github/nohang
2
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

add ignore_positive_oom_score_adj

Browse Source
This commit is contained in:
Alexey Avramov 2019-07-15 11:22:16 +09:00
parent 20493b9a50
commit d233b17bd5
4 changed files with 16 additions and 395 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

1
Makefile
View File

@ -21,7 +21,6 @@ install:
install -m0644 ./nohang.conf $(DESTDIR)/$(PREFIX)/etc/nohang/nohang.conf install -m0644 ./nohang.conf $(DESTDIR)/$(PREFIX)/etc/nohang/nohang.conf
install -m0644 ./nohang.conf $(DESTDIR)/$(PREFIX)/etc/nohang/nohang.conf.default install -m0644 ./nohang.conf $(DESTDIR)/$(PREFIX)/etc/nohang/nohang.conf.default
install -m0644 ./my_desktop.conf $(DESTDIR)/$(PREFIX)/etc/nohang/my_desktop.conf
install -d $(DESTDIR)/$(PREFIX)/etc/logrotate.d install -d $(DESTDIR)/$(PREFIX)/etc/logrotate.d
install -m0644 ./nohang.logrotate $(DESTDIR)/$(PREFIX)/etc/logrotate.d/nohang install -m0644 ./nohang.logrotate $(DESTDIR)/$(PREFIX)/etc/logrotate.d/nohang

365
my_desktop.conf
View File

@ -1,365 +0,0 @@
This is nohang config file.
Lines starting with #, tabs and spaces are comments.
Lines starting with @ contain optional parameters.
All values are case sensitive.
Be careful: nohang doesn't forbid you to shoot yourself in the foot.
The configuration includes the following sections:
0. Common zram settings
1. Memory levels to respond to as an OOM threat
2. Response on PSI memory metrics
3. The frequency of checking the level of available memory
(and CPU usage)
4. The prevention of killing innocent victims
5. Impact on the badness of processes via matching their names, cgroups and
cmdlines with specified regular expressions
6. Customize corrective actions: the execution of a specific command
instead of sending the SIGTERM signal
7. GUI notifications:
- low memory warnings
- OOM prevention results
8. Output verbosity
9. Misc
Just read the description of the parameters and edit the values.
Please restart the program after editing the config.
###############################################################################
0. Common zram settings
See https://www.kernel.org/doc/Documentation/blockdev/zram.txt
You maybe need to set `ignore_zram = False` if you has a big zram disksize.
ignore_zram = False
###############################################################################
1. Thresholds below which a signal should be sent to the victim
Sets the available memory levels at or below which SIGTERM or SIGKILL
signals are sent. The signal will be sent if MemAvailable and
SwapFree (in /proc/meminfo) at the same time will drop below the
corresponding values. Can be specified in % (percent) and M (MiB).
Valid values are floating-point numbers from the range [0; 100] %.
MemAvailable levels.
mem_min_sigterm = 10 %
mem_min_sigkill = 5 %
SwapFree levels.
swap_min_sigterm = 15 %
swap_min_sigkill = 5 %
Specifying the total share of zram in memory, if exceeded the
corresponding signals are sent. As the share of zram in memory
increases, it may fall responsiveness of the system. 90 % is a
usual hang level, not recommended to set very high.
Can be specified in % and M. Valid values are floating-point
numbers from the range [0; 90] %.
zram_max_sigterm = 50 %
zram_max_sigkill = 60 %
###############################################################################
2. Response on PSI memory metrics (it needs Linux 4.20 and up)
About PSI:
https://facebookmicrosites.github.io/psi/
Disabled by default (ignore_psi = True).
ignore_psi = True
Choose a path to PSI file.
By default it monitors system-wide file: /proc/pressure/memory
You also can set file to monitor one cgroup slice.
For example:
psi_path = /sys/fs/cgroup/unified/user.slice/memory.pressure
psi_path = /sys/fs/cgroup/unified/system.slice/memory.pressure
psi_path = /sys/fs/cgroup/unified/system.slice/foo.service/memory.pressure
Execute the command
find /sys/fs/cgroup -name memory.pressure
to find available memory.pressue files (except /proc/pressure/memory).
(actual for cgroup2)
psi_path = /proc/pressure/memory
Valid psi_metrics are:
some_avg10
some_avg60
some_avg300
full_avg10
full_avg60
full_avg300
some_avg10 is most sensitive.
psi_metrics = some_avg10
sigterm_psi_threshold = 60
sigkill_psi_threshold = 90
>= 0, float
psi_excess_duration = 60
psi_post_action_delay = 90
###############################################################################
3. The frequency of checking the amount of available memory
(and CPU usage)
Coefficients that affect the intensity of monitoring. Reducing
the coefficients can reduce CPU usage and increase the periods
between memory checks.
Why three coefficients instead of one? Because the swap fill rate
is usually lower than the RAM fill rate.
It is possible to set a lower intensity of monitoring for swap
without compromising to prevent OOM and thus reduce the CPU load.
Default values are well for desktop. On servers without rapid
fluctuations in memory levels the values can be reduced.
Valid values are positive floating-point numbers.
rate_mem = 4000
rate_swap = 1500
rate_zram = 6000
See also https://github.com/rfjakob/earlyoom/issues/61
max_sleep = 3
min_sleep = 0.1
Sleep time if soft threshold exceeded.
over_sleep = 0.05
###############################################################################
4. The prevention of killing innocent victims
Valid values are integers from the range [0; 1000].
min_badness = 30
Valid values are non-negative floating-point numbers.
Min delay if a victim doesn't respond to SIGTERM in 10 ms.
min_delay_after_sigterm = 3
post_zombie_delay = 0.1
victim_cache_time = 10
Valid values are True and False.
decrease_oom_score_adj = True
Valid values are integers from the range [0; 1000].
oom_score_adj_max = 0
###############################################################################
5. Impact on the badness of processes via matching their names,
cmdlines or UIDs with regular expressions using re.search().
See https://en.wikipedia.org/wiki/Regular_expression and
https://en.wikipedia.org/wiki/Perl_Compatible_Regular_Expressions
Enabling this options slows down the search for the victim
because the names, cmdlines or UIDs of all processes
(except init and kthreads) are compared with the
specified regex patterns (in fact slowing down is caused by
reading all /proc/*/cmdline and /proc/*/status files).
Use script `oom-sort` from nohang package to view
names, cmdlines and UIDs of processes.
5.1 Matching process names with RE patterns
Syntax:
@BADNESS_ADJ_RE_NAME badness_adj /// RE_pattern
New badness value will be += badness_adj
It is possible to compare multiple patterns
with different badness_adj values.
Example:
@BADNESS_ADJ_RE_NAME -500 /// ^sshd$
5.2 Matching CGroup-line (v1 and v2) with RE patterns
@BADNESS_ADJ_RE_CGROUP_V1 -50 /// ^/system\.slice/
@BADNESS_ADJ_RE_CGROUP_V1 50 /// /foo\.service$
@BADNESS_ADJ_RE_CGROUP_V1 -50 /// ^/user\.slice/
@BADNESS_ADJ_RE_CGROUP_V2 100 /// ^/workload
5.3 Matching eUIDs with RE patterns
@BADNESS_ADJ_RE_UID -100 /// ^0$
5.4 Matching realpath with RE patterns
@BADNESS_ADJ_RE_REALPATH 20 /// ^/usr/bin/foo
5.5 Matching cmdlines with RE patterns
A good option that allows fine adjustment.
Prefer chromium tabs and electron-based apps
@BADNESS_ADJ_RE_CMDLINE 200 /// --type=renderer
Prefer firefox tabs (Web Content and WebExtensions)
@BADNESS_ADJ_RE_CMDLINE 100 /// -appomni
@BADNESS_ADJ_RE_CMDLINE -200 /// ^/usr/lib/virtualbox
5.6 Matching environ with RE patterns
@BADNESS_ADJ_RE_ENVIRON 100 /// USER=user
Note that you can control badness also via systemd units via
OOMScoreAdjust, see
www.freedesktop.org/software/systemd/man/systemd.exec.html#OOMScoreAdjust=
###############################################################################
6. Customize corrective actions.
TODO: docs
Syntax:
KEY REGEXP SEPARATOR COMMAND
@SOFT_ACTION_RE_NAME ^foo$ /// kill -SEGV $PID
@SOFT_ACTION_RE_NAME ^bash$ /// kill -9 $PID
@SOFT_ACTION_RE_CGROUP_V1 ^/system\.slice/ /// systemctl restart $SERVICE
@SOFT_ACTION_RE_CGROUP_V1 /foo\.service$ /// systemctl restart $SERVICE
$PID will be replaced by process PID.
$NAME will be replaced by process name.
$SERVICE will be replaced by .service if it exists (overwise it will be
relpaced by empty line)
###############################################################################
7. GUI notifications:
- OOM prevention results and
- low memory warnings
gui_notifications = True
Enable GUI notifications about the low level of available memory.
Valid values are True and False.
gui_low_memory_warnings = True
Execute the command instead of sending GUI notifications if the value is
not empty line. For example:
warning_exe = cat /proc/meminfo &
warning_exe =
Can be specified in % (percent) and M (MiB).
Valid values are floating-point numbers from the range [0; 100] %.
mem_min_warnings = 25 %
swap_min_warnings = 35 %
zram_max_warnings = 40 %
psi_avg_warnings = 100
Valid values are floating-point numbers from the range [1; 300].
min_time_between_warnings = 20
Ampersands (&) will be replaced with asterisks (*) in process
names and in commands.
###############################################################################
8. Verbosity
Display the configuration when the program starts.
Valid values are True and False.
print_config = False
Print memory check results.
Valid values are True and False.
print_mem_check_results = True
min_mem_report_interval = 300
Print sleep periods between memory checks.
Valid values are True and False.
print_sleep_periods = False
print_total_stat = True
print_proc_table = False
Valid values:
None
cgroup_v1
cgroup_v2
realpath
cmdline
environ
extra_table_info = None
print_victim_info = True
print_victim_cmdline = True
max_ancestry_depth = 4
debug_gui_notifications = False
separate_log = True
psi_debug = False
###############################################################################
9. Misc
max_post_sigterm_victim_lifetime = 10
post_kill_exe =
forbid_negative_badness = True
###############################################################################
Use cases, feature requests and any questions are welcome:
https://github.com/hakavlad/nohang/issues

27
nohang
View File

@ -68,8 +68,7 @@ def check_config():
log(' min_delay_after_sigterm: {} sec'.format(min_delay_after_sigterm)) log(' min_delay_after_sigterm: {} sec'.format(min_delay_after_sigterm))
log(' post_zombie_delay: {} sec'.format(post_zombie_delay)) log(' post_zombie_delay: {} sec'.format(post_zombie_delay))
log(' victim_cache_time: {} sec'.format(victim_cache_time)) log(' victim_cache_time: {} sec'.format(victim_cache_time))
log(' decrease_oom_score_adj: {}'.format(decrease_oom_score_adj)) log(' ignore_positive_oom_score_adj: {}'.format(ignore_positive_oom_score_adj))
log(' oom_score_adj_max: {}'.format(oom_score_adj_max))
log('5. Impact on the badness of processes') log('5. Impact on the badness of processes')
@ -555,10 +554,10 @@ def pid_to_badness(pid):
oom_score = int(rline1('/proc/' + pid + '/oom_score')) oom_score = int(rline1('/proc/' + pid + '/oom_score'))
badness = oom_score badness = oom_score
if decrease_oom_score_adj: if ignore_positive_oom_score_adj:
oom_score_adj = int(rline1('/proc/' + pid + '/oom_score_adj')) oom_score_adj = int(rline1('/proc/' + pid + '/oom_score_adj'))
if badness > oom_score_adj_max and oom_score_adj > 0: if oom_score_adj > 0:
badness = badness - oom_score_adj + oom_score_adj_max badness = badness - oom_score_adj
if regex_matching: if regex_matching:
name = pid_to_name(pid) name = pid_to_name(pid)
@ -2633,10 +2632,13 @@ print_mem_check_results = conf_parse_bool('print_mem_check_results')
print_sleep_periods = conf_parse_bool('print_sleep_periods') print_sleep_periods = conf_parse_bool('print_sleep_periods')
gui_low_memory_warnings = conf_parse_bool('gui_low_memory_warnings') gui_low_memory_warnings = conf_parse_bool('gui_low_memory_warnings')
gui_notifications = conf_parse_bool('gui_notifications') gui_notifications = conf_parse_bool('gui_notifications')
decrease_oom_score_adj = conf_parse_bool('decrease_oom_score_adj')
ignore_psi = conf_parse_bool('ignore_psi') ignore_psi = conf_parse_bool('ignore_psi')
ignore_zram = conf_parse_bool('ignore_zram') ignore_zram = conf_parse_bool('ignore_zram')
debug_gui_notifications = conf_parse_bool('debug_gui_notifications') debug_gui_notifications = conf_parse_bool('debug_gui_notifications')
ignore_positive_oom_score_adj = conf_parse_bool('ignore_positive_oom_score_adj')
(mem_min_sigterm_kb, mem_min_sigterm_mb, mem_min_sigterm_percent (mem_min_sigterm_kb, mem_min_sigterm_mb, mem_min_sigterm_percent
@ -2823,18 +2825,7 @@ else:
exit(1) exit(1)
if 'oom_score_adj_max' in config_dict:
oom_score_adj_max = string_to_int_convert_test(
config_dict['oom_score_adj_max'])
if oom_score_adj_max is None:
errprint('Invalid oom_score_adj_max value, not integer\nExit')
exit(1)
if oom_score_adj_max < 0 or oom_score_adj_max > 1000:
errprint('Invalid oom_score_adj_max value\nExit')
exit(1)
else:
errprint('oom_score_adj_max not in config\nExit')
exit(1)
if 'min_time_between_warnings' in config_dict: if 'min_time_between_warnings' in config_dict:

10
nohang.conf
View File

@ -109,7 +109,7 @@ sigkill_psi_threshold = 90
>= 0, float >= 0, float
psi_excess_duration = 60 psi_excess_duration = 60
psi_post_action_delay = 90 psi_post_action_delay = 60
############################################################################### ###############################################################################
@ -164,11 +164,7 @@ victim_cache_time = 10
Valid values are True and False. Valid values are True and False.
decrease_oom_score_adj = False ignore_positive_oom_score_adj = False
Valid values are integers from the range [0; 1000].
oom_score_adj_max = 0
############################################################################### ###############################################################################
@ -337,7 +333,7 @@ print_victim_info = True
print_victim_cmdline = False print_victim_cmdline = False
max_ancestry_depth = 4 max_ancestry_depth = 3
debug_gui_notifications = False debug_gui_notifications = False