普罗米修斯客户端和PasteDeploy胶水
prometheus-paster的Python项目详细描述
这是prometheus_client和pastedeploy之间的胶水。
快速启动
配置示例
[composite:main]
use = egg:Paste#urlmap
/metrics = prom
/ = papp
[pipeline:papp]
pipeline = perf app
[app:app]
use = egg:someapp
[filter:perf]
use = egg:prometheus_paster#filter
name = someapp_call
[app:prom]
use = egg:prometheus_paster
普罗米修斯导出器终结点配置
您可以使用egg:prometheus_pasterwsgi应用程序公开端点。
[app:main]
use = egg:prometheus_paster
收集器过滤器配置
- 延迟：wsgi处理时间的度量名称（默认值
response_latency_seconds） - length:wsgi响应长度的度量名称（默认值
response_length_bytes） - path_regex：如果请求路径与此regex不匹配，则将忽略它。（默认值
.*） - 禁用延迟：如果设置为true，则不会收集延迟度量。
- 禁用长度：如果设置为true，则不会收集长度度量。
多指标（1）
如果要在同一个wsgi容器上收集两个不同的应用程序， 然后你可以放不同的过滤器。
[composite:main]
use = egg:Paste#urlmap
/metrics = prom
/app1 = app1
/app2 = app2
[app:app1]
use = egg:app1
filter-with = app1_perf
[filter:app1_perf]
use = egg:prometheus_paster#filter
latency = app1_latency
length = app1_length
[app:app2]
use = egg:app2
filter-with = app2_perf
[filter:app2_perf]
use = egg:prometheus_paster#filter
latency = app2_latency
length = app2_length
[app:prom]
use = egg:prometheus_paster
多指标（2）
另一个选择是堆叠过滤器。
[composite:main]
use = egg:Paste#urlmap
/metrics = prom
/app = app1_pipe
[pipeline:app1_pipe]
pipeline = path1_perf path2_perf app1
[app:app1]
use = egg:app1
[filter:path1_perf]
use = egg:prometheus_paster#filter
latency = path1_latency
length = path1_length
path_regex = /path1/.*
[filter:path2_perf]
use = egg:prometheus_paster#filter
latency = path2_latency
length = path2_length
path_regex = /path2/.*
[app:prom]
use = egg:prometheus_paster
欢迎加入QQ群-->： 979659372
