Airbnb on Amazon Web Services
There are so many companies that were benefitted by AWS starting from startups, MNCs and big enterprises. Airbnb is one such company among them.
Airbnb: Airbnb is a community marketplace that allows property owners and travellers to connect each other for the purpose of renting houses and vacation spaces.
It is a growing company that lists more than 7 million accommodation and it operates in nearly 25,000 cities and 192 countries. Hence it has to process and analyze around 50 GB of data daily and store nearly 10TB of static files including user images etc.
Like every other company it started from On-premise servers,data centers and After one year from launching it realized the pain and cost of managing, administration and scaling of hardware whenever required.
Finally in 2008 Airbnb moved to AWS and currently has around 200 EC2 instances for computing,Load balancing etc.
AWS: Amazon web services(AWS) is a cloud platform that offers services like Iaas(Infrastructure as a service)Paas(Platform as a service) and Saas(Software as a service).
AWS offers vast number of services from computing, storage, networking, databases, security to Machine learning, analytics, deployment, Content delivery networking, DNS services etc. It offers around 200 products and services.
The main reason behind the popularity of AWS is that providing almost all services and the integration between these services. Almost all IT operations can be done on AWS much like On-premise but with most cost-effectively and efficiently
Top Services used by Airbnb:
- Amazon EC2:
For hosting anything or for computation one needs server(system) that can be launched as bare metal,Virtual Machine or container.Managing these servers and increasing the hardware when needed is tedious task .Amazon EC2(Elastic Compute Cloud) is a virtual server that offers computing services much like normal PC and has storage from EBS. The instance can be launched with desired AMI (Amazon Machine Image like OS)within secs.
- Elastic Load Balancer(ELB):
Elastic Load Balancer automatically distributes incoming application traffic across multiple targets, such as Amazon EC2 instances.AWS provides 3 types of ELBs used according to the need.
- Auto scaling group:
Whenever load increases the servers need to be scaled immediately for 0 downtime. This can be made efficiently and quickly with Auto scaling groups which scales the instances based on some metrics specified.The entire thing is done automatically without human interaction.
- Amazon RDS(Relational database service):
Before switched to RDS, Airbnb was running on a single MySQL instance on top of Amazon EC2. They have dozens of background tasks that handle things like payments and analytics. Many of the queries that are run by these tasks are expensive SELECTs, and they starting to notice an impact on frontend performance.
It can be solved by having a read only server aside that does asynchronous replication and forwarding the read queries to that server(EC2).This also works for some range only.
Because of the above reason and the reliability of EC2 ,Airbnb shifted from Mysql to RDS.
It is as easy to setup as with a click and an API call and supports multi-AZ ,read replicas that can be maintained in different AZs
It also supports point-in-time recovery from snapshots.RDS takes snapshots of entire database by default and thus it not only helps with automated backups but also from data recovery.
This helps to know more about How RDS is more suitable
- Amazon S3(simple storage service):
It is an object storage service used to store any type of data and guarantees 99.999999999% of durability and 99.99% availability.It is like Gdrive and data stored in S3 can be accessed from anywhere.
It stores objects(files) in buckets(like folders) and by default replicates the buckets to ≥3 AZs based on the storage class selected.
Airbnb is using Amazon Simple Storage Service (Amazon S3) to house backups and static files, including 10 terabytes of user pictures
In addition to above services Airbnb uses Elastic Map Reduce service used to easily process and analyze 50 Gigabytes of data daily and Cloud Watch which allows the company to easily supervise all of its Amazon EC2 assets.
