A Server Management System that Adapts to Changing Demands
Based on preregistered settings, the number of servers automatically increases or decreases to lower costs, while maintaining stable service. The auto-scaling servers are withdrawn automatically after an event is disabled for maximum cost-efficiency.
- Flexible Capacity for
Unexpected Demand - It can often be difficult to predict required capacity for operating services, but there is no need to prepare an excessive number of servers to ensure high capacity. With Auto Scaling, the operating servers are constantly monitored to automatically connect to the server farm for increased demands and the server is automatically withdrawn after the high load condition is resolved.
- Efficient Server Operation for Services
that Experience Daily Fluctuations in User Traffic - For predictable user traffic services, the number of servers can be set to increase or decrease, according to the required time period. Efficient management of server operation costs is possible by focusing on increasing the number only at required times, rather than focusing on the worst, traffic.
Detailed Features
Create an Auto Scaling group to increase or decrease the number of servers.
Server status is monitored to guarantee the demanded capacity at all times
Auto Scale Products Based on Server Monitoring
Set thresholds for various monitoring metrics (CPU, memory, and traffic) that are linked to the monitoring service.
If a threshold is reached, the numbers of servers are automatically increased or decreased. This service is especially
useful for maintaining maximum flexibility during periods of unpredictable traffic, such as before or after the release
of a new game or service.
For example, engaging additional servers when CPU usage exceeds 50% on the monitoring data.
Operate Servers According to Traffic at Specified Times of Day
This is a method of increasing servers only during specific hours when traffic gets congested, depending on the characteristics
of the customer’s service. This service may be utilized when there are great differences between traffic levels
at various times or when the time to execute the batch processing task is set.
For example, you can add five more servers between 5PM and 11PM, when the number of users is high, and return to
the normal number of servers after 11PM.
Automatic Replacement of Malfunctioning Servers
The servers in the Auto Scaling group are checked periodically. If a server malfunctions or becomes unresponsive, that server is automatically replaced with a new server with the same configuration to maintain the capacity and stability of your service.
Integrated Monitoring of Auto Scaling Groups
All servers in the Auto Scaling group can be linked with monitoring services, which can automatically notify you of threshold limits. Server resources (CPU Usage, Load Average, Memory Usage, etc.) within a group are monitored according to your settings and can be viewed on the dashboard.
Designate the Servers to Return for Scale-in Using the API Query Method
Generally, when returning servers using the Auto Scaling function, the servers are returned from the earliest created order. If you want to designate the servers to be returned first, you can do so using the API Query method. You may utilize this feature to return servers with lower usage rates first.
Related Services
This service is free of charge.
Although no fees are charged for using Auto Scaling, fees on the servers created using this feature will be charged under the server usage fee.
Was this page helpful?