By Rohan Patel in software-dev — Jul 10, 2022

DVA-C01 studying - CloudWatch and CloudTrail

https://aws.amazon.com/cloudwatch/
monitoring service to monitor health and performance metrics of you applications and AWS resources
- by default can monitor compute, storage & content delivery, database & analytics, and a variety of other AWS services
- can even monitor estimated AWS charges
CloudWatch Agent - allows you to define your own metrics to monitor
OperatingSystem-Level metrics
- by default OS-level metrics are not sent to CloudWatch - need to use the CloudWatch agent to do this
  - memory usage, processess, CPU idle time, etc
- metric frequency - by default sent in 5-min intervals
  - can enable detailed monitoring for 1-min intervals for an extra charge
  - custom metrics default to 1-min intervals but high resolution metrics can be configured for 1-sec intervals (extra charge)
host-level metrics
- default EC2 metrics include CPU, network, disk, status check
CloudWatch Logs - allows you to monitor OS and application logs
- system and application logs - useful to track errors and troubleshoot
CloudWatch Alarms
- alarms you can create to monitor any CloudWatch metric
  - cpu or memory ulitization, latency, AWS bill charges, etc.
  - can set thresholds on the metric to trigger the alarm
- alarms can trigger things e.g. executing an Auto Scaling policy

a variable you want to monitor using CloudWatch
time-ordered sequence of data-points that are published to Cloudwatch
- each data-point has a timestamp and optional unit of measurment
for example, logging the CPU usage on an instance
uniquely defined by a name, namespace, and zero or more dimensions

a container for CloudWatch metrics
AWS provides its own (e.g. AWS/EC2 for EC2 metrics) and you can create your own for custom metrics data
- each data-point you want to publish to CloudWatch must have a specified namespace
- each metric you create must have a specifiec namespace
metrics in different namespaces are isolated from one another - they are not aggregated together

custom dashboard views you can create in CloudWatch to visualize/display specified metrics

CloudWatch API supports a long list of actions, https://docs.aws.amazon.com/AmazonCloudWatch/latest/APIReference/API_Operations.html
- actions allow you to publish, monitor, and alert on metrics
very useful when creating custom metrics
PutMetricData (super common action)
- publishes metric data-points to CloudWatch
- define the name of the metric, the namespace to publish it to, the value to publish, and timestamp for the data-point
PutMetricAlarm (super common action)
- creates an alarm associated with a metric to alert you if a threshold has been reached

https://aws.amazon.com/cloudtrail/
records API calls (user activity) in your AWS account - events related to creation, modification, or deletion of resources (also includes failed logins)
- stores log files in an S3 bucket
by default shows the last 90 days of activity
can be integrated with CloudWatch Logs
used for audit logs, not performance and metrics