Cloud watchingwithcloudwatch

CloudWatch

Cloud Watchingwith

Inside The Cloud – 2nd May 2013

John O’Shea, AWS Monitoring

Cloud Watching with CloudWatch

Cloud Watching with CloudWatch

Quick AWS Case Study

Cloud Watching with CloudWatch

Quick AWS Case Study

AWS Monitoring

Cloud Watching with CloudWatch

Quick AWS Case Study

AWS Monitoring

CloudWatch

Cloud Watching with CloudWatch

Quick AWS Case Study

AWS Monitoring

CloudWatch

Inside

Obama For America on AWS

quick casestudy

http://awsofa.info/

How & what is my

cloud doing?

MONITORING

WILL TELL YOU

AWS Monitoring

We provide the platform for monitoring

1. AWS (& Amazon)2. Your AWS Resources

Time Series Data (TSD)

A time series is a sequence of data points, measured typically at successive points in time spaced

at uniform time intervals.http://en.wikipedia.org/wiki/Time_series

Metrics

Forecasting, Estimation, Classification, Anomaly Detection

API calls, ad views, game play data, application

usage, checkouts, boxes shipped, robot activity, customer happiness!

Time Series Data (TSD)

Data that describes an application's performance and

behavior.

Metrics

Amazon CloudWatch provides monitoring for AWS cloud resources and the applications

customers run on AWS...

What is AWS CloudWatch?

Metrics Alarms

Monitoring

Collect data

Visualize

Act!

AWS Time Series Data

CloudWatch

Amazon EC2

Elastic Load

Balancing

Amazon EBS

Amazon EMR

Auto Scalin

g

Amazon RDS

DynamoDB

ElastiCache

Amazon SNS Amazon

SQS

AWS Storage Gateway

Monitoring

Collect data

Visualize

Act!

AWS CloudWatch Metrics

https://console.aws.amazon.com/cloudwatch/

Monitoring

Collect data

Act!

AWS CloudWatch Alarms

AWS CloudWatch Alarms

• Example Use Cases• EC2: Tell me if my instance needs attention• DynamoDB: Help me balance cost and

performance• Billing: Tell me when I'm exceeding my AWS

budget

• Fallacies of Distributed Computing• The network is reliable• Latency is zero• Bandwidth is infinite• The network is secure• Topology doesn't change• There is one administrator• Transport cost is zero• The network is homogeneous

Inside

Inside

• 1 datapoint per metric, per minute, per resource• AWS is VERY big, VERY dynamic, growing VERY

fast!• Aggregating data to produce meaningful

statistics• What is CPU utilization across my fleet?• Today it is 50% - Yay!• Yesterday it was 25% – Huh?!• Last week it was 5% – Yikes?!

Inside

• Data Access Patterns• Realtime data can be VERY valuable to wide

audiences – ‘Million Dollar Minutes’• LOTS of older data, that might still be VERY

valuable to narrow audiences• Be the Last Man Standing• What happens when there is a power event in a

DC?• Who watches the watchmen?

Time Series Data (TSD).

http://commons.wikimedia.org/wiki/File:Reunion_sausages_dsc07796.jpg

CloudWatch free tier• ‘Five-minute’ metrics on AWS

resources• 10 custom metrics, 10 alarms and

1 million API requests each monthGo PLAY!

aws.amazon.com/cloudwatch/

We’re hiring...

Work HardSolve problems of massive telemetric data scale, distributed systems/cloud computing, data visualization, and workflow.

We’re hiring...

Have FunAn environment you’ll love working in, teammates you’ll love working with

Work HardSolve problems of massive telemetric data scale, distributed systems/cloud computing, data visualization, and workflow.

We’re hiring...

Make HistoryCreate solutions to some of the most challenging technical problems in the industry.

Have FunAn environment you’ll love working in, teammates you’ll love working with

Work HardSolve problems of massive telemetric data scale, distributed systems/cloud computing, data visualization, and workflow.

Thanks!(Q & A?)

[email protected]

@joshea