Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

determine if nbresuse is running in docker or pod #66

Open
wants to merge 6 commits into
base: main
Choose a base branch
from
Open

determine if nbresuse is running in docker or pod #66

Changes from all commits
Commits
Show all changes
6 commits
Select commit Hold shift + click to select a range
ea6cde0
Determine if nbresuse running in docker
xianglei Sep 7, 2020
9677916
Merge pull request #1 from xianglei/xianglei-patched
xianglei Sep 7, 2020
8ac57ab
forgot import os module
xianglei Sep 7, 2020
72a3ba1
Update api.py
xianglei Sep 8, 2020
d4ec720
Update api.py
xianglei Sep 8, 2020
8aec768
fix calculate issue
xianglei Sep 10, 2020
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
71 changes: 59 additions & 12 deletions nbresuse/api.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -5,6 +5,7 @@
from notebook.base.handlers import IPythonHandler
from tornado import web
from tornado.concurrent import run_on_executor
import os

try:
# Traitlets >= 4.3.3
Expand All @@ -28,31 +29,46 @@ async def get(self):
all_processes = [cur_process] + cur_process.children(recursive=True)

# Get memory information
rss = sum([p.memory_info().rss for p in all_processes])

if callable(config.mem_limit):
mem_limit = config.mem_limit(rss=rss)
else: # mem_limit is an Int
mem_limit = config.mem_limit
# If running in a docker container or pod
dockerized = self.__if_dockerized()
if not dockerized:
# if running on physical server
rss = sum([p.memory_info().rss for p in all_processes])
if callable(config.mem_limit):
mem_limit = config.mem_limit(rss=rss)
else: # mem_limit is an Int
mem_limit = config.mem_limit
# unused rss, use system avail memory instead
else:
# if running in docker container or pod
mem_limit = self.__get_docker_physical_memory()
real_rss = self.__get_docker_real_rss_memory()
rss = real_rss

limits = {"memory": {"rss": mem_limit}}
if config.mem_limit and config.mem_warning_threshold != 0:
if config.mem_limit:
limits["memory"]["warn"] = (mem_limit - rss) < (
mem_limit * config.mem_warning_threshold
mem_limit * config.mem_warning_threshold
)

metrics = {"rss": rss, "limits": limits}

# Optionally get CPU information
if config.track_cpu_percent:
cpu_count = psutil.cpu_count()
cpu_percent = await self._get_cpu_percent(all_processes)
if not dockerized:
# original nbresuse
cpu_count = psutil.cpu_count()
cpu_percent = await self._get_cpu_percent(all_processes)
else:
# for docker use only
cpu_count = int(open('/sys/fs/cgroup/cpu/cpu.cfs_quota_us').read().strip()) / 100000
cpu_percent = await self._get_cpu_percent(all_processes) / cpu_count

if config.cpu_limit != 0:
limits["cpu"] = {"cpu": config.cpu_limit}
if config.cpu_warning_threshold != 0:
limits["cpu"]["warn"] = (config.cpu_limit - cpu_percent) < (
config.cpu_limit * config.cpu_warning_threshold
config.cpu_limit * config.cpu_warning_threshold
)

metrics.update(cpu_percent=cpu_percent, cpu_count=cpu_count)
Expand All @@ -68,5 +84,36 @@ def get_cpu_percent(p):
# about dead processes having no CPU usage
except:
return 0

return sum([get_cpu_percent(p) for p in all_processes])

@staticmethod
def __get_docker_physical_memory():
with open('/sys/fs/cgroup/memory/memory.limit_in_bytes') as limit:
mem_limit = int(limit.read().strip())
return mem_limit

@staticmethod
def __get_docker_memory_usage():
with open('/sys/fs/cgroup/memory/memory.usage_in_bytes') as usage:
all_used = int(usage.read().strip())
return all_used

@staticmethod
def __get_docker_real_rss_memory():
real_rss = 0
with open('/sys/fs/cgroup/memory/memory.stat') as stats:
for line in stats.readlines():
stat = line.split()
if stat[0] in ['rss', 'inactive_file', 'active_file']:
real_rss = real_rss + int(stat[1])
return real_rss

@staticmethod
def __if_dockerized():
if os.path.exists('/.dockerenv'):
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Is this a reliable way of detecting whether it is running in Docker? Does this file exist with other container engines?

Copy link

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

It's not present on Podman.
I found this stackoverflow question which provides a bit of background: https://superuser.com/questions/1021834/what-are-dockerenv-and-dockerinit

Copy link

@manics manics Dec 23, 2020 •
edited
Loading

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I just checked with Kubernetes K3S too, it's also not present.

However it looks like /sys/fs/cgroup/memory/memory.limit_in_bytes is set to <very large number> if there's no limit:

k3os-5708 [~]$ kubectl run tmp-shell --rm -i --tty --image busybox -- /bin/sh
If you don't see a command prompt, try pressing enter.
/ # cat /.dockerenv
cat: can't open '/.dockerenv': No such file or directory
/ # cat /sys/fs/cgroup/memory/memory.limit_in_bytes
9223372036854771712

k3os-5708 [~]$ kubectl run tmp-shell --rm -i --tty --image busybox --limits=memory=10Mi -- /bin/sh
If you don't see a command prompt, try pressing enter.
/ # cat /sys/fs/cgroup/memory/memory.limit_in_bytes
10485760

So one solution might be to see if /sys/fs/cgroup/memory/memory.limit_in_bytes exists, and if so use min(/sys/fs/cgroup/memory/memory.limit_in_bytes, total-system-memory), and skip the container check completely?

It's also worth noting that cgroups2 is different, I haven't figured out how to find the memory limit there.

# running in docker or pod
return True
else:
# running on physical server
return False