Cloud Elasticity

Use Cloud elasticity to cope with demand variability


If your current systems are struggling with everyday demands, or meeting those demands but your costs are sky high - then our blog on Right-sizing would be the best place to start.

If your current systems cope well with everyday demands, but fail miserably when hit with sudden and unpredicted surges in demand, then this blog on Cloud Elasticity is perfect for you. Read on.


What is Elasticity?


Elasticity is the property of something able to encompass much variety and change, be flexible and adaptable but also to resume its normal size and shape spontaneously after being stretched. A property not normally associated with your average desk side or data centre server, database, or NAS (Network Attached Storage) System.


How is Cloud Elastic?


Correctly designed, Cloud environments can be configured to provision and deprovision additional resources in an automated manner, such that at any point in time the available resources match the demand as closely as possible. The obvious benefit is that you are only paying for the resources you consume - you don’t have to provision your hardware up to the maximum demand you expect, and have it stand largely idle the rest of the time.


Where is Elasticity useful?


PolarSeven has helped many organisations deal with demand variability. Examples run across eCommerce, Streaming, and Education - where many candidates may access online training, or submit online assessments at the same time.


For one customer in Education, PolarSeven developed a serverless data ingestion pipeline. The challenge was to commit a large volume of post-course analytics, potentially from up to 100,000 users to a database in a very short space of time. The PolarSeven solution involved a serverless pipeline using Lambda and a nonSQL database that could capture the analytics data in a serverless queue, and trickle feed it to the SQL database at a pace that would not overload it.


For Clickview, an Australian-grown success story, often dubbed the ‘Netflix of Education’, PolarSeven deployed an elastic container solution with auto-scaling. The system was able to manage the varying demands of both compute and storage so that users’ online experience was not impaired when large audiences were accessing the same content at the same time.


In another case, where demand could increase 1000 fold in a matter of minutes, PolarSeven suggested a serverless solution. That was the challenge that Reading Cinemas, a theatre chain with 60 cinemas, faced when blockbuster movies went on sale. The solution, as well as migrating from on-premise to Cloud, was to re-architect the system into two parts: the relatively static movie schedule component, and the very dynamic eCommerce component built on serverless components.


How is Elasticity Achieved?


There are many approaches to elasticity whether you need elasticity in Compute, Database or Storage - or a combination of these.


For compute, one approach is to containerise your applications. A container works like a little self-contained application rocket: when you need more whoomph, the system automatically spawns a new container, drops in the payload and fires it up! Another approach is to go serverless. AWS Lambda, for example, is a serverless compute service that lets you run virtually any type of code with zero admin. Lambda automatically allocates computing power and runs the code in response to an incoming event.


If your challenge is a web application retrieving information from disk-based databases, your solution might be Amazon ElastiCache, a fully managed in-memory data store and cache service which improves performance overall, as well as coping with demand variability.


Bottom Line


If you are currently still on-premise, or have fixed capacity in a data centre and are challenged by sudden and/or steep increases in demand, contact us and we can help you explore the benefits of Cloud’s elasticity. Migrating to the Cloud is easier than you think.


If you are already on Cloud, and your platform does not respond well to demand variability, contact us and we can discuss how improving your elasticity will both help you better meet demand - keeping your customers happy - and save you money.

40 views

Related Posts

See All