There are two variants of system architecture for Magento. Peer configuration where all the components of the system, i.e. database, cache system and HTTP server, etc., are placed on a single server. Server cluster with functional load distribution and balancing.For the realization of these two variants, both dedicated servers and cloud technologies can be used.Let’s focus on the use of dedicated servers and cloud technologies for the building of online stores.A peculiarity of online store functioning is that its attendance has unstable “peak” nature. It means that during a promotion campaign the number of visitors rises considerably and then, as a rule, declines. Dedicated hostings: A performance increase of a peer system is an expensive move, as an increase in physical properties or purchase of a more powerful server takes place. When the load reduces you still continue paying for the resources that are not used.
Advantages: High speed of system response (TTFB) even with the rate of load more than average. Disadvantages: Substantial material and time costs on changes to configuration; Risks associated with the hardware failure. When your store outgrows a single-stack server Magento installation and you are in a need for a scalable cluster the cloud hosting offers a viable solution for your business. Database and caching system PaaS offering provides fault-tolerant instances and reduces your need to find qualified personal to manage them.Cloud infrastructure saves on server costs. You only pay for the computing power you use.
Cloud hostings for Magento.
- Highly available and distributed architecture
- Offers PaaS for databases and caching systems.
- Decreases your labour costs on supporting physical infrastructure
- Allows for easy and automated scaling of your environment
- Grows and shrinks automatically with changing user traffic patterns
- Allows for small incremental changes
- Saves your server budget
- Procurement of new servers takes minutes not days or weeks
The environment can easily scale up increasing the number of servers in the cluster to fulfil the temporary demand from users on your computing environment. An automated process can also shrink the environment removing unnecessary servers in off-peak hours.