Python psutil
last modified September 16, 2024
In this article we show how to retrieve information on processes and system utilization.
The psutil is a Python module that provides a simple interface to retrieve information about running processes, system users, networks, disks, and much more. It's a powerful tool for system monitoring and administration tasks.
The psutil.Process().info() method is used to retrieve detailed
information about a specific process. It returns a dictionary containing various
attributes of the process, such as:
- pid - Process ID
- ppid - Parent process ID
- name - Process name
- exe - Executable path
- cmdline - Command line arguments used to start the process
- status - Process status (e.g., 'running', 'sleeping', 'zombie')
- create_time - Process creation time
- user - User name associated with the process
- nice - Process nice level
- io_counters - I/O counters (read/write bytes, system calls)
- cpu_times - CPU times (user, system, idle, etc.)
- memory_info - Memory usage information (RSS, VMS, etc.)
In the examples, we use the argparse module to parse command line
arguments.
List memory usage
The example lists processes with memory usage above the given value in MB.
import psutil
import argparse
parser = argparse.ArgumentParser()
parser.add_argument('memory', help='memory value in MB')
args = parser.parse_args()
mem_limit = int(args.memory) * 1024 * 1024
p_it = psutil.process_iter(['name', 'memory_info'])
large_mem_ps = [{'pid': p.pid, 'name': p.info['name'], 'mem': p.info['memory_info'].rss}
for p in p_it if p.info['memory_info'].rss > mem_limit]
for p in large_mem_ps:
print(f'{p['pid']} {p['name']} {(p['mem'] / (1024 * 1024)):.2f}')
The program accepts one positional argument: the memory value in MB.
p_it = psutil.process_iter(['name', 'memory_info'])
The process_iter function returns a generator yielding a Process
instance for all running processes. We pick the name of the process and
memory_info attributes from the generated data.
large_mem_ps = [{'pid': p.pid, 'name': p.info['name'], 'mem': p.info['memory_info'].rss}
for p in p_it if p.info['memory_info'].rss > mem_limit]
We select all processes that use memory about our limit.
for p in large_mem_ps:
print(f'{p['pid']} {p['name']} {(p['mem'] / (1024 * 1024)):.2f}')
We go through the list and print the data.
CPU usage
The cpu_times attribute provides information about the CPU time
consumed by a process.
import psutil
processes = [
{'pid': p.pid, 'name': p.info['name'],
'sum_cpu_t': sum(p.info['cpu_times'])}
for p in psutil.process_iter(['name', 'cpu_times'])
]
sorted_ps = sorted(processes, key=lambda p: p['sum_cpu_t'])
# print the last five
for e in sorted_ps[-5:]:
print(f'pid: {e['pid']} name: {e['name']} cpu time: {e['sum_cpu_t']}')
The program lists top five processes by their CPU usage.
processes = [
{'pid': p.pid, 'name': p.info['name'],
'sum_cpu_t': sum(p.info['cpu_times'])}
for p in psutil.process_iter(['name', 'cpu_times'])
]
We get a list of processes, including their process ids, names, and CPU usage.
sorted_ps = sorted(processes, key=lambda p: p['sum_cpu_t'])
We sort the processes by the sum of their CPU times (user + system).
# print the last five
for e in sorted_ps[-5:]:
print(f'pid: {e['pid']} name: {e['name']} cpu time: {e['sum_cpu_t']}')
Since the list is sorted in asceding order, we print the last five of them.
Check log files of processes
The open_files attribute returns a list of file descriptors
associated with the process. Each file descriptor is represented as a tuple
containing the following information:
- path - The path of the file or device
- fd - The file descriptor number
- position - The current position within the file
- flags - Flags indicating the file access mode (read, write, etc.)
import psutil
for p in psutil.process_iter(['name', 'open_files']):
for file in p.info['open_files'] or []:
if file.path.endswith('.log'):
print(f"{p.pid:<5} {p.info['name']:<20} {file.path}")
The program lists all processes that have opened log files.
List processes
The following script list processes. We use a third-party rich library to display the output in a table on console.
import psutil
import argparse
from datetime import datetime
from rich import box
from rich.console import Console
from rich.table import Table
from datetime import date
def parse_arguments():
parser = argparse.ArgumentParser()
group = parser.add_mutually_exclusive_group(required=True)
group.add_argument('-a', '--all', action='store_true',
help='show all processes')
group.add_argument('-n', '--name', help='show info about process name')
args = parser.parse_args()
return args.all, args.name
def list_process(name):
now = f'{date.today()}'
table = Table(title=f'Process', box=box.MINIMAL,
caption=now, caption_justify='left')
table.add_column('id', style='cyan')
table.add_column('process name', style='grey69')
table.add_column('username')
table.add_column('create time', style='blue')
table.add_column('memory', style='green')
process_count = 0
for p in psutil.process_iter():
if name in p.name().lower():
ctime = datetime.fromtimestamp(p.create_time())
memory_percent = p.memory_percent()
table.add_row(f'{p.pid}', p.name(), p.username(),
ctime.isoformat(), f'{memory_percent:.2f}')
process_count += 1
if process_count > 0:
console = Console()
console.print(table, justify='center')
else:
print('no such process found')
def list_all_processes():
now = f'{date.today()}'
table = Table(title='Processes', box=box.MINIMAL,
caption=now, caption_justify='left')
table.add_column('id', style='cyan')
table.add_column('process name', style='grey69')
pnames = []
for p in psutil.process_iter():
pnames.append(p.name())
table.add_row(f'[bold]{p.pid}', f'[bold]{p.name()}')
console = Console()
console.print(table, justify='center')
print(len(pnames), 'processes')
print(len(set(pnames)), 'apps')
all_f, name = parse_arguments()
if all_f:
list_all_processes()
elif name:
list_process(name)
The script takes two options: the -a/-all lists all processes while
the -n/--name shows informatio about a specific process. These
options are mutually exclusive.
Source
In this article we have used predicates in Python.
Author
List all Python tutorials.