Database Reference
In-Depth Information
Monitoring DynamoDB tables
To start with, let's get familiar with AWS CloudWatch, which is a network-monitoring ser-
vice offered by Amazon Web Service. While creating a table, you can set alarms for vari-
ous events such as provisioning a throughput with exceeded exceptions, maintaining cer-
tain threshold, and so on. A table's size in terms of bytes or number of items is not avail-
able as a CloudWatch metric by default; however, this can be set up using the
DescribeTable
operation, and it is a good practice to keep watch on this metric. These
metrics are very important from the database admin's point of view, but as we are using a
hosted service here, we need not hire a DBA specialist. These metrics would help us keep
track of whether there is any sudden demand of resources happening or not, and if the peak
and slow time frames can be drawn for our application so that we can keep more read and
write throughput for peak time frames and lesser throughput for slow time frames.
CloudWatch gives us a detailed report of the metrics we want to monitor.
There are various ways of monitoring DynamoDB tables using CloudWatch as follows:
• AWS Management Console
• API
• Command-line interface
