What is Auto Scaling?
Auto-scaling is an approach that automatically scales up or scales down the number of computing resources, like CPU or RAM that are being allocated to your active application at any given point of time, based on its needs. eNlight Cloud has built-in intelligence to detect increase in resource requirements on a server in real-time, which it adds immediately; and most importantly, without any downtime. ALL with NO reboot or manual intervention needed! The server infrastructure setup on eNlight Cloud platform allows resources on Virtual Machines to scale-up or down depending upon the demand. The intelligent tool fetches the necessary resources from a large pool of free resources and allocates those to VMs during predicted as well as unpredicted bursts or spikes, allowing the company or personal website(s) to achieve optimum scalability and perform at optimum levels. Auto scaling responds automatically and ensures that applications continue running smoothly, without any interruptions, no matter how erratic the resource usage patterns maybe.
eNlight Cloud scales up and down your resources, i.e. the CPU and RAM as a percentage of resource consumption patterns. The scaling up is a steady but quick and automatic percentage of the pattern as required by your application, and so is the scaling down. It scales down and de-allocates resources to VMs when the demand falls during off peak hours. This ensures that you are only paying for what you use. eNlight Auto Scaling changes the economics of computing by allowing you to pay only for computing resources you actually use.
Due to thick provisions of disk on storage "Disk Scaling" is not supported in eNlight however eNlight comes with an extraordinary option called "Elastic Storage" which allows you to add multiple disks on the fly and also attach / detach them between multiple VMs without losing any data. eNlight is the best option for Disaster Recovery Hosting due to this unique feature.
Features of Auto Scaling
- Scale up computing resources seamlessly and automatically as per the demand.
- Scale down automatically when demand lulls to minimize costs.
- Define max limit for CPU, RAM & Bandwidth scaling to keep costs under check.
- Auto Scaling can be enabled or disabled on individual VMs.
- Facilitates optimum resource usage without compromising on performance.
- Available as a standard feature with eNlight Cloud and carries no additional fees.
Performance Models with eNlight Cloud
Cache memory is an extension of Random Access Memory (RAM) that a computer accesses more quickly than a regular Disk. The computer, when processing data, first looks in the cache memory that stores information from a previous reading of data and moves on to the much larger memory only if it is not able to retrieve it from the Cache. At times, the cache memory can grow very large unnecessarily reserving most of the RAM to store cache for performance; this is a normal scenario with Linux OS, to make the execution of a command that has been previously executed, easier. The amount of cache memory on a RAM can have huge implications on the performance of a machine. To make sure that you truly enjoy the freedom that eNlight brings to you, we offer two performance models depending on your preference :
- Economy based Model : With the economy based model, eNlight Cloud Storage devices are programmed to set a maximum limit to store cache on RAM, to stunt the growth of cache. Thus there is limited cache on RAM and hence the RAM on the server does not grow larger. The economy model thus focuses on limiting the cache to certain size and read data from the storage. As the name suggests the cost of a VM is kept low due to limited resource, compromising performance by accessing data from the disk and not from the RAM. This model suits those customers who has small size of database and doesn't require high performance for their dynamic applications plus to keep the VM invoices low.
- Performance based Model : As a part of the Performance based Model, eNlight ensure a considerably large cache size for the storage devices. This is done to increase performance and to avoid repetition of fetching the same data from the drive several times, so that the future requests can be processed immediately. This model warrants more investment, since the RAM size needs to be heftier, but it leads to efficient and reduced response time and comes highly recommended. This model best suits those customers who have heavy traffic on their application and the size of the dynamic data is also very large.
There are few scaling limitations depending upon the OS type you choose
- Linux Virtual Servers have CPU scaling capacity ranging from 1 CPU to 24 CPUs.
- Linux Virtual Servers have RAM scaling capacity ranging from 512 MB to 64 GB. (security restriction upon 16 GB)
- Windows Virtual Servers have RAM scaling capacity ranging from 1 GB to 4 GB.
- Windows Virtual Servers come with default 4 CPUs.