blob: 388e9b460c40c457094d6b78872e145124bef5a1 (
plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
|
plugin_name: python.d.plugin
modules:
- meta:
plugin_name: python.d.plugin
module_name: zscores
monitored_instance:
name: python.d zscores
link: https://en.wikipedia.org/wiki/Standard_score
categories:
- data-collection.other
icon_filename: ""
related_resources:
integrations:
list: []
info_provided_to_referring_integrations:
description: ""
keywords:
- zscore
- z-score
- standard score
- standard deviation
- anomaly detection
- statistical anomaly detection
most_popular: false
overview:
data_collection:
metrics_description: |
By using smoothed, rolling [Z-Scores](https://en.wikipedia.org/wiki/Standard_score) for selected metrics or charts you can narrow down your focus and shorten root cause analysis.
method_description: |
This collector uses the [Netdata rest api](https://github.com/netdata/netdata/blob/master/web/api/README.md) to get the `mean` and `stddev`
for each dimension on specified charts over a time range (defined by `train_secs` and `offset_secs`).
For each dimension it will calculate a Z-Score as `z = (x - mean) / stddev` (clipped at `z_clip`). Scores are then smoothed over
time (`z_smooth_n`) and, if `mode: 'per_chart'`, aggregated across dimensions to a smoothed, rolling chart level Z-Score at each time step.
supported_platforms:
include: []
exclude: []
multi_instance: true
additional_permissions:
description: ""
default_behavior:
auto_detection:
description: ""
limits:
description: ""
performance_impact:
description: ""
setup:
prerequisites:
list:
- title: Python Requirements
description: |
This collector will only work with Python 3 and requires the below packages be installed.
```bash
# become netdata user
sudo su -s /bin/bash netdata
# install required packages
pip3 install numpy pandas requests netdata-pandas==0.0.38
```
configuration:
file:
name: python.d/zscores.conf
description: ""
options:
description: |
There are 2 sections:
* Global variables
* One or more JOBS that can define multiple different instances to monitor.
The following options can be defined globally: priority, penalty, autodetection_retry, update_every, but can also be defined per JOB to override the global values.
Additionally, the following collapsed table contains all the options that can be configured inside a JOB definition.
Every configuration JOB starts with a `job_name` value which will appear in the dashboard, unless a `name` parameter is specified.
folding:
title: "Config options"
enabled: true
list:
- name: charts_regex
description: what charts to pull data for - A regex like `system\..*|` or `system\..*|apps.cpu|apps.mem` etc.
default_value: "system\\..*"
required: true
- name: train_secs
description: length of time (in seconds) to base calculations off for mean and stddev.
default_value: 14400
required: true
- name: offset_secs
description: offset (in seconds) preceding latest data to ignore when calculating mean and stddev.
default_value: 300
required: true
- name: train_every_n
description: recalculate the mean and stddev every n steps of the collector.
default_value: 900
required: true
- name: z_smooth_n
description: smooth the z score (to reduce sensitivity to spikes) by averaging it over last n values.
default_value: 15
required: true
- name: z_clip
description: cap absolute value of zscore (before smoothing) for better stability.
default_value: 10
required: true
- name: z_abs
description: "set z_abs: 'true' to make all zscores be absolute values only."
default_value: "true"
required: true
- name: burn_in
description: burn in period in which to initially calculate mean and stddev on every step.
default_value: 2
required: true
- name: mode
description: mode can be to get a zscore 'per_dim' or 'per_chart'.
default_value: per_chart
required: true
- name: per_chart_agg
description: per_chart_agg is how you aggregate from dimension to chart when mode='per_chart'.
default_value: mean
required: true
- name: update_every
description: Sets the default data collection frequency.
default_value: 5
required: false
- name: priority
description: Controls the order of charts at the netdata dashboard.
default_value: 60000
required: false
- name: autodetection_retry
description: Sets the job re-check interval in seconds.
default_value: 0
required: false
- name: penalty
description: Indicates whether to apply penalty to update_every in case of failures.
default_value: yes
required: false
examples:
folding:
enabled: true
title: "Config"
list:
- name: Default
description: Default configuration.
folding:
enabled: false
config: |
local:
name: 'local'
host: '127.0.0.1:19999'
charts_regex: 'system\..*'
charts_to_exclude: 'system.uptime'
train_secs: 14400
offset_secs: 300
train_every_n: 900
z_smooth_n: 15
z_clip: 10
z_abs: 'true'
burn_in: 2
mode: 'per_chart'
per_chart_agg: 'mean'
troubleshooting:
problems:
list: []
alerts: []
metrics:
folding:
title: Metrics
enabled: false
description: ""
availability: []
scopes:
- name: global
description: "These metrics refer to the entire monitored application."
labels: []
metrics:
- name: zscores.z
description: Z Score
unit: "z"
chart_type: line
dimensions:
- name: a dimension per chart or dimension
- name: zscores.3stddev
description: Z Score >3
unit: "count"
chart_type: stacked
dimensions:
- name: a dimension per chart or dimension
|
