Amazon has realised Elastic Block Storage recently and many other companies are following in their footsteps such as N2WS with a similar incentive and aim.
The general process of EBS is not too complicated and anyone can benefit from the system. Essentially a volume is formed from a storage amount of 1GB and can increase to 1TB – the storage can then be mounted to a device of your liking and that can then be used as a storage facility externally but based online.
The storage unit can be left at any moment and stay secure and safe. It can also move to another area of the system to use anytime. The volume can be snapshotted at any time to S3 and can also be restored to form a new volume from the snapshot.
EBS is a reliable system that has volumes built in already for ease and will not fail anyone that occurs any problems along the way. The S3 storage multiples data to numerous zones that are available and the EBS volumes are then in one area alone. The snapshot backups are important is this sense for the benefit of the future to protect the storage.
EBS is reliable in its system and providing volumes; the volumes are designed to aid those who require access to them, there are multiple servers available in one zone that will ultimately avoid any problems with losing masses of data from one asset. The longevity of the volume is depending on the volume size and also the percentage that the data has gathered since the previous snapshot.
The volumes work generally with about 20 GB which contains the modified data that was last snapshotted, from this we can expect an annual failure rate from 0.1% to 0.5%, this could result in a major loss in volume, this is also similar to hard disk that loss data quickly and are not as reliable.
The RAID storage system is also reliable and increases chances of data security, by having two EBS volumes in one software and working on much more than a single volume. It is possible to build a decent snapshot platform that is consistent and on-going with the scripts that specifically aid to recover data and form new volumes.
The overall performance of the volume system is based on pre-realised EBS volumes estimations that vary on manufacture. The tests have shown EBS volumes are effective and work well with volumes being created.
The EBS volumes essentially work on network attached to various storage platform, each time biting network bandwidth, the maximum is 1GB ps for the purpose of the network itself.
There have been numbers over 70mb a second with sysbench which can peak to higher levels and the performance is better overall. The performance has shown to exceed its expectations from other file-based cloud systems and can handle over four local drives in various instances.
The system creates the possibility of I/O transaction rate with mounting the various volumes at one time and then moving the file systems – this is also possible for streaming but not as beneficial.
With EBS it is possible to increase the I/O transaction rate further by mounting multiple EBS volumes on one instance and striping file systems across them.
For streaming performance, this doesn’t seem worthwhile, as the limit of the available instance network bandwidth is already reached with one volume, but it can increase the performance of random workloads as more heads can be seeking at a time.
Oct 07, 2017 0