Welcome to the “How to Write Better Apps” section.
In short, this section is a walkthrough of the best practices to improve your app performance and point out common user mistakes to avoid when working with ScyllaDB CQL.
We’re also going to talk about things that you should consider while working with multiple data centers.
The first thing that you should know is that we’re going to be using a ScyllaDB project called Scylla Monitoring, which provides a full dashboard to work with, displaying many metrics related to your current ScyllaDB usage in a really fancy way.
You can find graphs and monitoring tools by clicking on the monitoring button inside your cluster. But if you’re using the open-source version of ScyllaDB, make sure to install it according to the monitoring repository documentation.
After opening your monitoring stack, you’ll find dashboards with overview metrics where you can find gauges for average reads and writes requests per second, the number of nodes, P99 latency, and so on.
You’re also going to find dashboards like advanced reports, operational system metrics, detailed reports, and the one that you’re going to deep dive into in this section, which is the CQL dashboard.
At the CQL dashboard, you can find information about all of the executed queries inside your cluster. But better than that, you can also spot gauges for queries that slow down your cluster performance for some reasons like:
– non-prepared statements
– non-paged CQL reads
– non-token-aware queries
– spamming of ALLOW FILTERING
– and non-paged and reverse cql reads.
So, if you have everything set up, we can just get started.
