简体繁体 English

"在 activemq 上监控和警报"

[英]Monitoring and alerting on activemq

原文 2019-04-27 05:44:36 8 2 java/ activemq/ grafana/ spring-jms

Hi with activemq problems that i usually face are as under:- 1. Pushed message do not get consumed 2. As a result of point 1, queue size becomes huge 3. Messages in queue keep on retrying嗨，我通常面临的 activemq 问题如下：- 1. 推送的消息没有被消耗 2. 由于第 1 点，队列大小变得巨大 3. 队列中的消息不断重试

So is there any open source that anyone of you have used in project to send mails when queue size goes beyond a certain threshold.那么，当队列大小超过某个阈值时，你们中的任何人在项目中使用过任何开源来发送邮件吗？ Can we also send message details or store it somewhere to diagnose later why these messages were not getting consumed.我们是否也可以发送消息详细信息或将其存储在某个地方以便稍后诊断这些消息没有被消费的原因。

2 个解决方案

Zabbix can be a choice: https:\/\/www.zabbix.com\/integrations\/activemq<\/a> If the queue size becomes huge, it will create an alert. Zabbix 可以是一个选择： https<\/a> :\/\/www.zabbix.com\/integrations\/activemq 如果队列变大，它会创建一个警报。 Also, it has more variety of other alerts that can be useful.此外，它还有更多种类的其他有用的警报。

You can achieve this using alerting tools. 您可以使用警报工具实现此目的。
You have to collect the metrics from ActiveMQ jmx (or Jolokia over http): https://activemq.apache.org/jmx.html and react upon it. 您必须从ActiveMQ jmx（或Jolokia over http）收集指标： https ： //activemq.apache.org/jmx.html并对其做出反应。 A good stack (free, but not open source, if you don't need high availability) is Telegraf to collect metrics + InfluxDB to persist + Grafana to render metrics and send alerts. 一个好的堆栈（免费，但不是开源，如果你不需要高可用性）是Telegraf收集指标+ InfluxDB持久+ Grafana呈现指标和发送警报。

For example: 例如：

on the broker you'll find the StorePercentUsage to monitor the overall disk usage 在代理上，您将找到StorePercentUsage来监控整体磁盘使用情况
on a queue (=Destination) you'll find the QueueSize and the ConsumerCount. 在队列（= Destination）上，您将找到QueueSize和ConsumerCount。

You can easily build a rule to send an alert when the metrics are above/below a threshold. 您可以轻松构建规则，以在指标高于/低于阈值时发送警报。