Use Go open source performance monitoring software to automatically generate charts out of the box.
Introduction to uptrace
uptrace is an all-in-one tool, an open source monitoring system that optimizes performance and monitors errors and logs. Uptrace is a cost-effective tracing solution that helps you monitor, understand, and optimize complex distributed systems. Have confidence in your product and deliver better code faster than ever. According to the official website, it has the following functions.
Features Introduction
Application Overview
Monitor your entire application stack on a compact and information-rich dashboard. You can get a quick overview of all services, hosts, and systems. It is very simple to monitor the application system status.
Data Explorer
Analyze application performance using a fast and intuitive query language. You can use the full power of SQL to gain insights from your data. It also intelligently groups similar spans together so you can instantly see the most important metrics at the route, query or RPC level.
Time Series Indicators The Time
Series Indicators feature fills the gap left by tracking using modern indicators. Uptrace comes with pre-built dashboards for popular detection and the ability to query multiple metrics from a single chart using SQL-like expressions.
Performance Optimization
Uptrace provides insight into the performance of a single request that may involve multiple distributed services. You get full context on what's different, what's broken, related logs, and errors.
Error and Log Monitoring
Using span events, you can log errors and log actions that occur automatically, capturing all contextual information, significantly reducing the time spent identifying and fixing errors.
Performance Comparison
Compare trends before and after deployment or to a previous period to understand how the application changes over time and what actions caused the changes. Compare. analyze. improve.
Smart Alerts
Recover faster from incidents by receiving notifications when apps shut down or performance anomalies are detected. Uptrace analyzes historical data to predict future performance, then alerts you when incoming data doesn't match predictions. Alert thresholds can also be manually configured in case there are higher requirements on application behavior.
The multi-language client
Uptrace supports mainstream development languages as clients, such as Go, PHP, Java, Python, node.js, etc.
How to
use Uptrace is also very simple. The official provides two ways to use it. One is a self-built service, and the second is a service using Uptrace. It is also very simple to use.
Build your own service
To start using Uptrace, you need:
Create the ClickHouse database.
Install the uptrace binary.
Start using the OpenTelemetry distribution to send data for Uptrace.
uptrace needs to use the clickhouse database to store data. You can install the database like this.
clickhouse-client -q "CREATE DATABASE uptrace"
Once started, Uptrace will connect to the specified inconfig's ClickHouse database and automatically create the required tables and views.
Uptrace provides DEB and RPM packages for Linux amd64/arm64 systems. After installing the appropriate packages, you can obtain the following information:
# 程序可执行目录
/usr/bin/uptrace
# 配置文件
/etc/uptrace/uptrace.yml
# 系统服务
/lib/systemd/system/uptrace.service
# ystemd 服务使用的环境文件
/etc/uptrace/uptrace.conf
The service can be managed through the following commands.
# 服务状态
sudo systemctl status uptrace
# 重启服务
sudo systemctl restart uptrace
# 追踪服务
sudo journalctl -u uptrace -f
Use the DSN service
to register an account through this link address. You can use a Google account or a Github account to register. After registering, you will get a link like this. UPTRACE_DSN=https://[email protected]/973.
Next, select the appropriate client language and start writing data to the DSN.
