Push alerts

Configuring alerts in the manner described above will cover predominantly every alerting use case required for your team of on-call engineers. It is a way to seamlessly integrate into your existing workflows and alerting tools such as Prometheus and Zabbix.

However, some use cases would require independent alerts to be setup and operate in autonomy. For many users, Plumbr is their only monitoring and alerting system. To satisfy these use cases, we provide the ability to configure Push Alerts within Plumbr. Plumbr allows you to configure simplistic alerts from within the Plumbr UI. This will enable you to receive alerts without requiring external systems.

Because Plumbr is based on real user monitoring, our triggers also reflect these very parameters. An example alerting threshold would be: “Notify me if more than 2% of 1000 last user interactions resulted in failures”.

Interactions (or API calls) are chosen as the basis for these alerts. Configuring an alert this way allows you to prevent alert flapping by default. The rolling window will allow you to prevent spurious notifications. Presently, you can configure one notification per application. If you require multiple alerts to be setup for a single application, please contact us at support@plumbr.io.

Steps to take to configure alerts:

  • Make sure you’ve configured alerting channels as specified.
  • In order to configure an alert, ensure that you have administrator-level privileges on your account.
  • Open the application/API you want to setup the alert for.
  • Click on the “Alerts” side menu item.

On the alert screen you can configure 4 different alerts: error rate, throughput, latency and new errors.

Error rate alerts

  • You can configure a percentage threshold
  • As you set the threshold, you can see the Alert band (—) appear correspondingly.
  • Click on the “Setup alert” button to save this configuration.
  • You will receive alerts via the selected channel whenever the rolling error rate exceeds the selected threshold.

Latency alerts

  • It is possible to configure both median and 99th percentile alerts
  • As you set the threshold, you can see the Alert band (—) appear correspondingly.
  • 99th percentile is usually lot higher than median and can push the median chart series to the bottom where it is difficult to see. In that case it is possible to toggle the 99th percentile chart and threshold lines in the chart legend
  • Click on the “Set up alert” button to save this configuration.
  • You will receive alerts via the selected channel whenever latency exceeds the selected threshold.

Throughput alerts

  • It is possible to configure time window and interaction count
  • As you set the threshold, you can see the Alert band (—) appear correspondingly.
  • You will receive alerts via the selected channel whenever throughput goes over or under the threshold

New error alerts

  • Every time Plumbr detects a never before seen error an alert is triggered
  • Error history chart is always shown for the last 30 days