28e40a0f1b
* Initial metrics support based on Prometheus This commit introduces initial support for metrics in open5gs. The metrics code is added as libogsmetrics (lib/metrics/), with a well defined opaque API to manage different types of metrics, allowing for different implementations for different technologies to scrap the metrics (placed as lib/metrics/<impl>/. The implementation is right now selected at build time, in order to be able to opt-out the related dependencies for users not interested in the features. 2 implementations are already provided in this commit to start with: * void: Default implementation. Empty stubs, acts as a NOOP. * prometheus: open5gs processes become Prometheus servers, offering states through an http server to the Prometheus scrappers. Relies on libprom (prometheus-client-ci [1] project) to track the metrics and format them during export, and libmicrohttpd to make the export possible through HTTP. [1] https://github.com/digitalocean/prometheus-client-c The prometheus-client-c is not well maintained nowadays in upstream, and furthermore it uses a quite peculiar mixture of build systems (autolib on the main dir, cmake for libprom in a subdir). This makes it difficult to have it widely available in distros, and difficult to find it if it is installed in the system. Hence, the best is to include it as a meson subproject like we already do for freeDiameter. An open5gs fork is requried in order to have an extra patch adding a top-level CMakeList.txt in order to be able to includ eit from open5gs's meson build. Furthermore, this allows adding bugfixes to the subproject if any are found in the future. * [SMF] Initial metrics support * [SMF] Add metrics at gtp_node level * docs: Add tutorial documenting metrics with Prometheus
4.1 KiB
title
| title |
|---|
| Metrics with Prometheus |
0. Introduction
This tutorial explains how to export open5gs metrics to Prometheus, which can in turn be used to visualize or export them to other systems such as Grafana or StatsD.
When this method is used, any open5gs program exporting metrics becomes a Prometheus server, which is basically an HTTP server serving Prometheus data to the Prometheus scrapper.
Note: Only open5gs-smfd supports exporting metrics so far, though other may hopefully follow soon.
1. Enable Prometheus support during build
Open5GS programs use a generic internal API available in libogsmetrics. This
library implements the API based on configuration passed during open5gs build
time. By default, the library will be built using the void implementation,
which is basically a NO-OP implementation.
In order to use the Prometheus, the prometheus metrics implementation needs to
be selected at build time:
meson configure -Dmetrics_impl=prometheus build
This will enable building the implementation under lib/metrics/prometheus/, which uses:
- prometheus-client-c project (libprom): To generate the Prometheus expected output format of the metrics
- libmicrohttpd: To server the content generated by libprom as an HTTP server
The prometheus-client-c project is not currently well maintained, and uses a
weird mixture of build systems, which makes it difficult to make it available in
most Linux distributions. As a result, a fork of the project is available under
Open5GS GitHub namespace, with an extra patch applied making it possible to
include it as a subproject, which will be fetched and built automatically when
building the prometheus libmetrics implementation.
2. Configuring for runtime
By default the created Prometheus HTTP server will be listening on 0.0.0.0
port 9090.
This can be configured under the following config file options:
#
# metrics:
#
# <Metrics Server>
#
# o Metrics Server(http://<any address>:9090)
# metrics:
# addr: 0.0.0.0
# port: 9090
#
metrics:
addr: 0.0.0.0
port: 9090
Note: You may want to change the default IP address or port if you are running the Prometheus scrapper in the same host, since it will also spawn its own Prometheus server also in port 9090, which will collide.
3. Manual visualization
Simply open the web browser at the following URL (changing IP address and port as configured in previous section):
http://127.0.0.1:9090/metrics
Note: URL metrics/ (with a slash at the end) will not work.
You should see some output similar to this one below:
# HELP ues_active Active User Equipments
# TYPE ues_active gauge
ues_active 2
# HELP process_max_fds Maximum number of open file descriptors.
# TYPE process_max_fds gauge
process_max_fds 1024
# HELP process_virtual_memory_max_bytes Maximum amount of virtual memory available in bytes.
# TYPE process_virtual_memory_max_bytes gauge
process_virtual_memory_max_bytes -1
# HELP process_cpu_seconds_total Total user and system CPU time spent in seconds.
# TYPE process_cpu_seconds_total gauge
process_cpu_seconds_total 0
# HELP process_virtual_memory_bytes Virtual memory size in bytes.
# TYPE process_virtual_memory_bytes gauge
process_virtual_memory_bytes 3156643840
# HELP process_start_time_seconds Start time of the process since unix epoch in seconds.
# TYPE process_start_time_seconds gauge
process_start_time_seconds 402433
# HELP process_open_fds Number of open file descriptors.
# TYPE process_open_fds gauge
process_open_fds 23
3. Integration with Prometheus scrapper
Sample Prometheus scrapper configuration (~/prometheus.yml):
global:
scrape_interval: 10s
scrape_configs:
- job_name: open5gs-smfd
static_configs:
- targets: ["192.168.1.140:9091"]
Where 192.168.1.140:9091 is the IP address and port where open5gs-smfd is
serving its metrics, as configured in above sections.
The Prometheus scrapper can be easily started from a docker container:
docker run -p 9090:9090 -v /prometheus.yml:/etc/prometheus/prometheus.yml prom/prometheus
Then open your browser to be able to visualize the data: http://localhost:9090/graph