google cloud offers below services:
- compute
- storage
- big data
- machine learning
- application services
at the end of this notes you will be able to :
- identify the value of the google cloud products
- use application deployment environment on google cloud
- use google storage options
- interact with google cloud service
- describe the ways in which customers use google cloud .
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>.
TOPICS:
cloud computing
IAAS ,PAAS,SAAS
pricing and billing
google cloud hierarchy
IAM
VPC
compute engine
scaling virtual machines through load balancer
cloud DNS and CDN
google cloud storage options
storage classes and data transfer
cloud SQL
cloud spanner
Firestore
Bigtable
comparing storage options
containers in cloud
kubernetes
google kubernetes engine
hybrid and multi cloud
Anthos
App Engine
cloud run
development in cloud
deployment : insfrastructure as code
automating the deployment using the terraform.
monitoring
measuring the performance and reliability
integrated observability tools
monitoring tools
logging tools
cloud computing :
- customers get on demand service
- customers gets access to those service from anywhere.
- provider of those resources allocates them to users out of that pool
- resources are elastic and flexible
- customers only pay for what they use
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
Google cloud network :
it is designed to achieve the :
- highest possible throughput
- lowest latency
- 100 plus content caching nodes worldwide
- high demand content is cached for quiker access.
google cloud infrastructure is based on 5 major geographic locations:
- north america
- south america
- europe
- asia
- australia
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>..
latency :
it measures the time a packet of information takes to travel from its source to destination.
location is divided into regions and regions are divided into availability zones .
EX: Europe > london> (region: europe-west2)> europe-west2-a, europe-west2-b
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>.
Google cloud and AWS comparison :
| Features | Google Cloud | Amazon Web Services |
|---|---|---|
| Offered By | Amazon | |
| Computing Service | Google Compute Engine API (IaaS), App Engine (PaaS), Kubernetes Engine (Container), Cloud Functions (Serverless Functions) | Amazon Elastic Compute Cloud (IaaS), Elastic Beanstalk(PaaS), Elastic Compute Cloud Container Service (Container), AWS Lambda (Serverless Functions) |
| Database Services | Google Cloud SQL (RDBMS), Google Cloud Datastore, Google Cloud Bigtable (NoSQL Key–Value), Google Cloud Datastore (NoSQL: Indexed) | Amazon Relational Database Service (RDBMS), Amazon DynamoDB (NoSQL Key–Value), Amazon SimpleDB (NoSQL Indexed) |
| Storage Services | Google Cloud Storage (Object Storage), Google Compute Engine Persistent Disks (Block Storage), ZFS/Avere (File Storage), Google Cloud Storage Nearline (Cold Storage) | Amazon Simple Storage Service (Object Storage), Amazon Elastic Block Store (Block Storage), Amazon Elastic File System (File Storage), Amazon Glacier (Cold Storage) |
| Management Services | Stackdriver Monitoring (Monitoring), Google Cloud Deployment Manager (Deployment) | Amazon CloudWatch (Monitoring), AWS CloudFormation (Deployment) |
| Network Services | Virtual Private Cloud, Google Cloud Load Balancing, Google Cloud Interconnect, Google Cloud DNS | Amazon Virtual Private Cloud, Elastic Load Balancer, Direct Connect, Amazon Route 53 |
| Customization of instances | Google Cloud Platform provides a wide range of customization for any Instance | AWS provides limited customization. |
| Pricing | Google charges per minute basis | Amazon charges per hour basis |
| Cost | Google free tiers have no time limit. GCP provides $300 worth credit that can be used across all services. Hence, GCP is comparatively cheaper. | Amazon free tiers have a maximum validity of 12 months and later charges as per usage. Hence, AWS is costlier. |
| Downtime | GCP had reported more downtime compared to AWS | AWS had reported lesser downtime compared to GCP which makes it a clear winner in this case |
| Big data support | Big data analysis tool is AI First | Big data analysis tool is AWS Lambda. |
| AI/ML Support | Cloud Machine Learning Engine, Dialogflow Enterprise Edition, Cloud Natural Language, Cloud Speech API, Cloud Translation API, Cloud Video Intelligence, Cloud Job Discovery | Tools offered by AWS for AI/ML are SageMaker, Comprehend, Lex, Polly, Rekognition, Machine Learning, Translate, Transcribe, DeepLens, Deep Learning AMIs, Apache MXNet, TensorFlow |
| Availability | GCP is available in 29 geographic regions and 88 zones worldwide | AWS is available at 26 geographic regions and 84 zones worldwide |
| Companies using | Spotify, HSBC, Home Depot, Snapchat, Philips, Coca Cola, Domino’s and many more | Netflix, Twitch, LinkedIn, Facebook, ESPN, Citrix, Expedia and many more |
- per second billing
- running an compute engine instance for more than 25 percent of the month gives you the discount for every incremental minute you use for that instance.
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>..
how can I make sure I dont accidently run up a big google cloud bill ?
- we can set our budget .
- we can create an alert
- we can check reports in GCP
- GCP offers quotas , which avoids over consumption of the resources
- projects are separate entities in google cloud .
- project holds resources ,each of which belongs to one project .
- project can have different owners and users .
- projects are billed and managed separately.
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>.
Google cloud organization heirarchy :
organisation node
|
folder
- who has access to what should be defined so IAM comes into existence here.
- administrators can apply policies that define who can do what and on which resources.
- who can be : google account ,google group ,service account , cloud identity domain.
- what can be : role
- role : collection of the permissions. you grant a role to someone, you grant the collection of permissions to someone.
IAM roles :
- basic IAM role : contiains owner, editor, viewer, billing admin
- if several people are working together on the project that contains the sensitive data, basic roles are probably too broad.
- predefined role:
- if you want someone only to have an access to virtual machines. we can give him instance admin role which contains the following permission :
- get instances , delete instances , start instances ,stop instance etc.
- custom IAM role : you want someone to have permission to start and stop the instance , but not configure them. in that case we can create the custom IAM role. like : instance operator who will have permissions like :
- get instances , list instances, start instances , stop instances, and will not have the permission to configure the instances like changing the memory of the instance.
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
what if you want to give the permissions to virtual machine and not the person ?
you have applicaiton running on the virtual machine that is storing the data in cloud storage, but you dont want anyone on the internet to have access to that data. you can create service account to authenticate that VM to cloud storage.
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
How to access the google cloud ?
- google cloud console
- cloud sdk and cloud shell
- APIs
- cloud mobile app
SDK :
- set of tools to manage the resources and applications hosted on the google cloud.
- gcloud tool: provides the main CLI to access the google cloud services .
- gsutil : provide the access to cloud storage from the command line.
- bq : command line tool for big queery
- cloud shell : provides cli from browser to access the resources.
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>.
- connects google cloud resources to each other and to the internet.
- segmenting the networks
- using firewall rules to restrict the access to instances .
- creating the static route to forward traffic to specific destinations
- VPC networks are global and can have the subnets in any google cloud region.
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>....
Google compute engine :
- can create and run virtual machines on google infrastructure.
- each vm contains the power and functionality of a full fledged operating system.
- can be configured much like the physical server .
- can be created using the CLI , google cloud console , or compute engine API.
- can run windows server images and linux images, or any customised images.
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
google cloud storage products :
- cloud storage
- cloud SQL
- cloud spanner
- firestore
- cloud bigtable
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>.
Cloud Storage :
- offers durable and highly available object storage ,its not a file or the block storage.
- allow customer to store any amount of data.
- fully managed scalable service that has a wide variety of users.
- its a binary large object storage
- online content can be saved
- can be used for backup and archiving
- storage of intermediate results
- objects are stored into buckets.
- objects in bucket are immutable. which means if you update the object new version is created for that object . original object is not changed .
- cloud storage offers lifecycle policies: ex : delete the objects older than 365 days. or create the objects created before particular date. or keep only the 3 most recent versions.
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>.
storage classes in cloud storage :
- standard storage:
- best for frequently accessed data.
- nearline storage :
- best for infrequently accessed data. once per month
- coldline storage :
- best for the data which is accessed once per quarter.
- archive storage :
- best for the data which is accessed once per year.
Comments
Post a Comment