Amazon SimpleDB vs Amazon DynamoDB
Can someone simply explain the main differences between them and tell in which cases to choose one over the other.
This is addressed by the respective FAQ Q: How does Amazon DynamoDB differ from Amazon SimpleDB? Which should I use? (hash link no longer works, but use in-page Find to locate question within page) to some extent already, with the most compact summary at the end of the paragraph:
While SimpleDB has scaling limitations, it may be a good fit for smaller workloads that require query flexibility. Amazon SimpleDB automatically indexes all item attributes and thus supports query flexibility at the cost of performance and scale.
So it's a trade off between performance/scalability and simplicity/flexibility, i.e. for simpler scenarios it might still be easier getting started with SimpleDB to avoid the complexities of architecturing your application for DynamoDB (see below for a different perspective).
The linked FAQ entry references Werner Vogel's Amazon DynamoDB – a Fast and Scalable NoSQL Database Service Designed for Internet Scale Applications as well, which is indeed an elaborate and thus highly recommended read concerning the History of NoSQL at Amazon in general and Dynamo in particular; it contains many more insights addressing your question as well, e.g.
It became obvious that developers [even Amazon engineers] strongly preferred simplicity to fine-grained control as they voted "with their feet" and adopted cloud-based AWS solutions, like Amazon S3 and Amazon SimpleDB, over Dynamo. [addition mine]
Obviously DynamoDB has been introduced to address this and could thus be qualified as a successor of SimpleDB rather than 'just' amending their existing NoSQL offering:
We concluded that an ideal solution would combine the best parts of the original Dynamo design (incremental scalability, predictable high performance) with the best parts of SimpleDB (ease of administration of a cloud service, consistency, and a table-based data model that is richer than a pure key-value store).
Werner's Summary suggests DynamoDB to be a good fit for applications of any size now accordingly:
Amazon DynamoDB is designed to maintain predictably high performance and to be highly cost efficient for workloads of any scale, from the smallest to the largest internet-scale applications.
use SimpleDB or DynamoDB, it depends on your use case, I shared some of my experience using SimpleDB in some cases instead of DynamoDB. In another product, I used both SimpleDB and DynamoDB to store different data.
SimpleDB does not seem to be getting any love from Amazon these days - its hard to even find where to provision it in the AWS Console. Seems like SimpleDB is no longer being iterated on - use DynamoDB as your first choice for a Document Database on AWS.
SimpleDb is no longer really "iterated on". Meaning that there is no new development in the future for simpledb. It is "maintained and supported", but it won't be getting any better.
One of the differences used to be (as @Mason Zhang states in his article above) in indexing. DynamoDB used to limit you to creating indexes at the time of creating the table. However, now (since early 2014), there is the concept of Global Secondary Index (GSI). The GSI can be created on the table at any time. Upto 5 are supported. So, indexing is no longer a blocking issue for many use cases.
You should also know that SimpleDB has size and performance limits. (10GB and say, 25 requests/sec)
Perhaps eventually, DynamoDB will replace SimpleDB in all but the most simple use cases.
3 key differences:
- SimpleDB creates index for "EVERY" field in a table.
- DynamoDB you have to set indexing fields before creating the database, and cannot be modified.
- SimpleDB pricing is based on machine hours and storage capacity
- DynamoDB charge money by capacity of Read/Writes records per seconds.
- SimpleDB requires manual partitioning if data storage exceeds 10GB.
- DynamoDB automatically distributes data under the hood, thus provides very high scalability.
In simple terms both data stores are NoSql.
The difference lies on scalability ( and a few other aspects , but scaling carries the biggest value in my opinion). SimpleDB is quite similar to MongoDB but has a bunch of limitations when it comes to scaling.
But DynamoDB lets you provision small , and scale up as many Provisioned Throughput you need. And scale down when it's not required. ( ie. During a promotion , Celebrity referral signups etc scenarios like this will have a time-based spike of hardware requirements )
I believe the major difference in Simple DB vs Dynamo DB are
- Predictable performance in terms of latency and be able to handle high volumes without compromising latency and throughput. DynamoDB achieves this by using the partition key
- Read, write optimizations which can be customized
- Better eventual consistency model because the use of consistent hashing algorithm