [interval [samples]]

Writing and reading PCP archive folios:

pcp atop -w folio [-a] [-S] [interval [samples]]
pcp atop -r folio [-AcCdDfFgGmMnNopRsuvxy1] [-b [yy-mm-dd] hh:mm] [-e yy-mm-dd] hh:mm] [-L linelen] [-Plabel[,label]... [-Z]] [interval [samples]]

The program pcp-atop is an interactive monitor to view various aspects of load on a system. It shows the occupation of the most critical hardware resources (from a performance point of view) on system level, i.e. cpu, memory, disk and network.
It also shows which processes are responsible for the indicated load with respect to cpu and memory load on process level. Disk load is shown per process if "storage accounting" is active in the kernel.

Every interval (default: 10 seconds) information is shown about the resource occupation on system level (cpu, memory, disks and network layers), followed by a list of processes which have been active during the last interval (note that all processes that were unchanged during the last interval are not shown, unless the key 'a' has been pressed or unless sorting on memory occupation is done). If the list of active processes does not entirely fit on the screen, only the top of the list is shown (sorted in order of activity).
The intervals are repeated till the number of samples (specified as command argument) is reached, or till the key 'q' is pressed in interactive mode.

When invoked via the pcp(1) command, the PCPIntro(1) options -A/--align, -a/--archive, -h/--host, -O/--origin, -S/--start, -s/--samples, -T/--finish, -t/--interval, -v/--version, -z/, --hostzone and -z/, --timezone become indirectly available. Additionally, the --hotproc option can be used to request the per-process PCP metrics be used instead of the default proc metrics from pmdaproc(1).

When pcp-atop is started, it checks whether the standard output channel is connected to a screen, or to a file/pipe. In the first case it produces screen control codes (via the ncurses library) and behaves interactively; in the second case it produces flat ASCII-output.

In interactive mode, the output of pcp-atop scales dynamically to the current dimensions of the screen/window.
If the window is resized horizontally, columns will be added or removed automatically. For this purpose, every column has a particular weight. The columns with the highest weights that fit within the current width will be shown.
If the window is resized vertically, lines of the process/thread list will be added or removed automatically.

Furthermore in interactive mode the output of pcp-atop can be controlled by pressing particular keys. However it is also possible to specify such key as flag on the command line. In that case pcp-atop switches to the indicated mode on beforehand; this mode can be modified again interactively. Specifying such key as flag is especially useful when running pcp-atop with output to a pipe or file (non-interactively). These flags are the same as the keys that can be pressed in interactive mode (see section INTERACTIVE COMMANDS).
Additional flags are available to support storage of pcp-atop data in PCP archive format (see section PCP DATA STORAGE).

For the resource consumption on system level, pcp-atop uses colors to indicate that a critical occupation percentage has been (almost) reached. A critical occupation percentage means that is likely that this load causes a noticeable negative performance influence for applications using this resource. The critical percentage depends on the type of resource: e.g. the performance influence of a disk with a busy percentage of 80% might be more noticeable for applications/user than a CPU with a busy percentage of 90%.
Currently pcp-atop uses the following default values to calculate a weighted percentage per resource:

 Processor
A busy percentage of 90% or higher is considered `critical'.
 Disk
A busy percentage of 70% or higher is considered `critical'.
 Network
A busy percentage of 90% or higher for the load of an interface is considered `critical'.
 Memory
An occupation percentage of 90% is considered `critical'. Notice that this occupation percentage is the accumulated memory consumption of the kernel (including slab) and all processes; the memory for the page cache (`cache' and `buff' in the MEM-line) and the reclaimable part of the slab (`slrec`) is not implied!
If the number of pages swapped out (`swout' in the PAG-line) is larger than 10 per second, the memory resource is considered `critical'. A value of at least 1 per second is considered `almost critical'.
If the committed virtual memory exceeds the limit (`vmcom' and `vmlim' in the SWP-line), the SWP-line is colored due to overcommitting the system.
 Swap
An occupation percentage of 80% is considered `critical' because swap space might be completely exhausted in the near future; it is not critical from a performance point-of-view.

These default values can be modified in the configuration file (see separate man-page of pcp-atoprc(5)).

When a resource exceeds its critical occupation percentage, the concerning values in the screen line are colored red by default.
When a resource exceeded (default) 80% of its critical percentage (so it is almost critical), the concerning values in the screen line are colored cyan by default. This `almost critical percentage' (one value for all resources) can be modified in the configuration file (see separate man-page of pcp-atoprc(5)).
The default colors red and cyan can be modified in the configuration file as well (see separate man-page of pcp-atoprc(5)).

With the key 'x' (or flag -x), the use of colors can be suppressed.

GPU statistics can be gathered by pmdanvidia(1) which is a separate data collection daemon process. It gathers cumulative utilization counters of every Nvidia GPU in the system, as well as utilization counters of every process that uses a GPU. When pcp-atop notices that the daemon is active, it reads these GPU utilization counters with every interval.

Find a description about the utilization counters in the section OUTPUT DESCRIPTION.

When running pcp-atop interactively (no output redirection), keys can be pressed to control the output. In general, lower case keys can be used to show other information for the active processes and upper case keys can be used to influence the sort order of the active process/thread list.

Show generic output (default).

Per process the following fields are shown in case of a window-width of 80 positions: process-id, cpu consumption during the last interval in system and user mode, the virtual and resident memory growth of the process.

The subsequent columns depend on the used kernel:
When the kernel supports "storage accounting" (>= 2.6.20), the data transfer for read/write on disk, the status and exit code are shown for each process. When the kernel does not support "storage accounting", the username, number of threads in the thread group, the status and exit code are shown.
The last columns contain the state, the occupation percentage for the chosen resource (default: cpu) and the process name.

When more than 80 positions are available, other information is added.

Show memory related output.

Per process the following fields are shown in case of a window-width of 80 positions: process-id, minor and major memory faults, size of virtual shared text, total virtual process size, total resident process size, virtual and resident growth during last interval, memory occupation percentage and process name.

When more than 80 positions are available, other information is added.

For memory consumption, always all processes are shown (also the processes that were not active during the interval).

Show disk-related output.

When "storage accounting" is active in the kernel, the following fields are shown: process-id, amount of data read from disk, amount of data written to disk, amount of data that was written but has been withdrawn again (WCANCL), disk occupation percentage and process name.

Show scheduling characteristics.

Per process the following fields are shown in case of a window-width of 80 positions: process-id, number of threads in state 'running' (R), number of threads in state 'interruptible sleeping' (S), number of threads in state 'uninterruptible sleeping' (D), scheduling policy (normal timesharing, realtime round-robin, realtime fifo), nice value, priority, realtime priority, current processor, status, exit code, state, the occupation percentage for the chosen resource and the process name.

When more than 80 positions are available, other information is added.

Show various process characteristics.

Per process the following fields are shown in case of a window-width of 80 positions: process-id, user name and group, start date and time, status (e.g. exit code if the process has finished), state, the occupation percentage for the chosen resource and the process name.

When more than 80 positions are available, other information is added.

Show the command line of the process.

Per process the following fields are shown: process-id, the occupation percentage for the chosen resource and the command line including arguments.

Show cgroup v2 information.

Per process the following fields are shown: process-id, the command name, and the cgroup path name (horizontally scrollable).

Show GPU utilization.

Per process at least the following fields are shown: process-id, range of GPU numbers on which the process currently runs, GPU busy percentage on all GPUs, memory busy percentage (i.e. read and write accesses on memory) on all GPUs, memory occupation at the moment of the sample, average memory occupation during the sample, and GPU percentage.

When the pmdanvidia daemon does not run with root privileges, the GPU busy percentage and the memory busy percentage are not available on process level. In that case, the GPU percentage on process level reflects the GPU memory occupation instead of the GPU busy percentage (which is preferred).

Show the user-defined line of the process.

In the configuration file the keyword ownprocline can be specified with the description of a user-defined output-line.
Refer to the man-page of pcp-atoprc(5) for a detailed description.

Show the individual threads within a process (toggle).

Single-threaded processes are still shown as one line.
For multi-threaded processes, one line represents the process while additional lines show the activity per individual thread (in a different color). Depending on the option 'a' (all or active toggle), all threads are shown or only the threads that were active during the last interval. Depending on the option 'Y' (sort threads), the threads per process will be sorted on the chosen sort criterium or not.
Whether this key is active or not can be seen in the header line.

Sort the threads per process when combined with option 'y' (toggle).
Show the process activity accumulated per user.

Per user the following fields are shown: number of processes active or terminated during last interval (or in total if combined with command `a'), accumulated cpu consumption during last interval in system and user mode, the current virtual and resident memory space consumed by active processes (or all processes of the user if combined with command `a').
When "storage accounting" is active in the kernel, the accumulated read and write throughput on disk is shown. When the pmdabcc(1) module `netproc' has been installed, the number of receive and send network calls are shown.
The last columns contain the accumulated occupation percentage for the chosen resource (default: cpu) and the user name.

Show the process activity accumulated per program (i.e. process name).

Per program the following fields are shown: number of processes active or terminated during last interval (or in total if combined with command `a'), accumulated cpu consumption during last interval in system and user mode, the current virtual and resident memory space consumed by active processes (or all processes of the user if combined with command `a').
When "storage accounting" is active in the kernel, the accumulated read and write throughput on disk is shown. When the pmdabcc(1) module `netproc' has been installed, the number of receive and send network calls are shown.
The last columns contain the accumulated occupation percentage for the chosen resource (default: cpu) and the program name.

Show the process activity accumulated per Docker container.

Per container the following fields are shown: number of processes active or terminated during last interval (or in total if combined with command `a'), accumulated cpu consumption during last interval in system and user mode, the current virtual and resident memory space consumed by active processes (or all processes of the user if combined with command `a').
When "storage accounting" is active in the kernel, the accumulated read and write throughput on disk is shown. When the pmdabcc(1) module `netproc' has been installed, the number of receive and send network calls are shown.
The last columns contain the accumulated occupation percentage for the chosen resource (default: cpu) and the Docker container id (CID).

Sort the current list in the order of cpu consumption (default). The one-but-last column changes to ``CPU''.
Sort the current list in the order of GPU utilization (preferred, but only applicable when the pmdanvidia daemon runs under root privileges) or the order of GPU memory occupation). The one-but-last column changes to ``GPU''.
Sort the current list in the order of resident memory consumption. The one-but-last column changes to ``MEM''. In case of sorting on memory, the full process list will be shown (not only the active processes).
Sort the current list in the order of disk accesses issued. The one-but-last column changes to ``DSK''.
Sort the current list in the order of network bandwidth (received and transmitted). The one-but-last column changes to ``NET''.
Sort the current list automatically in the order of the most busy system resource during this interval. The one-but-last column shows either ``ACPU'', ``AMEM'', ``ADSK'' or ``ANET'' (the preceding 'A' indicates automatic sorting-order). The most busy resource is determined by comparing the weighted busy-percentages of the system resources, as described earlier in the section COLORS.
This option remains valid until another sorting-order is explicitly selected again.
A sorting-order for disk is only possible when "storage accounting" is active. A sorting-order for network is only possible when the pmdabcc(1) module `netproc' has been installed.

Miscellaneous interactive commands:

?
Request for help information (also the key 'h' can be pressed).
Request for version information (version number and date).
Gather and calculate the proportional set size of processes (toggle). Gathering of all values that are needed to calculate the PSIZE of a process is a very time-consuming task, so this key should only be active when analyzing the resident memory consumption of processes.
Get the WCHAN per thread (toggle). Gathering of the WCHAN string per thread is a relatively time-consuming task, so this key should only be made active when analyzing the reason for threads to be in sleep state.
Suppress colors to highlight critical resources (toggle).
Whether this key is active or not can be seen in the header line.
The pause key can be used to freeze the current situation in order to investigate the output on the screen. While pcp-atop is paused, the keys described above can be pressed to show other information about the current list of processes. Whenever the pause key is pressed again, pcp-atop will continue with the next sample.
Modify the interval timer (default: 10 seconds). If an interval timer of 0 is entered, the interval timer is switched off. In that case a new sample can only be triggered manually by pressing the key 't'.
Trigger a new sample manually. This key can be pressed if the current sample should be finished before the timer has exceeded, or if no timer is set at all (interval timer defined as 0). In the latter case pcp-atop can be used as a stopwatch to measure the load being caused by a particular application transaction, without knowing on beforehand how many seconds this transaction will last.

When viewing the contents of an archive folio, this key can be used to show the next sample from the folio.

When viewing the contents of an archive folio, this key can be used to show the previous sample from the folio.
When viewing the contents of an archive folio, this key can be used to move to a certain timestamp within the file (either forward or backward).
Reset all counters to zero to see the system and process activity since boot again.

When viewing the contents of an archive, this key can be used to rewind to the beginning of the file again.

Specify a search string for specific user names as a regular expression. From now on, only (active) processes will be shown from a user which matches the regular expression. The system statistics are still system wide. If the Enter-key is pressed without specifying a name, (active) processes of all users will be shown again.
Whether this key is active or not can be seen in the header line.
Specify a list with one or more PIDs to be selected. From now on, only processes will be shown with a PID which matches one of the given list. The system statistics are still system wide. If the Enter-key is pressed without specifying a PID, all (active) processes will be shown again.
Whether this key is active or not can be seen in the header line.
Specify a search string for specific process names as a regular expression. From now on, only processes will be shown with a name which matches the regular expression. The system statistics are still system wide. If the Enter-key is pressed without specifying a name, all (active) processes will be shown again.
Whether this key is active or not can be seen in the header line.
/
Specify a specific command line search string as a regular expression. From now on, only processes will be shown with a command line which matches the regular expression. The system statistics are still system wide. If the Enter-key is pressed without specifying a string, all (active) processes will be shown again.
Whether this key is active or not can be seen in the header line.
Specify a Docker container id of 12 (hexadecimal) characters. From now on, only processes will be shown that run in that specific Docker container (CID). The system statistics are still system wide. If the Enter-key is pressed without specifying a container id, all (active) processes will be shown again.
Whether this key is active or not can be seen in the header line.
Specify a comma-separated list of process state characters. From now on, only processes will be shown that are in those specific process states. Accepted states are: R (running), S (sleeping), D (disk sleep), T (stopped), t (tracing stop), X (dead), Z (zombie) and P (parked). The system statistics are still system wide. If the Enter-key is pressed without specifying a state, all (active) processes will be shown again.
Whether this key is active or not can be seen in the header line.
Specify search strings for specific logical volume names, specific disk names and specific network interface names. All search strings are interpreted as a regular expressions. From now on, only those system resources are shown that match the concerning regular expression. If the Enter-key is pressed without specifying a search string, all (active) system resources of that type will be shown again.
Whether this key is active or not can be seen in the header line.
The `all/active' key can be used to toggle between only showing/accumulating the processes that were active during the last interval (default) or showing/accumulating all processes.
Whether this key is active or not can be seen in the header line.
By default, pcp-atop shows/accumulates the processes that are alive and the processes that are exited during the last interval. With this key (toggle), showing/accumulating the processes that are exited can be suppressed.
Whether this key is active or not can be seen in the header line.
Show a fixed (maximum) number of header lines for system resources (toggle). By default only the lines are shown about system resources (CPUs, paging, logical volumes, disks, network interfaces) that really have been active during the last interval. With this key you can force pcp-atop to show lines of inactive resources as well.
Whether this key is active or not can be seen in the header line.
Suppress sorting of system resources (toggle). By default system resources (CPUs, logical volumes, disks, network interfaces) are sorted on utilization.
Whether this key is active or not can be seen in the header line.
1
Show relevant counters as an average per second (in the format `..../s') instead of as a total during the interval (toggle).
Whether this key is active or not can be seen in the header line.
Limit the number of system level lines for the counters per-cpu, the active disks and the network interfaces. By default lines are shown of all CPUs, disks and network interfaces which have been active during the last interval. Limiting these lines can be useful on systems with huge number CPUs, disks or interfaces in order to be able to run pcp-atop on a screen/window with e.g. only 24 lines.
For all mentioned resources the maximum number of lines can be specified interactively. When using the flag -l the maximum number of per-cpu lines is set to 0, the maximum number of disk lines to 5 and the maximum number of interface lines to 3. These values can be modified again in interactive mode.
Send a signal to an active process (a.k.a. kill a process).
Quit the program.
Show the next page of the process/thread list.
With the arrow-down key the list can be scrolled downwards with single lines.
^F
Show the next page of the process/thread list (forward).
With the arrow-down key the list can be scrolled downwards with single lines.
Show the previous page of the process/thread list.
With the arrow-up key the list can be scrolled upwards with single lines.
^B
Show the previous page of the process/thread list (backward).
With the arrow-up key the list can be scrolled upwards with single lines.
^L
Redraw the screen.

In order to store system and process level statistics for long-term analysis (e.g. to check the system load and the active processes running yesterday between 3:00 and 4:00 PM), pcp-atop can store the system and process level statistics in the PCP archive format, as an archive folio (see mkaf(1)).
All information about processes and threads is stored in the archive.
The interval (default: 10 seconds) and number of samples (default: infinite) can be passed as last arguments. Instead of the number of samples, the flag -S can be used to indicate that pcp-atop should finish anyhow before midnight.

A PCP archive can be read and visualized again with the -r option. The argument is a comma-separated list of names, each of which may be the base name of an archive or the name of a directory containing one or more archives. If no argument is specified, the file $PCP_LOG_DIR/pmlogger/HOST/YYYYMMDD is opened for input (where YYYYMMDD are digits representing the current date, and HOST is the hostname of the machine being logged). If a filename is specified in the format YYYYMMDD (representing any valid date), the file $PCP_LOG_DIR/pmlogger/HOST/YYYYMMDD is opened. If a filename with the symbolic name y is specified, yesterday's daily logfile is opened (this can be repeated so 'yyyy' indicates the logfile of four days ago).
The samples from the file can be viewed interactively by using the key 't' to show the next sample, the key 'T' to show the previous sample, the key 'b' to branch to a particular time or the key 'r' to rewind to the begin of the file.
When output is redirected to a file or pipe, pcp-atop prints all samples in plain ASCII. The default line length is 80 characters in that case; with the flag -L followed by an alternate line length, more (or less) columns will be shown.
With the flag -b (begin time) and/or -e (end time) followed by a time argument of the form [YY-MM-DD] HH:MM, a certain time period within the archive can be selected.

The first sample shows the system level activity since boot (the elapsed time in the header shows the time since boot). Note that particular counters could have reached their maximum value (several times) and started by zero again, so do not rely on these figures.

For every sample pcp-atop first shows the lines related to system level activity. If a particular system resource has not been used during the interval, the entire line related to this resource is suppressed. So the number of system level lines may vary for each sample.
After that a list is shown of processes which have been active during the last interval. This list is by default sorted on cpu consumption, but this order can be changed by the keys which are previously described.

If values have to be shown by pcp-atop which do not fit in the column width, another format is used. If e.g. a cpu-consumption of 233216 milliseconds should be shown in a column width of 4 positions, it is shown as `233s' (in seconds). For large memory figures, another unit is chosen if the value does not fit (Mb instead of Kb, Gb instead of Mb, Tb instead of Gb, ...). For other values, a kind of exponent notation is used (value 123456789 shown in a column of 5 positions gives 123e6).

OUTPUT DESCRIPTION - SYSTEM LEVEL

The system level information consists of the following output lines:

Process and thread level totals.
This line contains the total cpu time consumed in system mode (`sys') and in user mode (`user'), the total number of processes present at this moment (`#proc'), the total number of threads present at this moment in state `running' (`#trun'), `sleeping interruptible' (`#tslpi') and `sleeping uninterruptible' (`#tslpu'), the number of zombie processes (`#zombie'), the number of clone system calls (`clones'), and the number of processes that ended during the interval (`#exit') when process accounting is used. Instead of `#exit` the last column may indicate that process accounting could not be activated (`no procacct`).
If the screen-width does not allow all of these counters, only a relevant subset is shown.
CPU utilization.
At least one line is shown for the total occupation of all CPUs together.
In case of a multi-processor system, an additional line is shown for every individual processor (with `cpu' in lower case), sorted on activity. Inactive CPUs will not be shown by default. The lines showing the per-cpu occupation contain the cpu number in the field combined with the wait percentage.

Every line contains the percentage of cpu time spent in kernel mode by all active processes (`sys'), the percentage of cpu time consumed in user mode (`user') for all active processes (including processes running with a nice value larger than zero), the percentage of cpu time spent for interrupt handling (`irq') including softirq, the percentage of unused cpu time while no processes were waiting for disk I/O (`idle'), and the percentage of unused cpu time while at least one process was waiting for disk I/O (`wait').
In case of per-cpu occupation, the cpu number and the wait percentage (`w') for that cpu. The number of lines showing the per-cpu occupation can be limited.

For virtual machines, the steal-percentage (`steal') shows the percentage of cpu time stolen by other virtual machines running on the same hardware.
For physical machines hosting one or more virtual machines, the guest-percentage (`guest') shows the percentage of cpu time used by the virtual machines. Notice that this percentage overlaps the user-percentage!

When PMC performance monitoring counters are supported by the CPU and the kernel (and pmdaperfevent(1) runs with root privileges), the number of instructions per CPU cycle (`ipc') is shown. The first sample always shows the value 'initial', because the counters are just activated at the moment that pcp-atop is started.
When the CPU busy percentage is high and the IPC is less than 1.0, it is likely that the CPU is frequently waiting for memory access during instruction execution (larger CPU caches or faster memory might be helpful to improve performance). When the CPU busy percentage is high and the IPC is greater than 1.0, it is likely that the CPU is instruction-bound (more/faster cores might be helpful to improve performance).
Furthermore, per CPU the effective number of cycles (`cycl') is shown. This value can reach the current CPU frequency if such CPU is 100% busy. When an idle CPU is halted, the number of effective cycles can be (considerably) lower than the current frequency.
Notice that the average instructions per cycle and number of cycles is shown in the CPU line for all CPUs.
See also: http://www.brendangregg.com/blog/2017-05-09/cpu-utilization-is-wrong.html

In case of frequency scaling, all previously mentioned CPU percentages are relative to the used scaling of the CPU during the interval. If a CPU has been active for e.g. 50% in user mode during the interval while the frequency scaling of that CPU was 40%, only 20% of the full capacity of the CPU has been used in user mode.
In case that the kernel module `cpufreq_stats' is active (after issuing `modprobe cpufreq_stats'), the average frequency (`avgf') and the average scaling percentage (`avgscal') is shown. Otherwise the current frequency (`curf') and the current scaling percentage (`curscal') is shown at the moment that the sample is taken. Notice that average values for frequency and scaling are shown in the CPU line for every CPU.
Frequency scaling statistics are only gathered for systems with maximum 8 CPUs, since gathering of these values per CPU is very time consuming.

If the screen-width does not allow all of these counters, only a relevant subset is shown.

CPU load information.
This line contains the load average figures reflecting the number of threads that are available to run on a CPU (i.e. part of the runqueue) or that are waiting for disk I/O. These figures are averaged over 1 (`avg1'), 5 (`avg5') and 15 (`avg15') minutes.
Furthermore the number of context switches (`csw'), the number of serviced interrupts (`intr') and the number of available CPUs are shown.

If the screen-width does not allow all of these counters, only a relevant subset is shown.

GPU utilization (Nvidia).
Read the section GPU STATISTICS GATHERING in this document to find the details about the activation of the pmdanvidia daemon.

In the first column of every line, the bus-id (last nine characters) and the GPU number are shown. The subsequent columns show the percentage of time that one or more kernels were executing on the GPU (`gpubusy'), the percentage of time that global (device) memory was being read or written (`membusy'), the occupation percentage of memory (`memocc'), the total memory (`total'), the memory being in use at the moment of the sample (`used'), the average memory being in use during the sample time (`usavg'), the number of processes being active on the GPU at the moment of the sample (`#proc'), and the type of GPU.

If the screen-width does not allow all of these counters, only a relevant subset is shown.
The number of lines showing the GPUs can be limited.

Memory occupation (two lines).
These lines contain the total amount of physical memory (`tot'), the amount of memory which is currently free (`free'), the amount of memory in use as page cache including the total resident shared memory (`cache'), the amount of memory within the page cache that has to be flushed to disk (`dirty'), the amount of memory used for filesystem meta data (`buff'), the amount of memory being used for kernel mallocs (`slab'), the amount of slab memory that is reclaimable (`slrec'), the resident size of shared memory including tmpfs (`shmem'), the resident size of shared memory (`shrss') the amount of shared memory that is currently swapped (`shswp'), the amount of memory that is currently used for page tables (`pgtab'), the number of NUMA nodes in this system (`numnode'), the amount of memory that is currently claimed by vmware's balloon driver (`vmbal'), the amount of memory that is currently claimed by the ARC (cache) of ZFSonlinux (`zfarc'), the amount of memory that is claimed for huge pages (`hptot'), the amount of huge page memory that is really in use (`hpuse'), the amount of memory that is used for TCP sockets (`tcps'), and the amount of memory that is used for UDP sockets (`udps').

If the screen-width does not allow all of these counters, only a relevant subset is shown.

Swap occupation and overcommit info.
This line contains the total amount of swap space on disk (`tot') and the amount of free swap space (`free'), the size of the swap cache (`swcac'), the total size of compressed storage in zswap (`zpool'), the total size of the compressed pages stored in zswap (`zstor'), the total size of the memory used for KSM (`ksuse', i.e. shared), and the total size of the memory saved (deduped) by KSM (`kssav', i.e. sharing).
Furthermore the committed virtual memory space (`vmcom') and the maximum limit of the committed space (`vmlim', which is by default swap size plus 50% of memory size) is shown. The committed space is the reserved virtual space for all allocations of private memory space for processes. The kernel only verifies whether the committed space exceeds the limit if strict overcommit handling is configured (vm.overcommit_memory is 2).
Memory utilization per NUMA node (not shown for single NUMA node).
This line shows the total amount of physical memory of this node (`tot'), the amount of free memory (`free'), the amount of memory for cached file data (`file'), modified cached file data (`dirty'), recently used memory (`activ'), less recently used memory (`inact'), memory being used for kernel mallocs (`slab'), the amount of slab memory that is reclaimable (`slrec'), shared memory including tmpfs (`shmem'), total huge pages (`hptot') and the fragmentation percentage (`frag').
CPU utilization per NUMA node (not shown for single NUMA node).
This line shows the utilization percentages of all CPUs related to this NUMA node, categorized for system mode (`sys'), user mode (`user'), user mode for niced processes (`niced'), idle mode (`idle'), wait mode (`w' preceded by the node number), irq mode (`irq'), softirq mode (`sirq'), steal mode (`steal'), and guest mode (`guest') overlapping user mode.
Paging frequency.
This line contains the number of scanned pages (`scan') due to the fact that free memory drops below a particular threshold, the number of times that the kernel tries to reclaim pages due to an urgent need (`stall'), the number of process stalls to run memory compaction to allocate huge pages (`compact'), the number of NUMA pages migrated (`numamig'), and the total number of memory pages migrated successfully e.g. between NUMA nodes or for compaction (`migrate') are shown.
Also the number of memory pages the system read from block devices (`pgin'), the number of memory pages the system wrote to block devices (`pgout'), the number of memory pages the system read from swap space (`swin'), the number of memory pages the system wrote to swap space (`swout'), and the number of out-of-memory kills (`oomkill').
Pressure Stall Information.
This line contains percentages about resource pressure related to CPU, memory and I/O. Certain percentages refer to 'some' meaning that some processes/threads were delayed due to resource overload. Other percentages refer to 'full' meaning a loss of overall throughput due to resource overload.
The values `cpusome', `memsome', `memfull', `iosome' and `iofull' show the pressure percentage during the entire interval.
The values `cs' (cpu some), `ms' (memory some), `mf' (memory full), `is' (I/O some) and `if' (I/O full) each show three percentages separated by slashes: pressure percentage over the last 10, 60 and 300 seconds.
Logical volume/multiple device/disk utilization.
Per active unit one line is produced, sorted on unit activity. Such line shows the name (e.g. VolGroup00-lvtmp for a logical volume or sda for a hard disk), the percentage of elapsed time during which I/O requests were issued to the device (`busy') (note that for devices serving requests in parallel, such as RAID arrays, SSD and NVMe, this number does not reflect their performance limits), the number of read requests issued (`read'), the number of write requests issued (`write'), the number of discard requests issued (`discrd') if supported by kernel version, the number of KiBytes per read (`KiB/r'), the number of KiBytes per write (`KiB/w'), the number of KiBytes per discard (`KiB/d') if supported by kernel version, the number of MiBytes per second throughput for reads (`MBr/s'), the number of MiBytes per second throughput for writes (`MBw/s'), requests issued to the device driver but not completed (`inflt'), the average queue depth (`avq') and the average number of milliseconds needed by a request (`avio') for seek, latency and data transfer.
If the screen-width does not allow all of these counters, only a relevant subset is shown.

The number of lines showing the units can be limited per class (LVM, MDD or DSK) with the 'l' key or statically (see separate man-page of pcp-atoprc(5)). By specifying the value 0 for a particular class, no lines will be shown any more for that class.

Network Filesystem (NFS) mount at the client side.
For each NFS-mounted filesystem, a line is shown that contains the mounted server directory, the name of the server (`srv'), the total number of bytes physically read from the server (`read') and the total number of bytes physically written to the server (`write'). Data transfer is subdivided in the number of bytes read via normal read() system calls (`nread'), the number of bytes written via normal read() system calls (`nwrit'), the number of bytes read via direct I/O (`dread'), the number of bytes written via direct I/O (`dwrit'), the number of bytes read via memory mapped I/O pages (`mread'), and the number of bytes written via memory mapped I/O pages (`mwrit').
Network Filesystem (NFS) client side counters.
This line contains the number of RPC calls issues by local processes (`rpc'), the number of read RPC calls (`read`) and write RPC calls (`rpwrite') issued to the NFS server, the number of RPC calls being retransmitted (`retxmit') and the number of authorization refreshes (`autref').
Network Filesystem (NFS) server side counters.
This line contains the number of RPC calls received from NFS clients (`rpc'), the number of read RPC calls received (`cread`), the number of write RPC calls received (`cwrit'), the number of Megabytes/second returned to read requests by clients (`MBcr/s`), the number of Megabytes/second passed in write requests by clients (`MBcw/s`), the number of network requests handled via TCP (`nettcp'), the number of network requests handled via UDP (`netudp'), the number of reply cache hits (`rchits'), the number of reply cache misses (`rcmiss') and the number of uncached requests (`rcnoca'). Furthermore some error counters indicating the number of requests with a bad format (`badfmt') or a bad authorization (`badaut'), and a counter indicating the number of bad clients (`badcln').
Network utilization (TCP/IP).
One line is shown for activity of the transport layer (TCP and UDP), one line for the IP layer and one line per active interface.
For the transport layer, counters are shown concerning the number of received TCP segments including those received in error (`tcpi'), the number of transmitted TCP segments excluding those containing only retransmitted octets (`tcpo'), the number of UDP datagrams received (`udpi'), the number of UDP datagrams transmitted (`udpo'), the number of active TCP opens (`tcpao'), the number of passive TCP opens (`tcppo'), the number of TCP output retransmissions (`tcprs'), the number of TCP input errors (`tcpie'), the number of TCP output resets (`tcpor'), the number of UDP no ports (`udpnp'), and the number of UDP input errors (`udpie').
If the screen-width does not allow all of these counters, only a relevant subset is shown.
These counters are related to IPv4 and IPv6 combined.

For the IP layer, counters are shown concerning the number of IP datagrams received from interfaces, including those received in error (`ipi'), the number of IP datagrams that local higher-layer protocols offered for transmission (`ipo'), the number of received IP datagrams which were forwarded to other interfaces (`ipfrw'), the number of IP datagrams which were delivered to local higher-layer protocols (`deliv'), the number of received ICMP datagrams (`icmpi'), and the number of transmitted ICMP datagrams (`icmpo').
If the screen-width does not allow all of these counters, only a relevant subset is shown.
These counters are related to IPv4 and IPv6 combined.

For every active network interface one line is shown, sorted on the interface activity. Such line shows the name of the interface and its busy percentage in the first column. The busy percentage for half duplex is determined by comparing the interface speed with the number of bits transmitted and received per second; for full duplex the interface speed is compared with the highest of either the transmitted or the received bits. When the interface speed can not be determined (e.g. for the loopback interface), `---' is shown instead of the percentage.
Furthermore the number of received packets (`pcki'), the number of transmitted packets (`pcko'), the line speed of the interface (`sp'), the effective amount of bits received per second (`si'), the effective amount of bits transmitted per second (`so'), the number of collisions (`coll'), the number of received multicast packets (`mlti'), the number of errors while receiving a packet (`erri'), the number of errors while transmitting a packet (`erro'), the number of received packets dropped (`drpi'), and the number of transmitted packets dropped (`drpo').
If the screen-width does not allow all of these counters, only a relevant subset is shown.
The number of lines showing the network interfaces can be limited.

Infiniband utilization.
For every active Infiniband port one line is shown, sorted on activity. Such line shows the name of the port and its busy percentage in the first column. The busy percentage is determined by taking the highest of either the transmitted or the received bits during the interval, multiplying that value by the number of lanes and comparing it against the maximum port speed.
Furthermore the number of received packets divided by the number of lanes (`pcki'), the number of transmitted packets divided by the number of lanes (`pcko'), the maximum line speed (`sp'), the effective amount of bits received per second (`si'), the effective amount of bits transmitted per second (`so'), and the number of lanes (`lanes').
If the screen-width does not allow all of these counters, only a relevant subset is shown.
The number of lines showing the Infiniband ports can be limited.

OUTPUT DESCRIPTION - PROCESS LEVEL

Following the system level information, the processes are shown from which the resource utilization has changed during the last interval. These processes might have used cpu time or issued disk or network requests. However a process is also shown if part of it has been paged out due to lack of memory (while the process itself was in sleep state).

Per process the following fields may be shown (in alphabetical order), depending on the current output mode as described in the section INTERACTIVE COMMANDS and depending on the current width of your window:

The average size of one read-action on disk.
The average size of one write-action on disk.
Total bandwidth for received TCP and UDP packets consumed by this process (bits-per-second). This value can be compared with the value `si' on interface level (used bandwidth per interface).
This information will only be shown when the pmdabcc(1) module `netproc' has been installed.
Total bandwidth for sent TCP and UDP packets consumed by this process (bits-per-second). This value can be compared with the value `so' on interface level (used bandwidth per interface).
This information will only be shown when the pmdabcc(1) module `netproc' has been installed.
Aggregated block I/O delay, i.e. time waiting for disk I/O.
Path name of the cgroup (version 2) to which this process belongs. This path name is relative to the cgroup root directory, which is usually `/sys/fs/cgroup'.
Container ID of 12 hexadecimal digits, referring to the container in which the process/thread is running. If a process has been started and finished during the last interval, a `?' is shown because the container ID is not part of the standard process accounting record.
The name of the process. This name can be surrounded by "less/greater than" signs (`<name>') which means that the process has finished during the last interval.
Behind the abbreviation `CMD' in the header line, the current page number and the total number of pages of the process/thread list are shown.
The full command line of the process (including arguments). If the length of the command line exceeds the length of the screen line, the arrow keys -> and <- can be used for horizontal scroll.
Behind the verb `COMMAND-LINE' in the header line, the current page number and the total number of pages of the process/thread list are shown.
The occupation percentage of this process related to the available capacity for this resource on system level.
The identification of the CPU the (main) thread is running on or has recently been running on.
Container ID (OpenVZ). If a process has been started and finished during the last interval, a `?' is shown because the container ID is not part of the standard process accounting record.
The occupation percentage of this process related to the total load that is produced by all processes (i.e. total disk accesses by all processes during the last interval).
This information is shown when per process "storage accounting" is active in the kernel.
Effective group-id under which this process executes.
Date that the process has been finished. If the process is still running, this field shows `active'.
Time that the process has been finished. If the process is still running, this field shows `active'.
Virtual environment identified (OpenVZ only).
Effective user-id under which this process executes.
The exit code of a terminated process (second position of column `ST' is E) or the fatal signal number (second position of column `ST' is S or C).
Filesystem group-id under which this process executes.
Filesystem user-id under which this process executes.
When the pmdanvidia daemon does not run with root privileges, the GPU percentage reflects the GPU memory occupation percentage (memory of all GPUs is 100%).
When the pmdanvidia daemon runs with root privileges, the GPU percentage reflects the GPU busy percentage.
Busy percentage on all GPUs (one GPU is 100%).
When the pmdanvidia daemon does not run with root privileges, this value is not available.
Comma-separated list of GPUs used by the process during the interval. When the comma-separated list exceeds the width of the column, a hexadecimal value is shown.
The virtual amount of memory being locked (i.e. non-swappable) by this process (or user).
The number of page faults issued by this process that have been solved by creating/loading the requested memory page.
The occupation percentage of this process related to the available capacity for this resource on system level.
Average memory occupation during the interval on all used GPUs.
Busy percentage of memory on all GPUs (one GPU is 100%), i.e. the time needed for read and write accesses on memory.
When the pmdanvidia daemon does not run with root privileges, this value is not available.
Memory occupation at the moment of the sample on all used GPUs.
The number of page faults issued by this process that have been solved by reclaiming the requested memory page from the free list of pages.
The occupation percentage of this process related to the total load that is produced by all processes (i.e. consumed network bandwidth of all processes during the last interval).
This information will only be shown when the pmdabcc(1) module `netproc' has been installed.
The more or less static priority that can be given to a process on a scale from -20 (high priority) to +19 (low priority).
The number of active and terminated processes accumulated for this user or program.
Process-id.
The policies 'norm' (normal, which is SCHED_OTHER), 'btch' (batch) and 'idle' refer to timesharing processes. The policies 'fifo' (SCHED_FIFO) and 'rr' (round robin, which is SCHED_RR) refer to realtime processes.
Parent process-id.
The process' priority ranges from 0 (highest priority) to 139 (lowest priority). Priority 0 to 99 are used for realtime processes (fixed priority independent of their behavior) and priority 100 to 139 for timesharing processes (variable priority depending on their recent CPU consumption and the nice value).
The proportional memory size of this process (or user).
Every process shares resident memory with other processes. E.g. when a particular program is started several times, the code pages (text) are only loaded once in memory and shared by all incarnations. Also the code of shared libraries is shared by all processes using that shared library, as well as shared memory and memory-mapped files. For the PSIZE calculation of a process, the resident memory of a process that is shared with other processes is divided by the number of sharers. This means, that every process is accounted for a proportional part of that memory. Accumulating the PSIZE values of all processes in the system gives a reliable impression of the total resident memory consumed by all processes.
Since gathering of all values that are needed to calculate the PSIZE is a very time-consuming task, the 'R' key (or '-R' flag) should be active. Gathering these values also requires superuser privileges (otherwise '?K' is shown in the output).
When the kernel maintains standard io statistics (>= 2.6.20):
The read data transfer issued physically on disk (so reading from the disk cache is not accounted for).
Unfortunately, the kernel aggregates the data transfer of a process to the data transfer of its parent process when terminating, so you might see transfers for (parent) processes like cron, bash or init, that are not really issued by them.
Runqueue delay, i.e. time spent waiting on a runqueue.
The real group-id under which the process executes.
The amount of resident memory that the process has grown during the last interval. A resident growth can be caused by touching memory pages which were not physically created/loaded before (load-on-demand). Note that a resident growth can also be negative e.g. when part of the process is paged out due to lack of memory or when the process frees dynamically allocated memory. For a process which started during the last interval, the resident growth reflects the total resident size of the process at that moment.
The total resident memory usage consumed by this process (or user). Notice that the RSIZE of a process includes all resident memory used by that process, even if certain memory parts are shared with other processes (see also the explanation of PSIZE).
Realtime priority according the POSIX standard. Value can be 0 for a timesharing process (policy 'norm', 'btch' or 'idle') or ranges from 1 (lowest) till 99 (highest) for a realtime process (policy 'rr' or 'fifo').
The real user-id under which the process executes.
The current state of the (main) thread: `R' for running (currently processing or in the runqueue), `S' for sleeping interruptible (wait for an event to occur), `D' for sleeping non-interruptible, `Z' for zombie (waiting to be synchronized with its parent process), `T' for stopped (suspended or traced), `W' for swapping, and `E' (exit) for processes which have finished during the last interval.
The saved group-id of the process.
The status of a process.
The first position indicates if the process has been started during the last interval (the value N means 'new process').

The second position indicates if the process has been finished during the last interval.
The value E means 'exit' on the process' own initiative; the exit code is displayed in the column `EXC'.
The value S means that the process has been terminated unvoluntarily by a signal; the signal number is displayed in the in the column `EXC'.
The value C means that the process has been terminated unvoluntarily by a signal, producing a core dump in its current directory; the signal number is displayed in the column `EXC'.

The start date of the process.
The start time of the process.
The saved user-id of the process.
The swap space consumed by this process (or user).
CPU time consumption of this process in system mode (kernel mode), usually due to system call handling.
The average size of a received TCP buffer in bytes. This information will only be shown when the BCC PMDA is active and the `netproc' module is enabled.
The number of tcp_recvmsg()/tcp_cleanup_rbuf() calls from this process. This information will only be shown when the BCC PMDA is active and the `netproc' module is enabled.
The average size of a TCP buffer requested to be transmitted in bytes. This information will only be shown when the BCC PMDA is active and the `netproc' module is enabled.
The number of tcp_sendmsg() calls from this process. This information will only be shown when the BCC PMDA is active and the `netproc' module is enabled.
Total number of threads within this process. All related threads are contained in a thread group, represented by pcp-atop as one line or as a separate line when the 'y' key (or -y flag) is active.
Thread-id. All threads within a process run with the same PID but with a different TID. This value is shown for individual threads in multi-threaded processes (when using the key 'y').
Number of threads within this process that are in the state 'running' (R).
Number of threads within this process that are in the state 'interruptible sleeping' (S).
Number of threads within this process that are in the state 'uninterruptible sleeping' (D).
The average size of a received UDP buffer in bytes. This information will only be shown when the BCC PMDA is active and the `netproc' module is enabled.
The number of udp_recvmsg()/skb_consume_udp() calls from this process. This information will only be shown when the BCC PMDA is active and the `netproc' module is enabled.
The average size of a UDP buffer requested to be transmitted in bytes. This information will only be shown when the BCC PMDA is active and the `netproc' module is enabled.
The number of udp_sendmsg() calls from this process. This information will only be shown when the BCC PMDA is active and the `netproc' module is enabled.
CPU time consumption of this process in user mode, due to processing the own program text.
The virtual memory size of the private data used by this process (including heap and shared library data).
The amount of virtual memory that the process has grown during the last interval. A virtual growth can be caused by e.g. issuing a malloc() or attaching a shared memory segment. Note that a virtual growth can also be negative by e.g. issuing a free() or detaching a shared memory segment. For a process which started during the last interval, the virtual growth reflects the total virtual size of the process at that moment.
Virtual process-id (within an OpenVZ container). If a process has been started and finished during the last interval, a `?' is shown because the virtual process-id is not part of the standard process accounting record.
The total virtual memory usage consumed by this process (or user).
The virtual memory size of the (shared) text of all shared libraries used by this process.
The virtual memory size of the (private) stack used by this process
The virtual memory size of the (shared) text of the executable program.
Wait channel of thread in sleep state, i.e. the name of the kernel function in which the thread has been put asleep.
Since determining the name string of the kernel function is a relatively time-consuming task, the 'W' key (or '-W' flag) should be active.
When the kernel maintains standard io statistics (>= 2.6.20):
The write data transfer issued physically on disk (so writing to the disk cache is not accounted for). This counter is maintained for the application process that writes its data to the cache (assuming that this data is physically transferred to disk later on). Notice that disk I/O needed for swapping is not taken into account.
Unfortunately, the kernel aggregates the data transfer of a process to the data transfer of its parent process when terminating, so you might see transfers for (parent) processes like cron, bash or init, that are not really issued by them.
When the kernel maintains standard io statistics (>= 2.6.20):
The write data transfer previously accounted for this process or another process that has been cancelled. Suppose that a process writes new data to a file and that data is removed again before the cache buffers have been flushed to disk. Then the original process shows the written data as WRDSK, while the process that removes/truncates the file shows the unflushed removed data as WCANCL.

With the flag -P followed by a list of one or more labels (comma-separated), parseable output is produced for each sample. The labels that can be specified for system-level statistics correspond to the labels (first verb of each line) that can be found in the interactive output: "CPU", "cpu", "CPL", "GPU", "MEM", "SWP", "PAG", "PSI", "LVM", "MDD", "DSK", "NFM", "NFC", "NFS", "NET", "IFB", "LLC", "NUM" and "NUC".
For process-level statistics special labels are introduced: "PRG" (general), "PRC" (cpu), "PRE" (GPU), "PRM" (memory), "PRD" (disk, only if "storage accounting" is active).
With the label "ALL", all system and process level statistics are shown.

The command and command line in the parseable output might contain spaces and are therefore by default surrounded by parenthesis. However, since a space is often used as separator between the fields by parsing tools, with the additional flag -Z it is possible to exchange the spaces in the command (line) by underscores and omit the parenthesis.

For every interval all requested lines are shown whereafter pcp-atop shows a line just containing the label "SEP" as a separator before the lines for the next sample are generated.
When a sample contains the values since boot, pcp-atop shows a line just containing the label "RESET" before the lines for this sample are generated.

The first part of each output-line consists of the following six fields: label (the name of the label), host (the name of this machine), epoch (the time of this interval as number of seconds since 1-1-1970), date (date of this interval in format YYYY/MM/DD), time (time of this interval in format HH:MM:SS), and interval (number of seconds elapsed for this interval).

The subsequent fields of each output-line depend on the label:

Subsequent fields: total number of clock-ticks per second for this machine, number of processors, consumption for all CPUs in system mode (clock-ticks), consumption for all CPUs in user mode (clock-ticks), consumption for all CPUs in user mode for niced processes (clock-ticks), consumption for all CPUs in idle mode (clock-ticks), consumption for all CPUs in wait mode (clock-ticks), consumption for all CPUs in irq mode (clock-ticks), consumption for all CPUs in softirq mode (clock-ticks), consumption for all CPUs in steal mode (clock-ticks), consumption for all CPUs in guest mode (clock-ticks) overlapping user mode, frequency of all CPUs and frequency percentage of all CPUs.
Subsequent fields: total number of clock-ticks per second for this machine, processor-number, consumption for this CPU in system mode (clock-ticks), consumption for this CPU in user mode (clock-ticks), consumption for this CPU in user mode for niced processes (clock-ticks), consumption for this CPU in idle mode (clock-ticks), consumption for this CPU in wait mode (clock-ticks), consumption for this CPU in irq mode (clock-ticks), consumption for this CPU in softirq mode (clock-ticks), consumption for this CPU in steal mode (clock-ticks), consumption for this CPU in guest mode (clock-ticks) overlapping user mode, frequency of all CPUs, frequency percentage of all CPUs, instructions executed by all CPUs and cycles for all CPUs.
Subsequent fields: number of processors, load average for last minute, load average for last five minutes, load average for last fifteen minutes, number of context-switches, and number of device interrupts.
Subsequent fields: GPU number, bus-id string, type of GPU string, GPU busy percentage during last second (-1 if not available), memory busy percentage during last second (-1 if not available), total memory size (KiB), used memory (KiB) at this moment, number of samples taken during interval, cumulative GPU busy percentage during the interval (to be divided by the number of samples for the average busy percentage, -1 if not available), cumulative memory busy percentage during the interval (to be divided by the number of samples for the average busy percentage, -1 if not available), and cumulative memory occupation during the interval (to be divided by the number of samples for the average occupation).
Subsequent fields: page size for this machine (in bytes), size of physical memory (pages), size of free memory (pages), size of page cache (pages), size of buffer cache (pages), size of slab (pages), dirty pages in cache (pages), reclaimable part of slab (pages), total size of vmware's balloon pages (pages), total size of shared memory (pages), size of resident shared memory (pages), size of swapped shared memory (pages), huge page size (in bytes), total size of huge pages (huge pages), size of free huge pages (huge pages), size of ARC (cache) of ZFSonlinux (pages), size of sharing pages for KSM (pages), size of shared pages for KSM (pages), size of memory used for TCP sockets (pages), size of memory used for UDP sockets (pages), and size of pagetables (pages).
Subsequent fields: page size for this machine (in bytes), size of swap (pages), size of free swap (pages), size of swap cache (pages), size of committed space (pages), limit for committed space (pages), size of the swap cache (pages), size of compressed pages stored in zswap (pages), and total size of compressed pool in zswap (pages).
Subsequent fields: page size for this machine (in bytes), number of page scans, number of allocstalls, 0 (future use), number of swapins, number of swapouts, number of oomkills (-1 when counter not present), number of process stalls to run memory compaction, number of pages successfully migrated in total, number of NUMA pages migrated, number of pages read from block devices, and number of pages written to block devices.
Subsequent fields: PSI statistics present on this system (n or y), CPU some avg10, CPU some avg60, CPU some avg300, CPU some accumulated microseconds during interval, memory some avg10, memory some avg60, memory some avg300, memory some accumulated microseconds during interval, memory full avg10, memory full avg60, memory full avg300, memory full accumulated microseconds during interval, I/O some avg10, I/O some avg60, I/O some avg300, I/O some accumulated microseconds during interval, I/O full avg10, I/O full avg60, I/O full avg300, and I/O full accumulated microseconds during interval.
For every logical volume/multiple device/hard disk one line is shown.
Subsequent fields: name, number of milliseconds spent for I/O, number of reads issued, number of sectors transferred for reads, number of writes issued, number of sectors transferred for write, number of discards issued (-1 if not supported), number of sectors transferred for discards, number of requests currently in flight (not yet completed), and the average queue depth while the disk was busy.
Subsequent fields: mounted NFS filesystem, total number of bytes read, total number of bytes written, number of bytes read by normal system calls, number of bytes written by normal system calls, number of bytes read by direct I/O, number of bytes written by direct I/O, number of pages read by memory-mapped I/O, and number of pages written by memory-mapped I/O.
Subsequent fields: number of transmitted RPCs, number of transmitted read RPCs, number of transmitted write RPCs, number of RPC retransmissions, and number of authorization refreshes.
Subsequent fields: number of handled RPCs, number of received read RPCs, number of received write RPCs, number of bytes read by clients, number of bytes written by clients, number of RPCs with bad format, number of RPCs with bad authorization, number of RPCs from bad client, total number of handled network requests, number of handled network requests via TCP, number of handled network requests via UDP, number of handled TCP connections, number of hits on reply cache, number of misses on reply cache, and number of uncached requests.
First, one line is produced for the upper layers of the TCP/IP stack.
Subsequent fields: the verb "upper", number of packets received by TCP, number of packets transmitted by TCP, number of packets received by UDP, number of packets transmitted by UDP, number of packets received by IP, number of packets transmitted by IP, number of packets delivered to higher layers by IP, number of packets forwarded by IP, number of input errors (UDP), number of noport errors (UDP), number of active opens (TCP), number of passive opens (TCP), number of passive opens (TCP), number of established connections at this moment (TCP), number of retransmitted segments (TCP), number of input errors (TCP), and number of output resets (TCP).

Next, one line is shown for every interface.
Subsequent fields: name of the interface, number of packets received by the interface, number of bytes received by the interface, number of packets transmitted by the interface, number of bytes transmitted by the interface, interface speed, and duplex mode (0=half, 1=full).

Subsequent fields: name of the InfiniBand interface, port number, number of lanes, maximum rate (Mbps), number of bytes received, number of bytes transmitted, number of packets received, and number of packets transmitted.
Subsequent fields: NUMA node number, page size for this machine (in bytes), the fragmentation percentage of this node, size of physical memory (pages), size of free memory (pages), recently (active) used memory (pages), less recently (inactive) used memory (pages), size of cached file data (pages), dirty pages in cache (pages), slab memory being used for kernel mallocs (pages), slab memory that is reclaimable (pages), shared memory including tmpfs (pages), and total huge pages (pages).
Subsequent fields: NUMA node number, number of processors for this node, consumption for node CPUs in system mode (clock-ticks), consumption for node CPUs in user mode (clock-ticks), consumption for node CPUs in user mode for niced processes (clock-ticks), consumption for node CPUs in idle mode (clock-ticks), consumption for node CPUs in wait mode (clock-ticks), consumption for node CPUs in irq mode (clock-ticks), consumption for node CPUs in softirq mode (clock-ticks), consumption for node CPUs in steal mode (clock-ticks), and consumption for node CPUs in guest mode (clock-ticks) overlapping user mode.
For every process one line is shown.
Subsequent fields: PID (unique ID of task), name (between parenthesis or underscores for spaces), state, real uid, real gid, TGID (group number of related tasks/threads), total number of threads, exit code (in case of fatal signal: signal number + 256), start time (epoch), full command line (between parenthesis or underscores for spaces), PPID, number of threads in state 'running' (R), number of threads in state 'interruptible sleeping' (S), number of threads in state 'uninterruptible sleeping' (D), effective uid, effective gid, saved uid, saved gid, filesystem uid, filesystem gid, elapsed time of terminated process (hertz), is_process (y/n), OpenVZ virtual pid (VPID), OpenVZ container id (CTID), Container id (CID), indication if the task is newly started during this interval ('N'), and cgroup v2 path name (between parenthesis or underscores for spaces).
For every process one line is shown.
Subsequent fields: PID, name (between parenthesis or underscores for spaces), state, total number of clock-ticks per second for this machine, CPU-consumption in user mode (clockticks), CPU-consumption in system mode (clockticks), nice value, priority, realtime priority, scheduling policy, current CPU, sleep average, TGID (group number of related tasks/threads), is_process (y/n), runqueue delay in nanoseconds for this thread or for all threads (in case of process), wait channel of this thread (between parenthesis or underscores for spaces), block I/O delay (clockticks).
For every process one line is shown.
Subsequent fields: PID, name (between parenthesis or underscores for spaces), process state, GPU state (A for active, E for exited, N for no GPU user), number of GPUs used by this process, bitlist reflecting used GPUs, GPU busy percentage during interval, memory busy percentage during interval, memory occupation (KiB) at this moment cumulative memory occupation (KiB) during interval, and number of samples taken during interval.
For every process one line is shown.
Subsequent fields: PID, name (between parenthesis or underscores for spaces), state, page size for this machine (in bytes), virtual memory size (Kbytes), resident memory size (Kbytes), shared text memory size (Kbytes), virtual memory growth (Kbytes), resident memory growth (Kbytes), number of minor page faults, number of major page faults, virtual library exec size (Kbytes), virtual data size (Kbytes), virtual stack size (Kbytes), swap space used (Kbytes), TGID (group number of related tasks/threads), is_process (y/n), proportional set size (Kbytes) if in 'R' option is specified and virtually locked memory space (Kbytes).
For every process one line is shown.
Subsequent fields: PID, name (between parenthesis or underscores for spaces), state, obsoleted kernel patch installed ('n'), standard io statistics used ('y' or 'n'), number of reads on disk, cumulative number of sectors read, number of writes on disk, cumulative number of sectors written, cancelled number of written sectors, TGID (group number of related tasks/threads), obsoleted value ('n'), and is_process (y/n).
If the standard I/O statistics (>= 2.6.20) are not used, the disk I/O counters per process are not relevant. The counters 'number of reads on disk' and 'number of writes on disk' are obsoleted anyhow.
For every process one line is shown.
Subsequent fields: PID, name (between parenthesis or underscores for spaces), state, pmdabcc(1) module `netproc' loaded ('y' or 'n'), number of tcp_sendmsg() calls, cumulative size of TCP buffers requested to be transmitted, number of tcp_recvmsg()/tcp_cleanup_rbuf() calls, cumulative size of TCP buffers received, number of udp_sendmsg() calls, cumulative size of UDP buffers requested to be transmitted, number of udp_recvmsg()/skb_consume_udp() calls, cumulative size of UDP buffers transmitted, number of raw packets transmitted (obsolete, always 0), number of raw packets received (obsolete, always 0), TGID (group number of related tasks/threads) and is_process (y/n).

By sending the SIGUSR1 signal to pcp-atop a new sample will be forced, even if the current timer interval has not exceeded yet. The behavior is similar to pressing the `t` key in an interactive session.

By sending the SIGUSR2 signal to pcp-atop a final sample will be forced after which pcp-atop will terminate.

To monitor the current system load interactively with an interval of 5 seconds:

  pcp atop 5

To monitor the system load and write it to a file (in plain ASCII) with an interval of one minute during half an hour with active processes sorted on memory consumption:

  pcp atop -M 60 30 > /log/pcp-atop.mem

Store information about the system and process activity in a PCP archive folio with an interval of ten minutes during an hour:

  pcp atop -w /tmp/pcp-atop 600 6

View the contents of this file interactively:

  pcp atop -r /tmp/pcp-atop

View the processor and disk utilization of this file in parseable format:

  pcp atop -PCPU,DSK -r /tmp/pcp-atop.folio

View the contents of today's standard logfile interactively:

  pcp atop -r

View the contents of the standard logfile of the day before yesterday interactively:

  pcp atop -r yy

View the contents of the standard logfile of 2014, June 7 from 02:00 PM onwards interactively:

  pcp atop -r 20140607 -b 14:00

pcp-atop is based on the source code of the atop(1) command from https://atoptool.nl, maintained by Gerlof Langeveld (gerlof.langeveld@atoptool.nl), and aims to be command line and output compatible with it as much as possible. Some features of that atop command are not available in pcp-atop.

Some features of pcp-atop (such as reporting on the Apache HTTP daemon, Infiniband, NFS client mounts, hardware event counts, GPU statistics and per-process TCP and UDP statistics) are only activated if the corresonding PCP metrics are available. Refer to the documentation for pmdaapache(1), pmdainfiniband(1), pmdanfsclient(1), pmdanvidia(1), pmdaperfevent(1) and pmdabcc(1) for further details on activating these metrics.

The semantics of the per-process network statistics deviate slightly from the atop(1) tool: instead of the number of TCP/UDP packets sent/received (which may be inaccurate due to TCP segmentation offload), pcp-atop shows the number of tcp_sendmsg()/udp_sendmsg()/etc. kernel calls per process.

/etc/atoprc
Configuration file containing system-wide default values. See related man-page.
~/.atoprc
Configuration file containing personal default values. See related man-page.

Environment variables with the prefix PCP_ are used to parameterize the file and directory names used by PCP. On each installation, the file /etc/pcp.conf contains the local values for these variables. The $PCP_CONF variable may be used to specify an alternative configuration file, as described in pcp.conf(5).

For environment variables affecting PCP tools, see pmGetOptions(3).

PCPIntro(1), pcp(1), pcp-atopsar(1), pmdaapache(1), pmdabcc(1), pmdainfiniband(1), pmdanfsclient(1), pmdanvidia(1), pmdaproc(1), mkaf(1), pmlogger(1), pmlogger_daily(1) and pcp-atoprc(5).

PCP Performance Co-Pilot