AWS RDS – Maximum auto_explain.log_min_duration Parameter Explained

You are using a scalar function in your WHERE clause, which means that function needs to be called multiple times. You should consider calling that function once (since you already know the values of @N and @Ticker) at the start of the script to return the scalar and then pass that scalar in as a variable. While not perfect, this should reduce the multiple calls that is likely slowing down your query. A low-end RDS box is sharing resources and it's quite possible that you are spinning more CPU ticks than necessary.

DISCLAIMER: I WORK FOR SOLARWINDS

We have a product (Solarwinds Database Performance Analyzer) that can monitor your RDS instance and help troubleshoot performance issues such as this.

Changing Instance Class:

From the documentation "Modifying a DB Instance and Using the Apply Immediately Parameter", changing the instance class and setting "Apply Immediately" to false will result in:

• The change being applied during the next maintenance window.
• The change causing an outage to occur.

Since the instance is Multi-AZ and you're changing the instance class, your outage will be minimized to the amount of time it takes to failover to the standby replica. That isn't very long; likely worst-case a minute or two. You won't lose data that is already in the DB. Any existing connections will be dropped during failover. You should verify that applications connected to your database can withstand these conditions, or alternatively take them down for maintenance during this time.

Identifying RDS instance storage capacity:

You can identify how much of the RDS storage is currently in use by either:

1) Simply selecting your RDS instance from the RDS web console and viewing the "Storage" bar under the "Monitoring" header.

2) Viewing the "FreeStorageSpace" RDS cloudwatch metric for your RDS instance.

Further Reading:

• RDS Documentation - Viewing DB Instance Metrics