Why do we love Grafana?

If you like big graphs and you cannot lie, you’re probably familiar with Grafana. :)

The particular data points that are of interest, completely depends on what you want to use it for - while some data helps you in figuring out when an outage happened and the impact it has on other parts of your system, and some others help you proactively find issues before it becomes a big outage.

For example, collecting metrics about the CPU, RAM, Disk usage of the server, network load would give an insight into your Cluster Health. Keeping them in the right range could mean that we have enough resources for our workload if these numbers are under threshold. If it exceeds that range, we would have to reduce the usage by killing unnecessary processes or by deleting old log files or by increasing the capacity of the server to handle the load etc.

Another widely used metric correlation - monitoring your status code can give you a direct indication of whether your customer is facing a product usability problem. 

For example, the number of 5xx http status codes in a particular period of time may give us an insight into the severity of the customer impact.

But as is the case with any analysis, clear conclusions and root causes can be drawn out easily if you have more relevant data. However, there’s also such a thing as too much data. That being said, wouldn’t it be useful if your data could transform itself into something that makes sense as soon as you look at it? 

Enter Grafana

Grafana allows you to query and visualize data from anywhere. Grafana helps turn your monitoring metrics into beautiful graphs and helps you visualize your collected data better.

While Grafana allows for a few channels of alerting, the platform is not meant to replace your IT Alerting Tool. Its primary focus is to serve as a data visualization tool. Like we’ve mentioned in a few of our other posts, it’s very important for you to make sure your data is actionable. You can do this by integrating Grafana, like you would with any monitoring tool, to your incident management platform and configure it to notify you on the customer impacting alerts. 

In this particular post, we’ll be going through how you can use your Grafana data to set off alert triggers in Squadcast.

Side Note: At Squadcast, we use Grafana and absolutely love it! ❤️

What can you get out of the Grafana + Squadcast Integration?

1. Create Alerts from Grafana Data: You can send in relevant alerts from Grafana as incidents into Squadcast using a Webhook. 
   

2. Auto Resolve incidents: Alerts that are resolved in Grafana are automatically resolved in Squadcast. That way, you don’t have to do it twice! 
   

3. Send in Images: You can send in Grafana Images to Squadcast. All you’ll have to do is check the `Include Images` box and you’ll see the images as a part of the incident description on both our web app and mobile app.
   

Best Practices - Configuring Grafana & Squadcast

We understand that you will want all of your data in one place, however, it’s important for your incident management tool to know if an engineer needs to be pinged in the middle of his meal for this alert.  

A few things you can do to ensure that this integration works well for you: 

1. Send in actionable alerts only: This makes it easy for your on-call team to understand that any incident that comes into the incident management platform is of utmost importance. It also reduces the mental load and effort taken to check each alert that pours in. You can always check your 3rd party tool for any other data you’ll need. 
   However, if your processes are such that you will still need all of the data in one place, you can check out the following practices. 
   

2. Configure suppression rules: Configuring Suppression rules will not only ensure that all your data is available in your alerting platform but will also let Squadcast know that no notifications need to go for a few types of alerts. 
   You can set your rules for the kind of alerts that can be suppressed on Squadcast. 
   

3. Configure deduplication rules: There are cases where the same alert is triggered several times. It’s easy to get frustrated if you’re getting notified for the same alert 10 times in a span of 2 minutes. So, you can set your deduplication rules in Squadcast, which will group these same alerts as events under just one alert, and Squadcast will notify the relevant user or team.  
   

Turbocharge your observability data in Grafana by making it actionable! If you have other best practices to share or just need help with the integration set-up, feel free to drop a line to our Support Team.