[checkmk-commits] 6652 Predictive Levels for CPU utilization in Linux/Unix

Module: check_mk Branch: master Commit: da9ca35d2a2d6b89ef05ac88928be399ff393431 URL: http://git.mathias-kettner.de/git/?p=check_mk.git;a=commit;h=da9ca35d2a2d6b… Author: Óscar Nájera &lt;on(a)mathias-kettner.de&gt; Date: Mon Nov 19 15:31:10 2018 +0100 6652 Predictive Levels for CPU utilization in Linux/Unix Linux CPU utilization can now be monitored using predictive levels for the thresholds to warning and critical states. This confiruration is directly accessible in WATO. CKM-1284 Change-Id: Ie6d4cfb55541e3a9ea7794d7e867806d8b080b4f --- .werks/6652 | 12 +++ checks/kernel | 1 - cmk/gui/plugins/wato/check_parameters.py | 112 ++++++++++----------- .../cmk/gui/plugins/wato/test_check_parameters.py | 24 ++++- 4 files changed, 90 insertions(+), 59 deletions(-) diff --git a/.werks/6652 b/.werks/6652 new file mode 100644 index 0000000..8933212 --- /dev/null +++ b/.werks/6652 @@ -0,0 +1,12 @@ +Title: Predictive Levels for CPU utilization in Linux/Unix +Level: 1 +Component: checks +Compatible: compat +Edition: cre +Version: 1.6.0i1 +Date: 1543243736 +Class: feature + +Linux CPU utilization can now be monitored using predictive levels for the +thresholds to warning and critical states. This confiruration is directly +accessible in WATO. diff --git a/checks/kernel b/checks/kernel index 1edc2c0..06ffad7 100644 --- a/checks/kernel +++ b/checks/kernel @@ -85,7 +85,6 @@ check_info["kernel"] = { 'group': 'vm_counter', } -#. # .--kernel.util--CPU Utilization----------------------------------------. # | _ _ _ _ _ _ _ _ | # | | | | | |_(_) (_)______ _| |_(_) ___ _ __ | diff --git a/cmk/gui/plugins/wato/check_parameters.py b/cmk/gui/plugins/wato/check_parameters.py index c416dce..0044b7b 100644 --- a/cmk/gui/plugins/wato/check_parameters.py +++ b/cmk/gui/plugins/wato/check_parameters.py @@ -9358,18 +9358,16 @@ cpu_util_common_elements = [ "exceeding a utilization threshold over an extended period of time." "This is currently only supported on linux and windows agents " "as well as devices monitored through the host-resource mib"))), - ( - "core_util_graph", - Checkbox( - title=_("Graphs for individual cores"), - label=_("Enable performance graph for utilization of individual cores"), - help=_("This adds another graph to the performance CPU utilization " - "details page, showing utilization of individual cores. " - "Please note that this graph may be impractical on " - "device with very many cores. " - "This is currently only supported on linux and windows agents " - "as well as devices monitored through the host-resource mib")), - ), + ("core_util_graph", + Checkbox( + title=_("Graphs for individual cores"), + label=_("Enable performance graph for utilization of individual cores"), + help=_("This adds another graph to the performance CPU utilization " + "details page, showing utilization of individual cores. " + "Please note that this graph may be impractical on " + "device with very many cores. " + "This is currently only supported on linux and windows agents " + "as well as devices monitored through the host-resource mib"))), ] register_check_parameters( @@ -9392,62 +9390,62 @@ register_check_parameters( default_value=15, label=_("Compute average over last "), )), - ( - "levels", - Levels( - title=_("Levels on total CPU utilization"), - unit="%", - default_levels=(85, 90), - default_difference=(5, 8), - default_value=None, - ), - ), + ("levels", + Levels( + title=_("Levels on total CPU utilization"), + unit="%", + default_levels=(85, 90), + default_difference=(5, 8), + default_value=None, + )), ] + cpu_util_common_elements, ), None, match_type="dict", ) + +def transform_cpu_iowait(params): + if isinstance(params, tuple): + return {"iowait": params} + + return params + + register_check_parameters( subgroup_os, "cpu_iowait", _("CPU utilization on Linux/UNIX"), Transform( - Dictionary( - elements=[ - ( - "iowait", - Tuple( - title=_("Levels on disk wait (IO wait)"), - elements=[ - Percentage(title=_("Warning at a disk wait of"), default_value=30.0), - Percentage(title=_("Critical at a disk wait of"), default_value=50.0) - ], - help= - _("The CPU utilization sums up the percentages of CPU time that is used " - "for user processes, kernel routines (system), disk wait (sometimes also " - "called IO wait) or nothing (idle). " - "Currently you can only set warning/critical levels to the disk wait. This " - "is the total percentage of time all CPUs have nothing else to do then waiting " - "for data coming from or going to disk. If you have a significant disk wait " - "the the bottleneck of your server is IO. Please note that depending on the " - "applications being run this might or might not be totally normal.")), - ), - ("util", - Tuple( - title=_("Levels on total CPU utilization"), - elements=[ - Percentage(title=_("Warning at a utilization of"), default_value=90.0), - Percentage(title=_("Critical at a utilization of"), default_value=95.0) - ], - help= - _("Here you can set levels on the total CPU utilization, i.e. the sum of " - "<i>system</i>, <i>user</i> and <i>iowait</i>. The levels are always applied " - "on the average utiliazation since the last check - which is usually one minute." - ), - )), - ] + cpu_util_common_elements,), - forth=lambda old: not isinstance(old, dict) and {"iowait": old} or old, + Dictionary(elements=[ + ("util", + Levels( + title=_("Levels on total CPU utilization"), + unit="%", + default_levels=(90, 95), + default_difference=(5, 8), + default_value=None, + help= + _("The CPU utilization sums up the percentages of CPU time that is used " + "for user processes, kernel routines (system), disk wait (sometimes also " + "called IO wait) or nothing (idle). The levels are always applied " + "on the average utiliazation since the last check - which is usually one minute." + ), + )), + ("iowait", + Tuple( + title=_("Levels on disk wait (IO wait)"), + elements=[ + Percentage(title=_("Warning at a disk wait of"), default_value=30.0), + Percentage(title=_("Critical at a disk wait of"), default_value=50.0) + ], + help= + _("The disk wait is the total percentage of time all CPUs have nothing else to do but waiting " + "for data coming from or going to disk. If you have a significant disk wait " + "the the bottleneck of your server is IO. Please note that depending on the " + "applications being run this might or might not be totally normal."))), + ] + cpu_util_common_elements), + forth=transform_cpu_iowait, ), None, "dict", diff --git a/tests/unit/cmk/gui/plugins/wato/test_check_parameters.py b/tests/unit/cmk/gui/plugins/wato/test_check_parameters.py index afe72e3..e4f666c 100644 --- a/tests/unit/cmk/gui/plugins/wato/test_check_parameters.py +++ b/tests/unit/cmk/gui/plugins/wato/test_check_parameters.py @@ -1,6 +1,7 @@ import pytest from cmk.gui.exceptions import MKUserError -from cmk.gui.plugins.wato.check_parameters import forbid_re_delimiters_inside_groups +from cmk.gui.plugins.wato.check_parameters import (forbid_re_delimiters_inside_groups, + transform_cpu_iowait) @pytest.mark.parametrize('pattern', ["(test)$", 'foo\\b', '^bar', '\\bfoo\\b', '(a)\\b']) @@ -12,3 +13,24 @@ def test_validate_ps_allowed_regex(pattern): def test_validate_ps_forbidden_regex(pattern): with pytest.raises(MKUserError): forbid_re_delimiters_inside_groups(pattern, '') + + +(a)pytest.mark.parametrize(&#39;paramsrams, result', [ + ( + (10, 20), + { + 'iowait': (10, 20) + }, + ), + ({}, {}), + ( + { + 'util': (50, 60) + }, + { + 'util': (50, 60) + }, + ), +]) +def test_transform_cpu_iowait(params, result): + assert transform_cpu_iowait(params) == result