AWS Monitoring

We use Сloudwatch for gathering all metrics in AWS. It's the official way at AWS platform and below you can find all important metrics of Omaha-Server which we recommend to track:

1. RDS.(CloudWatch -> Metrics -> RDS -> Per-Database Metrics)

Metric	Description	Units	Critical Limit
CPUUtilization	The percentage of CPU utilization.	Percents	90%
DatabaseConnections	The number of database connections in use.	Count	Depends on instance type (please, see below)
FreeableMemory	The amount of available random access memory.	Bytes	90 percent of maximum. More information in Specifications for DB Instance Classes
FreeStorageSpace	The amount of available storage space.	Bytes	70-90 percents of maximum.

If you reached a critical limit you should upgrade your current instances' class to higher one.

Max amount of connections to database for different RDS types:

t2.micro	66
t2.small	150
m3.medium	296
t2.medium	312
m3.large	609
t2.large	648
m4.large	648
m3.xlarge	1237
r3.large	1258
m4.xlarge	1320
m2.xlarge	1412
m3.2xlarge	2492
r3.xlarge	2540

2. ElastiCache (CloudWatch -> Metrics -> ElastiCache > Cache Node Metrics)

The most important metrics which we recommend to use:

Metric	Description	Units	Critical Limit
CPUUtilization	The percentage of CPU utilization.	Percents	90%
FreeableMemory	The amount of free memory available on the host.	Bytes	80 percent of maximum. ElastiCache Instance Classes

3. Elastic Beanstalk. How to track Elastic Beanstalk metrics:

• Go to the Elastic Beanstalk service
• Choose an application
• Open the Monitoring tab
• Click on the Edit button, if you need additional metrics

The most important metrics:

Resource	Metric	Description	Units
AWSEBAutoScalingGroup	CPUUtilization	The percentage of allocated EC2 compute units that are currently in use on the instance.	Percent
AWSEBAutoScalingGroup	NetworkIn	The number of bytes received on all network interfaces	Bytes
AWSEBAutoScalingGroup	NetworkOut	The number of bytes sent out on all network interfaces	Bytes
AWSEBAutoScalingGroup	DiskReadBytes	Bytes read from all instance store volumes available to the instance	Bytes
AWSEBAutoScalingGroup	DiskWriteBytes	Bytes written to all instance store volumes available to the instance.	Bytes
AWSEBLoadBalancer	Latency	The time elapsed, after the request leaves the load balancer until the headers of the response are received.	Seconds
AWSEBLoadBalancer	HealthyHostCount	The number of targets that are considered healthy.	Hosts
AWSEBLoadBalancer	RequestCount	The number of requests received by the load balancer.	Requests

3.1 Elastic Beanstalk allows you to configure autoscaling which based on metrics. Follow the next steps:

1. Choose an Elastic Beanstalk application
2. Click on the Configuration tab
3. Go to the scaling section(Click on the gear button)
4. Configure a scaling trigger.

4. Inspecting celery queues.

Finding the number of tasks in a queue:
$ redis-cli -h HOST -p PORT -n DATABASE_NUMBER llen QUEUE_NAME

The default queue is named celery. To get all available queues, invoke:
$ redis-cli -h HOST -p PORT -n DATABASE_NUMBER keys \*

Note! The commands should be executed from an EC2 instance which has access to your Redis node.