Differences between S3 and EBS

Amazon Elastic Block Storage (Amazon EBS) is a new type of storage designed specifically for Amazon EC2 instances. Amazon EBS allows you to create volumes that can be mounted as devices by EC2 instances. Amazon EBS volumes behave as if they were raw unformatted external hard drives and can be formatted using a file system such as ext3 (Linux) or NTFS (Windows) and mounted on an EC2 instance; files are accessed through the file system . They have user supplied device names and provide a block device interface.

For a 20 GB volume, Amazon estimates an annual failure rate for EBS volumes from 1-in-200 to 1-in-1000.  The failure rate increases as the size of the volume increases.  Therefore you either need to keep an up-to-date snapshot on S3, or have a backup of the contents somewhere else that you can restore quickly enough to meet your needs in the event of a failure.  

EBS accounts can have a  maximum of 20 volumes unless a higher limit is requested from Amazon. The maximum size of a volume is 1 TB and the storage on a volume is limited to the provisioned size and cannot be changed. EBS volumes can only be accessed from an EC2 instance in the same availability zone whereas snapshots on S3 can be accessed from any availability zone. 

Amazon S3 provides a simple web services interface that can be used to store and retrieve any amount of data, at any time, from anywhere on the web. It gives any developer access to the same highly scalable, reliable, fast, inexpensive data storage infrastructure that Amazon uses to run its own global network of web sites. The service aims to maximize benefits of scale and to pass those benefits on to developers. S3 needs software to be able to read and write files but is hugely scalable, stores 6 copies of data for HA and redundancy, and is rumoured to be written in Erlang and is hugely scalable.

S3 accounts can have a maximum of 100 buckets, each with unlimited storage and an unlimited number of files. The maximum size of a single file is 5 GB.

S3 is subject to “eventual consistency”  which means that there may be a delay in writes appearing in the system whereas EBS has no consistency delays. Also EBS can only be accessed by one machine at a time whereas snapshots on S3 can be shared,  

In terms of performance S3 has the higher latency and also has higher variation in latency. S3 write latency can also be higher than read latency . EBS on the other hand has lower latency with less variation. It also has writeback caching for very low write latency. However be aware that writeback caching and out-of-order flushing could result in either an unpredictable file system or a database corruption

In terms of throughput S3 has  maximum  throughput (single threaded ) of approximately 20 MB/s or 25 MB/s  for multithreaded. This is on a small instance. This rises to 50 MB/s on the large and  extra large instances. EBS has a maximum  throughput limited by the network, This isapproximately 25 MB/s on a small instance and 50 MB/s on large instances,  and 100 MB/s on  extra large instances. As both S3 and EBS are shared resources they are subject to slowdown under heavy load.

For file listing S3 is slow and search is by prefix only  whereas EBS has fast directory listing and searching. S3 is performance optimized by using multiple buckets. The write performance is  optimized by writing keys in sorted order . EBS single volume performance is similar to a disk drive with writeback caching.

There is an alternative to EBS for EC2 and that is PersistentFS. With PersistentFS you mount a drive and use it like any other, but, and here is the crunch, the storage for the device is actually realized in many little chunks in an S3 storage bucket. PersistentFS is a closed-source product based on the FUSE approach.

S3 costs 15 cents per GB for storage actually used and 1 cent per 10,000 GETs, and 1 cent per 1,000 PUTs. EBS costs 10 cents per GB provisioned and 1 cent per 100,000 I/O’s. For a pricing of PersistentFS and how this compares to both S3 and EBS I suggest you read this post on the Amazon forums which as posted by the PersistentFS team.

Be Sociable, Share!