Comments (10)
brian-brazil
commented on September 7, 2024
There is, however you should never want to do this outside of unittests.
You should probably be removing the label in question.
from client_java.
badprogrammist
commented on September 7, 2024
Is it true that my assumption can induce OutOfMemory error?
from client_java.
brian-brazil
commented on September 7, 2024
Having many children can run you out of memory, but you're using the library in very much the wrong way if that's an issue. Prometheus is a metrics system, not a logging system.
from client_java.
badprogrammist
commented on September 7, 2024
I haven't been using that library in wrong way yet. But I'm concerned about this problem after I having looked through entire library's code and haven't found the code that removes children or collectors at all.
from client_java.
brian-brazil
commented on September 7, 2024
In normal usage you'd never remove children, they're intended to stick around until the process dies.
If this is a problem for you, then you're using the library in the wrong way.
from client_java.
badprogrammist
commented on September 7, 2024
Could you please tell me whether I'm using library properly? I have a web application and it writes metrics on each request to this application. I had debugged this code and saw that the instances of the collector's children are accumulated.
from client_java.
brian-brazil
commented on September 7, 2024
What labels are on these metrics, and what are their cardinalities? It sounds like you have something as a label that you shouldn't have.
from client_java.
badprogrammist
commented on September 7, 2024
I described two metrics:
private static final Counter requestsCounter = Counter.build()
.name('http_requests_total')
.help('Request amount')
.labelNames('service', 'code', 'method', 'url')
.register()
private static final Histogram requestsDuration = Histogram.build()
.name('http_request_duration_milliseconds')
.help('Request process duration in milliseconds')
.labelNames('service', 'code', 'method', 'url')
.register()
Each http request it writes metrics like this:
requestsCounter
.labels(componentName, String.valueOf(response.getStatus()), request.getMethod(), uri)
.inc()
requestsDuration
.labels(componentName, String.valueOf(response.getStatus()), request.getMethod(), uri)
.observe(duration)
The metrics that are transferred:
...
http_request_duration_milliseconds_bucket{service="skybonds-data-api",code="200",method="GET",url="${api.root.path}/debug/metrics/prometheus",le="0.005",} 0.0
http_request_duration_milliseconds_bucket{service="skybonds-data-api",code="200",method="GET",url="${api.root.path}/debug/metrics/prometheus",le="0.01",} 0.0
http_request_duration_milliseconds_bucket{service="skybonds-data-api",code="200",method="GET",url="${api.root.path}/debug/metrics/prometheus",le="0.025",} 0.0
http_request_duration_milliseconds_bucket{service="skybonds-data-api",code="200",method="GET",url="${api.root.path}/debug/metrics/prometheus",le="0.05",} 0.0
http_request_duration_milliseconds_bucket{service="skybonds-data-api",code="200",method="GET",url="${api.root.path}/debug/metrics/prometheus",le="0.075",} 0.0
http_request_duration_milliseconds_bucket{service="skybonds-data-api",code="200",method="GET",url="${api.root.path}/debug/metrics/prometheus",le="0.1",} 0.0
http_request_duration_milliseconds_bucket{service="skybonds-data-api",code="200",method="GET",url="${api.root.path}/debug/metrics/prometheus",le="0.25",} 0.0
http_request_duration_milliseconds_bucket{service="skybonds-data-api",code="200",method="GET",url="${api.root.path}/debug/metrics/prometheus",le="0.5",} 0.0
http_request_duration_milliseconds_bucket{service="skybonds-data-api",code="200",method="GET",url="${api.root.path}/debug/metrics/prometheus",le="0.75",} 0.0
http_request_duration_milliseconds_bucket{service="skybonds-data-api",code="200",method="GET",url="${api.root.path}/debug/metrics/prometheus",le="1.0",} 0.0
http_request_duration_milliseconds_bucket{service="skybonds-data-api",code="200",method="GET",url="${api.root.path}/debug/metrics/prometheus",le="2.5",} 0.0
http_request_duration_milliseconds_bucket{service="skybonds-data-api",code="200",method="GET",url="${api.root.path}/debug/metrics/prometheus",le="5.0",} 0.0
http_request_duration_milliseconds_bucket{service="skybonds-data-api",code="200",method="GET",url="${api.root.path}/debug/metrics/prometheus",le="7.5",} 0.0
http_request_duration_milliseconds_bucket{service="skybonds-data-api",code="200",method="GET",url="${api.root.path}/debug/metrics/prometheus",le="10.0",} 0.0
http_request_duration_milliseconds_bucket{service="skybonds-data-api",code="200",method="GET",url="${api.root.path}/debug/metrics/prometheus",le="+Inf",} 1.0
http_request_duration_milliseconds_count{service="skybonds-data-api",code="200",method="GET",url="${api.root.path}/debug/metrics/prometheus",} 1.0
http_request_duration_milliseconds_sum{service="skybonds-data-api",code="200",method="GET",url="${api.root.path}/debug/metrics/prometheus",} 11.0
...
from client_java.
brian-brazil
commented on September 7, 2024
If you've lots of services or urls that's going to be a problem.
It's also reocmmended to use seconds everywhere, and not break out latency by status code.
from client_java.
badprogrammist
commented on September 7, 2024
Ok, thanks a lot
from client_java.
Related Issues (20)
- Endpoint actuator/prometheus not working on macOS HOT 1
- Publish a BOM again
- Implement latest changes in the Prometheus Protobuf format
- [httpserver exporter] No option to specify a wait time for ongoing requests when closing the server HOT 3
- Explicit Exception for Duplicate Labels during Scrape HOT 1
- Stress test failure in DropwizardExportsTest HOT 2
- Stress test failure in SlidingWindowTest HOT 1
- dropwizard5 MapperConfig doesn't rename the metric? HOT 4
- How to remove all metrics in the defaultRegistry HOT 5
- Failing while building simpleclient_httpserver version parent-0.15.0 using mvn clean install HOT 4
- log4j2 instrumentation support for 1.X?
- Flaky test - SlidingWindowTest
- Filter parameter causes HTTP Status 400 – Bad Request HOT 1
- info required for compatible version with openjdk21 HOT 2
- Counts in ClassicHistogramBuckets cannot be negative. HOT 1
- StatefulMetric clear() invocation causes noLabels increase to stop reporting on collect()
- Integrating Native Histograms with Spring 2.7.x and Prometheus client_java 1.3.1: Challenges in Writing Native Histograms to Response Object
- Inconsistencies between PrometheusNaming.sanitizeMetricName() methods
- UnknownDataPointSnapshot constructor ignores labels
- [FeatureRequest] Restore MetricsFilter in the servlet modules
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from client_java.