To deal with sudden increases in website traffic and storage problems, the company Scalr provides a platform that allows to build a scalable infrastructure and manage capacity issues.

Using Scalable Infrastructure Helps Handling Unexpected Website Issues


For start-up companies, growth in business is hard to manage without the proper infrastructure in place. Scalr, an open source cloud management application, monitors the activity of a website and offers automation to clients' operations to handle website glitches. If website loads are to increase, the application adds server capacity and when loads decrease, less machines are subsequently used. This feature, which is known as auto-scaling, is a sysadmin robot that routinely checks server activity. If a website wants to launch a new product and is expecting thousands of users to sign up, Scalr will provision for the necessary resources to ensure smooth activity. It scales a website's database, application servers, and configures load balancing servers.

Auto-Scaling a Distinguishable Feature

Other cloud-computing projects such as Rackspace and Eucalyptus offer cloud solutions to companies; however, the auto-scaling feature is one that remains unique to Scalr. Scaling can be based on CPU, RAM, or database issues. If a database gets overloaded, Scalr will automatically decide to add capacity to the database depending on disk I/O. Without such technology, a monitor check would not be able to succeed if a server dies.  As a result, the sysadmin would start fixing the problem – which could take anywhere from a few minutes to several hours. However, with the auto-scaling that Scalr provides, a replace command is sent when a server fails and then provisions for a new one from the cloud. As a result, downtime is reduced as well as recovery cost.

Database Failure No Longer an Obstacle

In order to prevent a potential server crash, Scalr backups data at regular intervals and utilizes Amazon EBS for database storage. The Elastic Block Store enables users to rent virtual computers allowing them to run their own computer applications. This virtual machine can terminate allow one to elastically store data according to the company's needs. If there is a master database replicating data to two slaves and the master database crashes, Scalr will then send the promote command. This command will promote the slave to master reconfiguration replication and then will then communicate with Amazon EBS to configure the slave to act as a new master.


By Marcus Burke