Cloud Basics - Scalability and Elasticity

High Scalability / Elasticity

Your ability to increase your capacity based on the increasing demand of traffic, memory and computing power. Cloud offer high scalability which is otherwise not an easy task to achieve using manual approach. The enterprises need to achieve scalability as their number of end user increase, so do their data hits and network traffic. In order to meet the increased requirement for the system, it needs to be scaled. There are two ways of scaling:


Vertical Scaling/ Scaling Up 

It means upgrade to a bigger server or powerful machines. If on-premises, this has to be done manually. you need to buy expensive machines/hardware and software to upgrade the server to host the software components. Cloud manages this automatically and you just have to pay accordingly, you don't have to buy any physical infrastructure, just scale up manually based on your specified conditions (e.g; more than 85% of CPU consumption or 70% of memory consumption etc). Cloud portal has scaling page to accomplish this. If you want scaling to be performed automatically, that can also be managed by cloud.

Horizonal Scaling/ Scaling Out 

Another way of adding more servers of the same size, this means adding more instances of the servers to serve the requests rather than upgrading to the faster machines. Horizontal scaling is usually the cost effective solution but depends on the the system you are hosting as well.

You can achieve this by manually specifying the number of instances and max and min no of instances in cloud. Or leave it to cloud to manage it for you

Click to enlarge the picture


Scaling Down/Scaling In


As the name indicates this can be achieved or managed by cloud automatically depending upon the traffic. and hence your get the optimised and scaled solutions based on your requirements. In On-premises this has to be done manually and requires lots of effort and time. also the systems has to stay down during this process. cloud manages scaling invisibility and end users don't experience any lag or system down time. This is because of the high availability feature mentioned the article High Availability.

High Elasticity 

Your ability to automatically increase or decrease your capacity based on the current demand of traffic, memory and computing power 

Azure VM Scale Sets 

The horizontal scaling is achieved through the underlying infrastructure of VM scale sets. This is automatically increasing or decreasing in response to demand or a defined schedule based on the organization's preference. 
 I will discuss this topic in detail in subsequent posts.

SQL Server Stretch Database 

Dynamically stretch warm and cold transactional data from Microsoft  SQL Server 2016 to Microsoft Azure 

Vertical Scaling is generally hard for traditional architecture so you'll usually 
only see horizontal scaling described with Elasticity. 

For more information about infrastructure used in scalability read about Load Balancer and the High Availability 

Comments

Post a Comment

Popular posts from this blog

How to prepare for AZ-204 Developing for Microsoft Azure Certification Exam

Image
  Here is the list of learning path to prepare for AZ-204 Developing for Microsoft Azure exam from my personal experience. First foremost, subscribe to free Azure plan for portal, this will give you free access for some services for 30 days and for others 12 months access. you can browse through Azure portal and see its services. Exam expects 2 years experience of MS Azure, but if you just starting and want to go ahead with learning, you need to start from somewhere and this should be your first step. Take a Udemy course if you can buy one, go through all modules once. For absolute beginner, this is good step as the good instructor will quickly guide you through all the steps/services and give you a general idea of the content. However these courses are never comprehensive. you should go through them again and again if you are beginner or don’t have enough knowledge about a certain topic. you might want to go thro...
Read more

Cloud Basics - Availability and Redundancy

 High Availability  Your ability for your service to remain available by ensuring there is  *no single point Of failure and/or ensure a certain level of performance. High availability is achieved by replicating the images in multiple zones or data centres. So in case of any geo-disaster the alternative systems serve the users, hence it is called highly available. There are three types of redundancies: Local Redundancy Zone Redundancy Global/Geo Redundancy Local redundancy The default behaviour is locally replicating the data/resource images in multiple copies in the same datacentre. The resources by default have three copies. so if one copy fail to serve the other two are still available. This mode has less availability comparing to the other two but still it provides some piece of mind. Zones Redundancy  In this mode three copies of the resources are kept in three different zones or data centres in the same location/city. For example, UK West is a location which may...
Read more