Google Cloud Platform(GCP) : Part- 1

Cloud Computing

Cloud  computing  is  a way  of using  I.T.  that  has these  five important traits:

• Get  computing  resources  on-demand  and  self-service.
• Access  these  resources  over  the  internet  from  anywhere  we  want
• The  provider of  these  resources has a  big  pool  of them  and  allocates  them
• Resources are elastic
• Pay only for what we use

GCP Architectures

Virtualize  data centers  brought  us Infrastructure  as  a Service,  IaaS,  and  Platform  as  a  Service, PaaS offerings.
IaaS  offerings provide  raw,  compute,  storage,  and  network  organized in  ways  that  are  familiar from  data  centers.
PaaS  offerings on  the  other  hand, bind  application  code  we write  to  libraries that give  access  to  the  infrastructure  our application  needs.
In  the  IaaS  model, we  pay  for what  we allocate. In  the  PaaS  model, we pay  for  what  we  use. Google's popular  applications  like, Search, Gmail,  Docs  and  Drive  are  Software  as a Service  applications.

Google Network

It's designed  to  give  its  users  the  highest possible throughput  and  the  lowest  possible latencies  for their applications. When  an  Internet user  sends  traffic  to  a Google resource,  Google  responds to  the user's request  from  an  edge  network location  that  will provide  the  lowest  latency. Google's  Edge-caching  network  cites content close  to  end users  to  minimize  latency.

GCP Regions and Zones

A zone  is a  deployment  area  for Google  Cloud  Platform  Resources. Zones  are  grouped into  regions, independent  geographic areas,  and  we  can  choose  what regions  our GCP  resources  are  in. All  the zones  within  a  region  have  fast  network  connectivity  among  them. Locations  within  regions usually  have  round  trip  network latencies  of under  five  milliseconds. Zone  is  a single failure  domain  within  a  region. As part  of  building  a  fault tolerant  application,  we  can spread the resources  across multiple  zones  in  a  region. That helps protect  against  unexpected failures. We  can  run  resources  in different regions  too.  Lots  of  GCP  customers  do  that,  both to bring  their applications closer  to  users  around  the  world, and  also  to  protect  against  the  loss of an  entire  region, say, due  to  a  natural disaster.
A few  Google  Cloud  platform  services  support  placing  resources  in  what we call  a  Multi-Region. For  example, Google cloud  storage  lets  us to  place  data within  the  Europe  Multi-Region. That  means, it  will be stored  redundantly  in at  least  two  geographic locations, separated  by  at  least  160  kilometers within  Europe.

Pricing

Google  was  the  first  major  cloud  provider  to  build  by  the second,  rather  than  rounding  up  to  bigger  units  of  time  for  its  virtual  machines as  a  service  offering. Google offers per second billing. Charges  for  rounding  can  really  add  up  for  customers  who  are  creating and  running  lots  of virtual  machines. Computer  engine  offers  automatically  applied sustained use discounts,  which  are  automatic  discounts that we get  for running  a  virtual machine  for a significant  portion  of the  billing  month.  When  we run  an instance  from  more  than  25  percent  of a  month, computer engine  automatically  gives  us a  discount  for every  incremental  minute  we  use  it.

Open APIs

Google helps  its  customers  avoid  feeling  locked  in. GCP  services  are  compatible with  open source products. For example, Bigtable  uses  the interface  of the open source  database  Apache  HBase,  which  gives  customers  the  benefit  of code  portability. Another example, Cloud  Dataproc  offers the  open source  big  data environment  Hadoop,  as a  managed service. Google  publishes key  elements  of technology  using  open source  licenses  to  create  ecosystems that provide  customers  with  options  other  than  Google. For example,  TensorFlow is  an  open source  software library  for  machine  learning  developed inside  Google.
Many GCP  technologies  provide  interoperability. Kubernetes  gives  customers  the  ability  to  mix and match  microservices  running  across  different  clouds,  and  Google Stackdriver  lets  customers  monitor workload  across  multiple cloud  providers.

Why GCP

Google  Cloud  Platform  lets  us choose  from  computing, storage,  big  data,  machine  learning  and application  services  for  web,  mobile,  analytics  and  backend  solutions.  It's global, it's  cost  effective, it's open  source  friendly  and  it's designed for security. Google Cloud  Platforms products and  services  can  be  broadly  categorized  as  compute,  storage, big  data,  machine  learning, networking and  operations and  tools.

We are born not to be Average, 

We are born to be Awesome..