Autoscaling
o
Based
on custom policies, schedules, or metric monitoring
§
Can
use any metric CloudWatch measures
o
Scaling
out ensures performance to handle demand
§
Adding
instances
o
Scaling
in minimizes cost when resources are no longer needed
§
Removing
instances
Ø
Auto Scaling is enabled by Amazon
CloudWatch
o
No
additional fees other than CloudWatch fees
o
All
instances launched using Auto Scaling will have detailed monitoring automatically
enabled
Ø
Can configure Auto Scaling with web
console, command-line tools, or APIs
o
This
section will focus on the command-line tools
o
You
can explore the web console in the exercise bonus
Why Use Auto Scaling?
Ø
Ensure
that a minimum number of instances is always running
o
Auto Scaling will make sure that number of
instances is always running
o
If an instance terminates for any reason, a new
one will be started
o
Can also be used to keep instance count at a
fixed size
§
Delete “extra” instances started by mistake
Ø
Automatically
scale instances to meet demand
o
Can specify quantity of instances to launch or
terminate as demand requires
§
For example, add or remove five instances at a
time
o
Can specify a “cool-down” period to wait before
taking action
Two Steps to Get Started With Auto Scaling
Ø
Create a
launch configuration, specifying
o
A name for the launch configuration
o
The image (AMI) to use when creating instances
o
The instance type
# aws autoscaling
create-launch-configuration --launch-configuration-name user01LC--image-id
ami-1235564a --instance-type m1.small –-key-name Crs1205EC2Key--security-groups
sg-7a91c610
Ø
Create an
Auto Scaling group, specifying
o
A name for the group
o
The launch configuration to use
o
One or more subnets (separated by commas) to run
the instances in
o
Minimum instance count
o
Maximum instance count
# aws autoscaling create-auto-scaling-group
--auto-scaling-group-name user01ASGrp --launch-configuration-name user01LC
--vpc-zone-identifier subnet-41767929c --min-size 2 --max-size 4
Ø
After creating an Auto Scaling group
o
The
minimum number of instances will automatically be started
o
No
other action is needed for this to happen
Ø
The Auto Scaling group ensures that
the minimum number of instances are always running
o
New
instances will automatically be started if necessary
Checks, Policies, and Actions
Ø
Health check
o
A
check of the health of an instance within an Auto Scaling group
o
Uses
standard EC2 health checks by default
o
Can
also use custom Elastic Load Balancing (ELB) health checks
§
Will
be seen shortly
o
If
an instance is marked not healthy, it is immediately scheduled for replacement
§
By
default, it is not allowed a chance to recover
Ø
Policy
o
Instructions
on how to change the instance count size
o
Can
specify the size as an absolute size, increment, decrement, or percentage of
current size
Ø
Triggers
o
Combination
of an Auto Scaling policy and CloudWatch alarm
Three Types of Scaling
Ø
Manual
scaling
o
Change the Auto Scaling group instance counts
# aws autoscaling update-auto-scaling-group
--auto-scaling-group-name user01ASGrp --min-size 4 --max-size 10
Ø
Scaling
on a schedule
o
Useful when you have a predictable schedule of
required demand
o
For example: Run additional servers between
10:00 a.m. and 4:00 p.m.
Ø
Scaling
based on triggers and policies
o
Utilize CloudWatch alarm to execute a policy
o
Must define separate policies to scale up and
scale down; e.g.:
§
Scale up when bandwidth exceeds a threshold
§
Scale down when average bandwidth falls back
down
o
This is the truly “automatic” part of Auto
Scaling
Creating a Policy
Ø
Create a new policy
#aws autoscaling put-scaling-policy
–-policy-name <PolicyName> --adjustment-type <type> --auto-scaling-group-name
<as-group> --scaling-adjustment <adjustment>
o
Type
can be ExactCapacity, ChangeInCapacity, PercentChangeInCapacity
Ø
Examples:
o
Scale
up by two instances
# aws autoscaling put-scaling-policy
–-policy-name IncrBy2Pol --adjustment-type ChangeInCapacity --auto-scaling-group-name
user01ASGrp --scaling-adjustment 2
o
Scale
down by two instances
# aws autoscaling put-scaling-policy
–-policy-name DecrBy2Pol --adjustment-type ChangeInCapacity --auto-scaling-group-name
user01ASGrp --scaling-adjustment -2
o
Scale
up by 50 percent
--adjustment-type PercentChangeInCapacity
–-scaling-adjustment 50
Testing a Policy
Ø
Policies
can be tested by executing them manually
o
Good
idea to test before utilizing it from a trigger
# aws autoscaling execute-policy
--auto-scaling-group-name user01ASGrp --policy-name team02IncrBy2Pol
Ø
Note:
An Auto Scaling group always ensures that the running instance count adheres to
the minimum and maximum sizes specified in the group
o
Executing
a policy will not be allowed to violate this
o
For
example, if a maximum size is eight, there will never be more than eight
instances
§
Even
if the policy says to add 10 instances
Creating a Trigger
Cool-Down Period
The Auto Scaling cooldown period is a configurable setting for
your Auto Scaling group that helps to ensure that Auto Scaling doesn't launch
or terminate additional instances before the previous scaling activity takes
effect. After the Auto Scaling group dynamically scales using a simple scaling
policy, Auto Scaling waits for the cooldown period to complete before resuming
scaling activities. When you manually scale your Auto Scaling group, the
default is not to wait for the cooldown period, but you can override the
default and honor the cooldown period. Note that if an instance becomes
unhealthy, Auto Scaling does not wait for the cooldown period to complete
before replacing the unhealthy instance
Ø
A time period to suspended Auto
Scaling
o
Goal
is to prevent Auto Scaling from reacting too quickly
§
For
example, before a previous scaling activity has finished
Ø
Can specify a cool-down period for
an Auto Scaling group
o
Provide
the following option to the create-auto-scaling-group command when creating the
group:
--default-cooldown <seconds>
--default-cooldown <300s>
Ø
Can also specify a cool-down period
for each policy
o
Overrides
the group’s cool-down period for that policy
o
Provide
the following option to the put-scaling-policy when creating a policy:
--default-cooldown <seconds>
Managing Auto Scaling Resources
Ø
All
launch configurations defined can be displayed:
aws autoscaling
describe-launch-configurations
Ø
All
Auto Scaling groups defined can be displayed:
aws autoscaling describe-auto-scaling-groups
Ø
All
policies defined can be displayed:
aws autoscaling describe-policies
Benefits of Auto Scaling
Adding Auto Scaling to your
application architecture is one way to maximize the benefits of the AWS cloud.
When you use Auto Scaling, your applications gain the following benefits:
·
Better
fault tolerance. Auto Scaling can detect when an instance is unhealthy,
terminate it, and launch an instance to replace it. You can also configure Auto
Scaling to use multiple Availability Zones. If one Availability Zone becomes
unavailable, Auto Scaling can launch instances in another one to compensate.
·
Better
availability. Auto Scaling can help you ensure that your application always
has the right amount of capacity to handle the current traffic demands.
·
Better
cost management. Auto Scaling can dynamically increase and decrease
capacity as needed. Because you pay for the EC2 instances you use, you save
money by launching instances when they are actually needed and terminating them
when they aren't needed.
