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
class PrometheusService < MonitoringService
include ReactiveCaching
self.reactive_cache_key = ->(service) { [service.class.model_name.singular, service.project_id] }
self.reactive_cache_lease_timeout = 30.seconds
self.reactive_cache_refresh_interval = 30.seconds
self.reactive_cache_lifetime = 1.minute
# Access to prometheus is directly through the API
prop_accessor :api_url
with_options presence: true, if: :activated? do
validates :api_url, url: true
end
after_save :clear_reactive_cache!
def initialize_properties
if properties.nil?
self.properties = {}
end
end
def title
'Prometheus'
end
def description
'Prometheus monitoring'
end
def help
<<-MD.strip_heredoc
Retrieves the Kubernetes node metrics `container_cpu_usage_seconds_total`
and `container_memory_usage_bytes` from the configured Prometheus server.
If you are not using [Auto-Deploy](https://docs.gitlab.com/ee/ci/autodeploy/index.html)
or have set up your own Prometheus server, an `environment` label is required on each metric to
[identify the Environment](https://docs.gitlab.com/ce/user/project/integrations/prometheus.html#metrics-and-labels).
MD
end
def self.to_param
'prometheus'
end
def fields
[
{
type: 'text',
name: 'api_url',
title: 'API URL',
placeholder: 'Prometheus API Base URL, like http://prometheus.example.com/'
}
]
end
# Check we can connect to the Prometheus API
def test(*args)
client.ping
{ success: true, result: 'Checked API endpoint' }
rescue Gitlab::PrometheusError => err
{ success: false, result: err }
end
def metrics(environment)
with_reactive_cache(environment.slug) do |data|
data
end
end
# Cache metrics for specific environment
def calculate_reactive_cache(environment_slug)
return unless active? && project && !project.pending_delete?
memory_query = %{(sum(container_memory_usage_bytes{container_name!="POD",environment="#{environment_slug}"}) / count(container_memory_usage_bytes{container_name!="POD",environment="#{environment_slug}"})) /1024/1024}
cpu_query = %{sum(rate(container_cpu_usage_seconds_total{container_name!="POD",environment="#{environment_slug}"}[2m])) / count(container_cpu_usage_seconds_total{container_name!="POD",environment="#{environment_slug}"}) * 100}
{
success: true,
metrics: {
# Average Memory used in MB
memory_values: client.query_range(memory_query, start: 8.hours.ago),
memory_current: client.query(memory_query),
# Average CPU Utilization
cpu_values: client.query_range(cpu_query, start: 8.hours.ago),
cpu_current: client.query(cpu_query)
},
last_update: Time.now.utc
}
rescue Gitlab::PrometheusError => err
{ success: false, result: err.message }
end
def client
@prometheus ||= Gitlab::Prometheus.new(api_url: api_url)
end
end