Amazon Redshift Maintenance

Keeping our dashboards and analytic pipelines healthy, require a performant cluster. For this reason, we need to periodically perform maintenance tasks on our Amazon Redshift Instance.

We will see why it is important to involve the data analyst in the maintenance process and how the knowledge of the data and the queries to be performed, can help not only keep the cluster healthy but also maximize its performance.

First, we will see what the Vacuum process is and how we can perform it. Why we need it and why it is important to know the nature of your data for laying out a good Vacuuming strategy.

Then investigate what Column Compression Encoding is. How it affects the utilization of our cluster and the speed of our queries. How the deep knowledge of the data that the analyst has, can help decide which encoding is optimal.

Finally, we will go through some different sources of information that we can use to monitor the performance of the executed queries. Last but not least provide some hints on how to use this information to identify what causes the queries to slow down and what maintenance technique to perform to fix them.