0% found this document useful (0 votes)
81 views15 pages

Amazon Web Services - How AWS Pricing Works June 2015

This document provides an overview of how pricing works on Amazon Web Services (AWS). It discusses that AWS uses a utility-based pricing model where customers pay for only the specific cloud computing services and resources they use, with no long-term commitments. Pricing varies by service, but common characteristics include paying hourly for compute instances and per gigabyte for storage and data transfer. The document provides pricing examples for several AWS services and suggests ways customers can optimize costs, such as using reserved capacity.

Uploaded by

igrowrajesh
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
81 views15 pages

Amazon Web Services - How AWS Pricing Works June 2015

This document provides an overview of how pricing works on Amazon Web Services (AWS). It discusses that AWS uses a utility-based pricing model where customers pay for only the specific cloud computing services and resources they use, with no long-term commitments. Pricing varies by service, but common characteristics include paying hourly for compute instances and per gigabyte for storage and data transfer. The document provides pricing examples for several AWS services and suggests ways customers can optimize costs, such as using reserved capacity.

Uploaded by

igrowrajesh
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 15

Amazon

 Web  Services  –  How  AWS  Pricing  Works   June  2015  

 
 
 

   
 
 
 
 
How  AWS  Pricing  Works  
2016  
 
 
 
(Please  consult  http://aws.amazon.com/whitepapers/  for  the  latest  version  of  this  paper)  
   

Page  1  of  15    


Amazon  Web  Services  –  How  AWS  Pricing  Works   June  2015  

Table  of  Contents  


Table  of  Contents  ...................................................................................................................................................................  2  

Abstract  ..................................................................................................................................................................................  3  

Introduction  ...........................................................................................................................................................................  3  

Fundamental  Pricing  Characteristics  ......................................................................................................................................  4  

Amazon  Elastic  Compute  Cloud  (Amazon  EC2)  ..................................................................................................................  5  

Amazon  Simple  Storage  Service  (Amazon  S3)  ....................................................................................................................  6  

Amazon  Elastic  Block  Store  (Amazon  EBS)  .........................................................................................................................  6  

Amazon  Relational  Database  Service  (Amazon  RDS)  .........................................................................................................  7  

Amazon  CloudFront  ...........................................................................................................................................................  7  

How  to  Further  Save  Costs  .....................................................................................................................................................  8  

Conclusion  ..............................................................................................................................................................................  9  

Resources  .............................................................................................................................................................................  11  

Cost  Calculation  Example  .....................................................................................................................................................  12  

   

Page  2  of  15    


Amazon  Web  Services  –  How  AWS  Pricing  Works   June  2015  

Abstract  
While  the  number  and  types  of  services  offered  by  AWS  has  increased  dramatically,  our  philosophy  on  pricing  has  not  
changed:  at  the  end  of  each  month,  you  pay  only  for  what  you  use,  and  you  can  start  or  stop  using  a  product  at  any  
time.  No  long-­‐term  contracts  are  required.  

Pricing  information  for  each  service  on  our  website  is  available  at  http://aws.amazon.com/pricing/.  Our  strategy  of  
pricing  each  service  independently  gives  you  tremendous  flexibility  to  choose  the  services  you  need  for  each  project  and  
to  pay  only  for  what  you  use.    

This  whitepaper  will  help  you  understand  how  to  effectively  estimate  the  costs  of  running  your  specific  project  on  AWS.  
We  provide  several  examples  that  leverage  the  AWS  Simple  Monthly  Calculator.  For  each  example,  this  paper  discusses  
its  architecture,  example  usage  of  each  service,  cost  breakdown  for  each  service,  and  the  total  estimated  monthly  
charge.        

Introduction  
AWS  offers  a  range  of  cloud  computing  services.  For  each  service,  you  
pay  for  exactly  the  amount  of  resources  you  actually  need.  This  utility-­‐
style  pricing  model  is  explained  below:       AWS  Free  Tier  
• Pay  as  you  go.  No  minimum  commitments  or  long-­‐term  
To  help  new  AWS  customers  
contracts  required.  You  replace  your  upfront  capital  expense  
with  low  variable  cost  and  pay  only  for  what  you  use.  There  is  no  
get  started  in  the  cloud,  AWS  
need  to  pay  upfront  for  excess  capacity  or  get  penalized  for   offers  a  free  usage  tier.  If  
under-­‐planning.  For  compute  resources,  you  pay  on  an  hourly   you’re  a  new  AWS  customer,  
basis  from  the  time  you  launch  a  resource  until  the  time  you   you  can  run  a  free  Amazon  
terminate  it.  For  data  storage  and  transfer,  you  pay  on  a  per   EC2  Micro  Instance  for  a  year  
gigabyte  basis.  We  charge  based  on  the  underlying  infrastructure   while  also  leveraging  a  free  
and  services  that  you  consume.  You  can  turn  off  your  cloud   usage  tier  for  Amazon  S3,  
resources  and  stop  paying  for  them  when  you  don’t  need  them.   Amazon  Elastic  Block  Store,  
Amazon  Elastic  Load  
• Pay  less  when  you  reserve.  For  certain  products,  you  can  invest  
Balancing,  AWS  data  transfer  
in  reserved  capacity  and  get  a  significantly  discounted  hourly  
and  other  AWS  services.  For  
rate,  which  results  in  overall  savings  up  to  60%  (depending  on  
the  type  of  instance  you  reserve)  over  equivalent  On-­‐Demand   more  information,  go  to  
capacity.   http://aws.amazon.com/free  

• Pay  even  less  per  unit  by  using  more.  You  save  more,  as  you  
grow  bigger.  For  storage  and  data  transfer,  pricing  is  tiered.  The  
more  you  use,  the  less  you  pay  per  gigabyte.  For  compute,  you  get  volume  discounts  up  to  10%  when  you  
reserve  more.  

• Pay  even  less  as  AWS  grows.  Most  importantly,  we  are  constantly  focused  on  reducing  our  data  center  
hardware  costs,  improving  our  operational  efficiencies,  lowering  our  power  consumption,  and  generally  
lowering  the  cost  of  doing  business.  These  optimizations  and  AWS’s  substantial  and  growing  economies  of  scale  
result  in  passing  savings  back  to  you  in  the  form  of  lower  pricing.  Since  2006,  AWS  has  lowered  pricing  44  times.  

Page  3  of  15    


Amazon  Web  Services  –  How  AWS  Pricing  Works   June  2015  

• Custom  pricing.  What  if  none  of  our  pricing  models  work  for  your  project?  Custom  pricing  is  available  for  high  
volume  projects  with  unique  requirements.  For  assistance,  contact  us  to  speak  with  a  sales  representative.  

AWS  also  offers  a  variety  of  services  for  no  additional  charge:  

• Amazon  VPC:  Amazon  Virtual  Private  Cloud  (Amazon  VPC)  lets  you  provision  a  logically  isolated  section  of  the  
Amazon  Web  Services  (AWS)  Cloud  where  you  can  launch  AWS  resources  in  a  virtual  network  that  you  define.  
• AWS  Elastic  Beanstalk:  AWS  Elastic  Beanstalk  is  an  even  easier  way  for  you  to  quickly  deploy  and  manage  
applications  in  the  AWS  cloud.  
• AWS  CloudFormation:  AWS  CloudFormation  gives  developers  and  systems  administrators  an  easy  way  to  create  
a  collection  of  related  AWS  resources  and  provision  them  in  an  orderly  and  predictable  fashion.  
• AWS  Identity  and  Access  Management  (IAM):  AWS  IAM  controls  your  users’  access  to  AWS  services  and  
resources.  
• Auto  Scaling:  Auto  Scaling  automatically  adds  or  removes  Amazon  EC2  instances  according  to  conditions  you  
define.  With  Auto  Scaling,  the  number  of  Amazon  EC2  instances  you’re  using  increases  seamlessly  during  
demand  spikes  to  maintain  performance,  and  decreases  automatically  during  demand  lulls  to  minimize  costs.    
• AWS  OpsWorks:  AWS  OpsWorks  is  an  application  management  service  that  makes  it  easy  to  deploy  and  operate  
applications  of  all  shapes  and  sizes.  
• Moreover,  you  can  consolidate  all  your  accounts  using  Consolidated  Billing  and  get  tiering  benefits.  

Fundamental  Pricing  Characteristics  


There  are  three  fundamental  characteristics  you  pay  for  with  AWS:  compute,  storage,  and  data  transfer  out.  These  
characteristics  vary  slightly  depending  on  the  AWS  product  you  are  using.  However,  fundamentally  these  are  the  core  
characteristics  that  have  the  greatest  impact  on  cost.      

Although  you  are  charged  for  data  transfer  out,  there  is  no  charge  for  
inbound  data  transfer  or  for  data  transfer  between  other  Amazon  Web  
Services  within  the  same  region.  The  outbound  data  transfer  is  aggregated   Free  Inbound  Data  
across  Amazon  EC2,  Amazon  S3,  Amazon  RDS,  Amazon  SimpleDB,  Amazon  
SQS,  Amazon  SNS,  and  Amazon  VPC  and  then  charged  at  the  outbound  
Transfer  
data  transfer  rate.    This  charge  appears  on  the  monthly  statement  as  AWS  
There  is  no  charge  for  
Data  Transfer  Out.  
inbound  data  transfer  
The  rest  of  this  section  breaks  down  the  pricing  characteristics  for  four   across  all  Amazon  Web  
commonly  used  AWS  products:  Amazon  Elastic  Compute  Cloud  (Amazon   Services  in  all  regions.  There  
EC2),  Amazon  Simple  Storage  Service  (S3),  Amazon  Elastic  Block  Store   are  no  outbound  data  
(EBS),  Amazon  Relational  Database  Service  (Amazon  RDS),  and  Amazon   transfer  charges  between  
CloudFront.       Amazon  Web  Services  
within  the  same  region.  
Pricing  is  available  for  each  AWS  product  and  its  specific  pricing  
characteristics  at  http://aws.amazon.com/pricing/.    

Page  4  of  15    


Amazon  Web  Services  –  How  AWS  Pricing  Works   June  2015  

Amazon  Elastic  Compute  Cloud  (Amazon  EC2)    


Amazon  Elastic  Compute  Cloud  (Amazon  EC2)  is  a  web  service  that  provides  resizable  compute  capacity  in  the  cloud.  
Amazon  EC2’s  simple  web  service  interface  allows  you  to  obtain  and  configure  capacity  with  minimal  friction.  It  provides  
complete  control  of  your  computing  resources  on  Amazon’s  proven  computing  environment.  Amazon  EC2  changes  the  
economics  of  computing  by  charging  you  only  for  capacity  that  you  actually  use.  

When  you  begin  to  estimate  the  cost  of  using  Amazon  EC2,  you  need  to  consider  the  following:  

• Clock  Hours  of  Server  Time  -­‐  Resources  incur  charges  when  they  are  running.  For  example,  from  the  time  
Amazon  EC2  instances  are  launched  until  they  are  terminated,  or  from  the  time  Elastic  IPs  are  allocated  until  the  
time  they  are  de-­‐allocated.  
• Machine  Configuration  -­‐  Consider  the  physical  capacity  of  the  Amazon  EC2  instance  you  choose.  Instance  pricing  
varies  with  the  AWS  region,  OS,  number  of  cores,  and  memory.    
• Machine  Purchase  Type  -­‐  With  On-­‐Demand  Instances,  you  pay  for  compute  capacity  by  the  hour  with  no  
required  minimum  commitments.  Reserved  Instances  give  you  the  option  to  make  a  low  one-­‐time  payment  –  or  
no  payment  at  all  -­‐  for  each  instance  you  want  to  reserve  and  in  turn  receive  a  significant  discount  on  the  hourly  
usage  charge  for  that  instance.  With  Spot  Instances,  you  can  bid  for  unused  Amazon  EC2  capacity.  For  more  
information  how  to  further  save  costs  using  Reserved  Instances  and  Spot  Instances,  see  How  to  Further  Save  
Costs.  
• Number  of  Instances  -­‐  You  can  provision  multiple  instances  of  your  Amazon  EC2  and  Amazon  EBS  resources  to  
handle  peak  loads.    
• Load  Balancing  -­‐  An  Elastic  Load  Balancer  can  be  used  to  distribute  traffic  among  Amazon  EC2  instances.  The  
number  of  hours  the  Elastic  Load  Balancer  runs  and  the  amount  of  data  it  processes  contribute  to  the  monthly  
cost.  
• Detailed  Monitoring  -­‐  You  can  use  Amazon  CloudWatch  to  monitor  your  EC2  instances.  By  default,  basic  
monitoring  is  enabled  (and  available  at  no  additional  cost);  however,  for  a  fixed  monthly  rate,  you  can  opt  for  
detailed  monitoring,  which  includes  seven  preselected  metrics  recorded  once  a  minute.    Partial  months  are  
charged  on  an  hourly  pro  rata  basis,  at  a  per  instance-­‐hour  rate.  
• Auto  Scaling  -­‐  Auto  Scaling  automatically  adjusts  the  number  of  Amazon  EC2  instances  in  your  deployment  
according  to  conditions  you  define.  This  service  is  available  at  no  additional  charge  beyond  Amazon  CloudWatch  
fees.  
• Elastic  IP  Addresses  -­‐  You  can  have  one  Elastic  IP  (EIP)  address  associated  with  a  running  instance  at  no  charge.  
• Operating  Systems  and  Software  Packages  -­‐  Operating  System  prices  are  included  in  the  instance  prices.  To  
view  a  list  of  currently  available  operating  systems  that  you  can  use  with  your  Amazon  EC2  instances  and  their  
prices,  go  to  Amazon  Elastic  Compute  Cloud  (Amazon  EC2).    There  are  no  additional  licensing  costs  to  run  the  
following  commercial  operating  systems:  Red  Hat  Enterprise  Linux,  SUSE  Enterprise  Linux,  Windows  Server,  and  
Oracle  Enterprise  Linux.    Additionally,  AWS  has  made  it  easy  for  you  and  has  partnered  with  Microsoft,  IBM  and  
several  other  vendors  so  you  can  run  commercial  software  packages  on  your  Amazon  EC2  instances.  For  
example,  Microsoft  SQL  Server  on  Windows,  IBM  Software.    For  commercial  software  packages  that  AWS  does  
not  provide,  such  as  nonstandard  operating  systems,  Oracle  Applications,  Windows  Server  applications  such  as  
Microsoft  SharePoint  and  Microsoft  Exchange,  you  need  to  obtain  a  license  from  the  vendors.  You  can  also  bring  
your  existing  license  to  the  cloud  through  specific  vendor  programs  such  as  Microsoft  License  Mobility  through  
Software  Assurance  Program.    

Page  5  of  15    


Amazon  Web  Services  –  How  AWS  Pricing  Works   June  2015  

 
For  more  information  about  Amazon  EC2  pricing,  go  to  http://aws.amazon.com/ec2/.  
 

Amazon  Simple  Storage  Service  (Amazon  S3)    


Amazon  Simple  Storage  Service  (Amazon  S3)is storage for the Internet. It provides  a  simple  web  services  
interface  that  can  be  used  to  store  and  retrieve  any  amount  of  data,  at  any  time,  from  anywhere  on  the  web.  

 
When  you  begin  to  estimate  the  cost  of  Amazon  S3,  you  need  to  consider  the  following:  

 
• Storage  Class  -­‐  Standard  Storage  is  designed  to  provide  99.999999999%  durability  and  99.99%  availability.  
Standard  –  Infrequent  Access  (SIA)  is  a  storage  option  within  Amazon  S3  that  you  can  use  to  reduce  your  costs  
by  storing  less  frequently  accessed  data  at  slightly  lower  levels  of  redundancy  than  Amazon  S3’s  standard  
storage.  Standard  –  Infrequent  Access  is  designed  to  provide  the  same  99.999999999%  durability  as  S3  with  
99.9%  availability  in  a  given  year.  Each  class  has  different  rates.  
• Storage  -­‐  The  number  and  size  of  objects  stored  in  your  Amazon  S3  buckets  as  well  as  type  of  storage.  
• Requests  -­‐  The  number  and  type  of  requests.    GET  requests  incur  charges  at  different  rates  than  other  requests,  
such  as  PUT  and  COPY  requests.    
• Data  Transfer  -­‐  The  amount  of  data  transferred  out  of  the  Amazon  S3  region.  
 
For  more  information  about  Amazon  S3  pricing,  go  to  http://aws.amazon.com/pricing/s3/.  

Amazon  Elastic  Block  Store  (Amazon  EBS)  


Amazon  Elastic  Block  Store  (EBS)  provides  block  level  storage  volumes  for  use  with  Amazon  EC2  instances.  Amazon  EBS  
volumes  are  off-­‐instance  storage  that  persists  independently  from  the  life  of  an  instance.  They  are  analogous  to  virtual  
disks  in  the  cloud.  Amazon  EBS  provides  three  volume  types:  General  Purpose  (SSD),  Provisioned  IOPS  (SSD),  and  
Magnetic.  The  three  volume  types  differ  in  performance  characteristics  and  cost,  so  you  can  choose  the  right  storage  
performance  and  price  for  the  needs  of  your  applications.    

• Volumes  –  Volume  storage  for  all  EBS  volume  types  is  charged  by  the  amount  you  provision  in  GB  per  month,  
until  you  release  the  storage.  
• Input  Output  Operations  per  Second  (IOPS)  -­‐  I/O  is  included  in  the  price  of  General  Purpose  (SSD)  volumes,  
while  for  EBS  Magnetic  volumes,  I/O  is  charged  by  the  number  of  requests  you  make  to  your  volume.  With  
Provisioned  IOPS  (SSD)  volumes,  you  are  also  charged  by  the  amount  you  provision  in  IOPS  (multiplied  by  the  
percentage  of  days  you  provision  for  the  month.    
• Snapshot  -­‐  Amazon  EBS  provides  the  ability  to  back  up  snapshots  of  your  data  to  Amazon  S3  for  durable  
recovery.    If  you  opt  for  EBS  snapshots,  the  added  cost  is  per  GB-­‐month  of  data  stored.  
• Data  Transfer  -­‐  Take  into  account  the  amount  of  data  transferred  out  of  your  application.  Inbound  data  transfer  
is  free,  and  outbound  data  transfer  charges  are  tiered.  
For  more  information  about  Amazon  S3  pricing,  go  to  http://aws.amazon.com/pricing/s3/.  

Page  6  of  15    


Amazon  Web  Services  –  How  AWS  Pricing  Works   June  2015  

Amazon  Relational  Database  Service  (Amazon  RDS)    


Amazon  Relational  Database  Service  (Amazon  RDS)  is  a  web  service  that  makes  it  easy  to  set  up,  operate,  and  scale  a  
relational  database  in  the  cloud.  It  provides  cost-­‐efficient  and  resizable  capacity  while  managing  time-­‐consuming  
database  administration  tasks,  so  you  can  focus  on  your  applications  and  business.  

When  you  begin  to  estimate  the  cost  of  Amazon  RDS,  you  need  to  consider  the  following:  

• Clock  Hours  of  Server  Time  -­‐  Resources  incur  charges  when  they  are  running.  For  example,  from  the  time  you  
launch  a  DB  instances  until  you  terminated  the  DB  instance.  
• Database  Characteristics  -­‐  The  physical  capacity  of  the  database  you  choose  will  affect  how  much  you  are  
charged.  Database  characteristics  vary  depending  on  the  database  engine,  size,  and  memory  class.      
• Database  Purchase  Type  -­‐  When  you  use  On-­‐Demand  DB  Instances,  you  pay  for  compute  capacity  for  each  hour  
your  DB  Instance  runs,  with  no  required  minimum  commitments.  With  Reserved  DB  Instances,  you  can  make  a  
low,  one-­‐time,  up-­‐front  payment  for  each  DB  Instance  you  wish  to  reserve  for  a  1-­‐year  or  3-­‐year  term.    
• Number  of  Database  Instances  -­‐  With  Amazon  RDS,  you  can  provision  multiple  DB  instances  to  handle  peak  
loads.    
• Provisioned  Storage  -­‐  There  is  no  additional  charge  for  backup  storage  of  up  to  100%  of  your  provisioned  
database  storage  for  an  active  DB  Instance.  After  the  DB  Instance  is  terminated,  backup  storage  is  billed  per  
gigabyte  per  month.  
• Additional  Storage  -­‐  The  amount  of  backup  storage  in  addition  to  the  provisioned  storage  amount  is  billed  per  
gigabyte  per  month.  
• Requests  -­‐  The  number  of  input  and  output  requests  to  the  database.  
• Deployment  Type  -­‐  You  can  deploy  your  DB  instance  to  a  single  Availability  Zone  (analogous  to  a  stand-­‐alone  
data  center)  or  multiple  Availability  Zones  (analogous  to  secondary  data  center  for  enhanced  availability  and  
durability).  Storage  and  I/O  charges  vary,  depending  on  the  number  of  Availability  Zones  you  deploy  to.  
• Data  Transfer  -­‐  Inbound  data  transfer  is  free,  and  outbound  data  transfer  costs  are  tiered.  
 
Depending  on  the  needs  for  your  application,  it’s  possible  to  optimize  your  costs  for  Amazon  RDS  database  instances  by  
purchasing  reserved  Amazon  RDS  database  instances.  To  purchase  Reserved  Instances,  you  make  a  low,  one-­‐time  
payment  for  each  instance  you  want  to  reserve  and  in  turn  receive  a  significant  discount  on  the  hourly  usage  charge  for  
that  instance.      

For  more  information  about  Amazon  RDS  pricing,  go  to  http://aws.amazon.com/pricing/rds/.  

Amazon  CloudFront    
Amazon  CloudFront  is  a  web  service  for  content  delivery.  It  integrates  with  other  Amazon  Web  Services  to  give  you  an  
easy  way  to  distribute  content  to  end  users  with  low  latency,  high  data  transfer  speeds,  and  no  required  minimum  
commitments.  

When  you  begin  to  estimate  the  cost  of  Amazon  CloudFront,  you  need  to  consider  the  following:  

Page  7  of  15    


Amazon  Web  Services  –  How  AWS  Pricing  Works   June  2015  

• Traffic  Distribution  -­‐  Data  transfer  and  request  pricing  vary  across  geographic  regions,  and  pricing  is  based  on  
the  edge  location  through  which  your  content  is  served.      
• Requests  -­‐  The  number  and  type  of  requests  (HTTP  or  HTTPS)  made  and  the  geographic  region  in  which  the  
requests  are  made.    
• Data  Transfer  Out  -­‐  The  amount  of  data  transferred  out  of  your  Amazon  CloudFront  edge  locations.  
 
For  more  information  about  Amazon  CloudFront  pricing,  go  to  http://aws.amazon.com/pricing/cloudfront/.  

How  to  Further  Save  Costs  


Depending  on  the  needs  of  your  application,  you  can  optimize  your  costs  for  Amazon  EC2  instances  by  purchasing  EC2  
Reserved  Instances  or  Spot  Instances.  On-­‐Demand  Instances  are  a  good  option  if  you  run  your  Amazon  EC2  Instances a  
couple  of  hours  a  day  or  a  few  days  per  week;  however,  if  you  plan  to  run  your  Amazon  EC2  Instances  more  than  that,  
Reserved  Instances  can  save  you  money.  To  obtain  Reserved  Instances,  you  choose  duration  (1  or  3  years)  and  payment  
terms  (All  Upfront,  Partial  Upfront  or  No  Upfront)  for  each  instance  you  want  to  reserve,  and  in  turn  you  receive  a  
significant  discount  on  the  hourly  usage  charge  for  that  instance.  

You  can  use  Amazon  Elastic  Compute  Cloud  (Amazon  EC2)  Reserved  Instances  to  reserve  capacity  and  receive  a  discount  
on  your  instance  usage  compared  to  running  On-­‐Demand  instances.  The  discounted  usage  price  is  reserved  for  the  
duration  of  your  contract,  allowing  you  to  predict  compute  costs  over  the  term  of  the  Reserved  Instance.  When  you  
want  to  use  the  capacity  you  reserved,  you  launch  an  EC2  instance  with  the  same  configuration  as  the  reserved  capacity  
that  you  purchased.  Amazon  Web  Services  (AWS)  will  automatically  apply  the  usage  price  that  is  associated  with  your  
capacity  reservation.  You  are  charged  that  price  for  your  EC2  instance  for  every  hour  in  the  term  for  as  long  as  you  own  
the  Reserved  Instance.  

When  the  term  of  your  Reserved  Instance  ends,  and  you  do  not  renew  by  purchasing  another  Reserved  Instance,  you  
can  continue  using  the  EC2  instance  without  interruption.  However,  you  will  now  be  charged  at  the  On-­‐Demand  rate.  

To  purchase  an  Amazon  EC2  Reserved  Instance,  you  must  select  an  instance  type  (such  as  m1.small),  platform  
(Linux/UNIX,  Windows),  Availability  Zone,  tenancy  (dedicated  or  default),  and  term  (either  one-­‐year  or  three-­‐year).   The
Reserved  Instance  Marketplace  allows  other  AWS  customers  to  list  their  Reserved  Instances  for  sale.  3rd  party  Reserved  
Instances  are  often  listed  at  lower  prices  and  shorter  terms.  These  Reserved  Instances  are  no  different  than  Reserved  
Instances  purchased  directly  from  AWS.  

In  addition,  you  can  use  Auto  Scaling  or  other  AWS  services  to  launch  the  On-­‐Demand  instances  that  utilize  your  
Reserved  Instance  benefits  

The  following  table  shows  an  example  of  potential  savings  for  Amazon  EC2  Reserved  Instances  for  a  1-­‐year  and  3-­‐year  
term  running  a  single  m3.large  Linux  instances  in  the  US  East  Region.      

Page  8  of  15    


Amazon  Web  Services  –  How  AWS  Pricing  Works   June  2015  

Table  1:  Savings  Comparison  of  1  and  3  Year  Reserved  Instances  over  On-­‐Demand  Instances    

Prices  shown  for  US  East  Region  as  of  July  20th  2014  

   
As  shown  in  table  1,  savings  by  paying  1  year  Partial  or  all  Upfront  Reserved  Instances  will  save  on  average  38%  over  on  
demand  purchasing.  For  3  year  RI’s,  the  savings  on  average  will  exceed  60%  and  29%  for  No  Upfront.    

Why  Should  I  Use  Reserved  Instances?  


Cost  Effective  
Reserved  Instances  provide  you  with  a  significant  discount  (over  60%)  compared  to  On-­‐Demand  Instance  pricing.    
Reliable  
Reserved  Instances  provide  a  capacity  reservation  so  that  you  can  have  confidence  in  your  ability  to  launch  the  number  
of  instances  you  have  reserved  when  you  need  them.  

Flexible  
You  have  the  flexibility  to  pay  all,  part,  or  nothing  upfront.  The  more  you  pay  up  front,  the  more  you  save.  If  your  
requirements  change,  you  can  modify  or  sell  your  Reserved  Instance.  
     
Your  Reserved  Instance  will  be  available  for  the  operating  system  (Linux/UNIX,  SUSE  Linux  Enterprise,  Red  Hat  Enterprise  
Linux  or  Windows)  and  Availability  Zone  in  which  you  purchased  it.  For  more  information  about  Reserved  Instances,  go  
to  Amazon  EC2  Reserved  Instance  pages.  

Spot  Instances  are  unused  Amazon  EC2  capacity  that  you  bid  for.  Instances  are  charged  at  Spot  Price,  which  is  set  by  
Amazon  EC2  and  fluctuates  periodically  depending  on  the  supply  of,  and  demand  for,  Spot  Instance  capacity.  If  your  
maximum  bid  exceeds  the  current  Spot  Price,  your  bid  request  is  fulfilled,  and  your  instances  will  run  until  either  you  
choose  to  terminate  them  or  the  Spot  Price  increases  above  your  maximum  bid,  whichever  is  sooner.  To  learn  more  
about  Spot  Instances,  go  to  http://aws.amazon.com/ec2/spot-­‐instances/.  

Conclusion  
While  the  number  and  types  of  services  offered  by  AWS  has  increased  dramatically,  our  philosophy  on  pricing  has  not  
changed.  You  pay  as  you  go,  pay  for  what  you  use,  pay  less  as  you  use  more,  and  pay  even  less  when  you  reserve  
capacity.  Projecting  costs  for  a  use  case,  such  as  web  application  hosting,  can  be  challenging,  because  a  solution  typically  

Page  9  of  15    


Amazon  Web  Services  –  How  AWS  Pricing  Works   June  2015  

uses  multiple  features  across  multiple  AWS  products,  which  in  turn  means  there  are  more  factors  and  purchase  options  
to  consider.    

The  best  way  to  estimate  costs  is  to  examine  the  fundamental  characteristics  for  each  AWS  product,  estimate  your  
usage  for  each  characteristic,  and  then  map  that  usage  to  the  prices  posted  on  the  website.  To  help  you  understand  how  
AWS  pricing  works  in  the  context  of  real-­‐world  solutions,  see  a  worked  out  example  later  in  this  paper.    

You  can  use  the  AWS  Simple  Monthly  Calculator  to  estimate  your  monthly  bill.  The  calculator  provides  per  service  cost  
breakdown,  as  well  as  an  aggregate  monthly  estimate.  You  can  also  use  the  calculator  to  see  an  estimation  and  
breakdown  of  costs  for  common  solutions.    

AWS  has  introduced  a  free  usage  tier  to  help  you  get  started  with  AWS.  Take  advantage  of  the  Free  Usage  Tier,  and  get  
started  today!    

   

Page  10  of  15    


Amazon  Web  Services  –  How  AWS  Pricing  Works   June  2015  

Resources  
Resource   Description  
AWS  Simple  Monthly  
Calculator   The  AWS  Simple  Monthly  Calculator  helps  estimate  your  monthly  bill.  

The  AWS  Architecture  Center  provides  you  with  the  necessary  guidance  and  
AWS  Architecture  Center  
best  practices  to  build  highly  scalable  and  reliable  applications  in  the  AWS  
Cloud.  
The  AWS  Economics  Center  provides  access  to  information,  tools,  and  
AWS  Economics  Center  
resources  to  compare  the  costs  of  Amazon  Web  Services  with  IT  
infrastructure  alternatives.  
AWS  Account  Activity  Page   View  your  current  charges  and  account  activity,  itemized  by  service  and  by  
usage  type.  Previous  months’  billing  statements  are  also  available.  
Usage  reports  are  available  to  download  for  each  service.  Specifying  usage  
AWS  Usage  Reports  
types,  timeframe,  service  operations,  and  more  can  customize  reports.  

 
   

Page  11  of  15    


Amazon  Web  Services  –  How  AWS  Pricing  Works   June  2015  

Cost  Calculation  Example  


This  section  of  the  document  uses  the  AWS  Simple  Monthly  Calculator  to  provide  costing  example  of  a  common  use  case  
of  a  dynamic  website  hosted  on  AWS  using  Amazon  EC2,  Auto  Scaling,  and  Amazon  RDS.  The  Amazon  EC2  instance  runs  
the  web  and  application  tiers,  and  we  use  Auto  Scaling  to  match  the  number  of  instances  to  the  traffic  load.  Amazon  RDS  
uses  one  DB  instance  for  its  primary  storage.  This  DB  instance  is  deployed  across  multiple  Availability  Zones.  

Architecture  

Elastic  Load  Balancing  balances  traffic  across  one  or  more  Amazon  EC2  instances.  The  Amazon  EC2  instances  belong  to  
an  Auto  Scaling  group,  and  the  Auto  Scaling  group  either  adds  or  subtracts  Amazon  EC2  instances,  depending  on  
variations  in  the  traffic  load.  Deploying  Amazon  RDS  across  multiple  Availability  Zones  enhances  data  durability  and  
availability.  Amazon  RDS  provisions  and  maintains  a  standby  in  a  different  Availability  Zone  for  automatic  failover  in  the  
event  of  planned  or  unplanned  outages.  The  following  illustration  shows  the  example  architecture  for  a  dynamic  website  
using  Amazon  EC2,  Auto  Scaling,  and  one  Amazon  RDS  database  instance  across  multiple  Availability  Zones.    

 
Figure  1:  Complex  Dynamic  Site  Architecture  

Daily  Usage  Profile  

You  can  monitor  daily  usage  for  your  application  so  that  you  can  better  estimate  your  costs.  For  instance,  you  can  look  
at  the  daily  pattern  to  figure  out  how  your  application  handles  traffic.  For  each  hour,  track  how  many  hits  you  get  at  
your  website.  Then  track  how  many  instances  are  running.  Add  up  the  total  number  of  hits  for  that  day.  Examine  the  

Page  12  of  15    


Amazon  Web  Services  –  How  AWS  Pricing  Works   June  2015  

number  of  Amazon  EC2  instances  that  run  each  hour,  and  then  take  the  average.  You  can  use  the  number  of  hits  per  day  
and  the  average  number  of  instances  for  your  calculations.  

Amazon  EC2  Cost  Breakdown  

The  following  table  shows  the  characteristics  for  Amazon  EC2  we  have  identified  for  this  dynamic  site.  

Characteristic   Estimated  Usage   Description  


24  hrs.  /day   At  an  average  of  30.5  days  in  a  month,  the  instance  runs  
Clock  Hours  of  Server  Time  
732  hours/month  
m3.medium  instance,  
Machine  Characteristics   3.75  GB  of  memory,  1  vCPU,  4  GB  of  SSD  storage  
ephemeral  storage  
For  this  application,  we  need  only  the  local  storage  which  
Additional  Storage   No  EBS  volumes  
is  packaged  with  the  AMI.  
There  are  approximately  1,000,000  hits  per  day.  Each  
Data  In:  5  GB/day   response  transfers  out  about  50  KB,  and  each  request  
Data  Transfer  
Data  Out:  50  GB  /day   transfers  in  about  5  KB.  

Instance  Scale   4   On  average  in  a  given  day,  there  are  4  instances  running.  
Hourly  usage:  732  
Elastic  Load  Balancing  is  used  24  hours/day,  7  days/week.  
hrs./month  
Elastic  Load  Balancing   Elastic  Load  Balancing  processes  a  total  of  55  GB/day  
Data  processed:  
(data  in  +  data  out)  
1677.5  GB/month  
We  use  Elastic  Load  Balancing  to  balance  traffic  across  
Elastic  IP  Address   None  
instances.  
Detailed  monitoring  is  not  enabled,  so  we  get  basic  
Detailed  Monitoring   None  
monitoring  free  of  charge.  
 

The  total  cost  for  one  month  is  the  sum  of  the  cost  of  the  running  instances,  overall  AWS  data  transfer,  Elastic  Load  
Balancers,  and  the  data  processed  by  the  Elastic  Load  Balancers.  

Variable   Formula   Calculation  


  Instance  cost  per  hour   $0.070  
       Number  of  instances          4  
Amazon  EC2  Instance  Cost  
     X                        Uptime  in  hours        X                        732  
$204.96    
(Hours  used  X  Hourly  Rate)   732  X  $0.025      
Elastic  Load  Balancer  Cost        +(Data  processed  (GB)  X  Process  Rate)        X                1677.5  X  $0.008    
$31.72  
(Data  in  (GB)  X  Data  In  Rate)   152.5  X  $0.00      
AWS  Data  Transfer  Cost        +(Data  out  (GB)  X  Data  Out  Rate)        +              (1525-­‐1)  1  X  $0.12    
  $182.88  
Estimated  Cost  for  Amazon  EC2     $419.56  
                                                                                                                       
1
 Since  up  to  one  gigabyte  is  free  per  month,  we  subtract  one.  
 

Page  13  of  15    


Amazon  Web  Services  –  How  AWS  Pricing  Works   June  2015  

We  use  the  AWS  Simple  Monthly  Calculator  to  estimate  this  cost.  Using  the  calculator,  we  arrive  at  the  total  cost  for  one  
month’s  usage  for  Amazon  EC2  of  $419.562.  

Amazon  RDS  Cost  Breakdown  

The  following  table  shows  the  characteristics  for  Amazon  RDS  we  have  identified  for  this  complex  dynamic  site.  

Characteristic   Estimated  Usage   Description  


Clock  Hours  of  Server  Time   24  hrs/day   24*30.5  =  732  hours  per  month  

Database  Characteristics   Medium  RDS   3.75  GiB  memory,  1  virtual  core,  Moderate  network  
instance   performance  

Provisioned  Storage   100  GB/month   Amazon  provides  5  GB  to  1  TB  of  associated  storage  
capacity  for  your  primary  data  set.  

Requests   300M  I/O   We  have  1,000,000  hits  per  day  at  a  rate  of  5  I/O  
requests/month   requests  per  hit  on  site  with  30.5  days  in  a  month  on  
average.  This  is  a  total  of  150M  I/O  requests  per  
  month,  but  since  the  write  I/O  request  will  double  
because  data  is  also  replicated  to  the  standby  
instance,  we  have  a  total  of  300M.  
Deployment  Type   Multiple   We  will  run  our  database  instance  across  multiple  
Availability  Zones   Availability  Zones.  
Additional  Backup  Storage   None   We’ll  use  up  to  the  provisioned  amount,  which  is  100  
GB.        
Data  Transfer   Data  in:  0  GB   There  is  no  data  transfer  from  RDS  to  the  Internet.  
Data  out:  0  GB  
Database  Instance  Scale   We  need  one  database  instance.   1  
Special  Features   There  are  no  reserved  database  instances.   None  
 
Because  we  do  not  have  data  transfer  in  or  out,  and  we  do  not  have  additional  backup  storage,  the  total  cost  for  one  
month  is  the  sum  of  the  cost  of  the  running  instances,  provisioned  storage,  and  I/O  requests.  

Variable   Formula   Calculation  


Instance  Cost     Instance  cost  per  hour   $0.180  
       Number  of  instances          1  
     X                        Uptime  in  hours    X                        732  
$131.76  
Provisioned  Storage   Storage  rate   $0.20  
     X                      Storage  Amount  (GB)   X                        100  
  $20.00  
                                                                                                                       
2
 Depending  on  regional  factors  and  special  offers,  the  cost  you  get  from  the  calculator  will  be  slightly  different.  

Page  14  of  15    


Amazon  Web  Services  –  How  AWS  Pricing  Works   June  2015  

I/O  Requests   I/O  rate   $0.10  


     X          Number  of  requests    X        300                        
  $30.00  
Estimated  cost  for  Amazon  RDS     $181.76  
 
We  use  the  AWS  Simple  Monthly  Calculator  to  estimate  this  cost.  The  total  cost  for  one  month’s  usage  for  Amazon  RDS  
portion  of  this  scenario  is  $181.763.  

Total  Cost  

To  calculate  the  total  cost  for  this  example,  we  add  the  cost  for  Amazon  EC2,  Amazon  RDS,  and  AWS  Data  Transfer  Out  
and  subtract  any  discount  that  falls  into  the  AWS  Free  Usage  Tier.    The  total  cost  of  the  dynamic  site  in  this  scenario  is  
estimated  at  $601.3213  per  month.    

                                                                                                                       
3
 Depending  on  regional  factors  and  special  offers,  the  cost  you  get  from  the  AWS  Simple  Monthly  Calculator  will  be  slightly  
different.  
13
 Depending  on  regional  factors  and  special  offers,  the  cost  you  get  from  the  AWS  Simple  Monthly  Calculator  will  be  slightly  
different.  

Page  15  of  15    

You might also like