Introduction to Database Systems CSE 444 Lecture 24 Databases as a Service YongChul Kwon CSE 444 Fall 2010 References Amazon SimpleDB Website Part of the Amazon Web services Google App Engine Datastore Website Part of the Google App Engine Microsoft SQL Azure Website Part of the Azure platform CSE 444 Fall 2010 Motivation Running a DBMS is challenging Need to hire a skilled database administrator DBA Need to provision machines hardware software configuration If business picks up may need to scale quickly In general workload varies over time Solution Use a DBMS service All machines are hosted in service provider s data centers Data resides in those data centers Pay per use policy Elastic scalability Zero administration CSE 444 Fall 2010 Basic Features Data storage and query capabilities High availability guarantees Operations and admin tasks handled by provider Elastic scalability Clients pay exactly for the resources they consume consumption can grow shrink dynamically No capital expenditures Fast provisioning CSE 444 Fall 2010 Outline Database in the Cloud DIY Database in the Cloud RDBMS as a Service Database as a Service Discussion Technical challenges behind databases as a service Broader impacts of databases as a service CSE 444 Fall 2010 Outline Database in the Cloud DIY Database in the Cloud Amazon EC2 Database software RDBMS as a service Database as a service CSE 444 Fall 2010 DIY Database in the Cloud Setting up a database server Buy land building room server rack Buy computers and network gears Install the hardware Install OS Install database server How can we do this using the cloud CSE 444 Fall 2010 Amazon Web Services Since 2006 Infrastructure web services platform in the cloud Amazon Elastic Compute Cloud Amazon EC2 Amazon Simple Storage Service Amazon S3 Amazon SimpleDB Amazon CloudFront Amazon Simple Queue Service Amazon SQS CSE 444 Fall 2010 Amazon EC2 Amazon Elastic Compute Cloud Amazon EC2 Rent compute power on demand server instances Select required power small large or extra large instance Share resources with other users Variety of operating systems Includes Amazon Elastic Block Store Off instance storage that persists independent from life of instance Highly available and highly reliable CSE 444 Fall 2010 Amazon S3 Amazon Simple Storage Service Amazon S3 Storage for the Internet Web services interface that can be used to store and retrieve any amount of data at any time from anywhere on the web Some key features Write read and delete uniquely identified objects containing from 1 byte to 5 gigabytes of data each Objects are stored in buckets located in US or Europe A bucket can be accessed from anywhere Authentication Reliability CSE 444 Fall 2010 Let s hook up what we ve got Buy an EC2 instance Buy database software Pre licensed Amazon Machine Image Launch your EC2 instance Hooray Took less than 5 minutes to here CSE 444 Fall 2010 Summary DIY Database in the Cloud Pros Fast provision of resource Full control over software Choose your favorite database software Cons Full control over software Operating system backup fault tolerance security performance tuning scaling Can cloud provider manage such detail CSE 444 Fall 2010 Outline Database in the Cloud DIY Database in the Cloud RDBMS as a service Amazon Relational Database Service Microsoft SQL Azure Database as a service CSE 444 Fall 2010 RDBMS as a Service Cloud provider manages Operating system Backup Fault tolerance Scaling User manages Schema Data Indexes Use your favorite SQL Two services Amazon Relational Database Service Microsoft SQL Azure CSE 444 Fall 2010 Amazon Relational Database Service Features Backed by MySQL Easy administration Backup replication adding more resources CPU disk Automated or initiated by AWS API or via management console Database size 5GB 1TB per instance How to scale Adding more resources to master Adding more read only replicas CSE 444 Fall 2010 Microsoft Windows Azure Platform Internet scale cloud computing and services platform Provides an operating system and a set of developer services that can be used individually or together Windows Azure Platform Windows Azure SQL Azure AppFabric CSE 444 Fall 2010 DataMarket SQL Azure Virtualized SQL Azure Server cluster Azure manages physical cluster SQL Azure Server A logical SQL Server Automatically replicated Hosts up to 149 databases 5 GB or 50 GB per database How to scale Partition across multiple databases Automatically distributed across multiple machines CSE 444 Fall 2010 Summary RDBMS as a Service Remove burden of database administration Amazon RDS automate common administration tasks SQL Azure separate logical physical database cluster Pros No change to existing applications Support full SQL transaction consistencies Highly available scalable Cons Managing schema indexes partitioning tuning CSE 444 Fall 2010 Outline Database in the Cloud DIY Database in the Cloud RDBMS as a service Database as a service Amazon SimpleDB Google App Engine Google Fusion Table Big Query CSE 444 Fall 2010 Database as a Service RDBMS as a Service is good But user still has to know a lot about database Can cloud manage following Schema Partitioning Indexing Two services Amazon SimpleDB Google App Engine CSE 444 Fall 2010 Amazon SimpleDB Web service providing the core database functions of data indexing and querying Partitioning Data partitioned into domains queries run within domain Schema No fixed schema Objects are defined with attribute value pairs CSE 444 Fall 2010 Amazon SimpleDB 2 3 Indexing Automatically indexes all attributes Support for writing PUT and DELETE items in a domain Support for querying select output list from domain name where expression sort instructions limit limit GET by key Selection sort A simple form of aggregation count Query execution time is limited to 5 second but can continue CSE 444 Fall 2010 Amazon SimpleDB 3 3 Availability and consistency Fully indexed data is stored redundantly across multiple servers and data centers Takes time for the update to propagate to all storage locations The data will eventually be consistent but an immediate read might not show the change Consistent read Conditional Put Delete Integration with other services Developers can run their applications in Amazon EC2 and store their data objects in Amazon S3 Amazon SimpleDB can then be used to query the object metadata from within the application in Amazon EC2 and return pointers to the objects stored in Amazon S3 CSE 444 Fall 2010 Google App
View Full Document
Unlocking...