Common problems

Available in Classic and VPC

You may encounter the following issues while using the Auto Scaling service. Check the causes and solutions for each issue and take an action accordingly:

LAUNCH process stops and server creation error

The LAUNCH process is paused and the server is not created.

Cause

Administrative suspension may be the cause of the suspension of the LAUNCH process. Administrative suspension is a situation in which the use of a Launch Configuration is suspended because it is determined to be defective. If the management agent inside the OS fails to start (wait for up to 60 minutes) while rebooting to apply settings during the initial boot of the server, rebooting fails. If the same operation fails even after attempting up to 5 times, it is considered a Launch Configuration defect.

Solution

Under the administrative suspension situation, you need to determine the cause of repeated server creation failures, take action, and then resume from the process management settings.
To resume from the Process Management settings, follow these steps:

1. From the NAVER Cloud Platform console, click > Services > Compute > Auto Scaling > Auto Scaling Group, in order.
2. Select an Auto Scaling Group and click [Settings] button.
3. Click the [Process management] tab.
4. Select the check box for the LAUNCH items and click [Restart].

Scale-out error

• The number of servers should increase, but new servers are not created.
• Scale-out has not been performed.

Cause

• If the server image selected when creating Launch Configuration has been faded out, scale-out might not occur.
• Each server type has its limitations. Due to server type limitations, servers may not be created correctly.
• If you exceed the number of servers that can be created, an error may occur.
• If the subnet used when creating a server is not for "General", the server may not be created.

Solution

• Check whether the server image is faded out when creating Launch Configuration. Scale-out using fade-out server images is not supported. Set up Scale-out using a different server image.
• Increase your server creation limit or terminate an existing server.
• Check the limits for the server type you are to create.
• If the subnet used when creating a server is not for "General", the server may not be created.

Server automatic termination

• Servers keep being terminated automatically.
• Servers are repeatedly being terminated and created when the desired capacity is set to be 1.
• Servers are randomly terminated regardless of the set schedule.

Cause

• If the created server fails a health check with the Load Balancer and the server is not successfully created, the server may be automatically terminated.
• If the cooldown time is too short, the server may be automatically terminated.

Solution

Adjust cooldown time
Adjust the cooldown time. If the cooldown time is too short, the health check request may proceed when the server has not finished its booting, causing it to be recognized as an unhealthy server and terminated. To edit the cooldown time, see edit Auto Scaling Group information.

Check health check grace period
See the health check grace period. If the health check grace period is too short, the health check may proceed before it completes binding to the Load Balancer, and the health check may fail and the server to be terminated. It's set as 300 as default, so check the time your server boots and adjust accordingly. To edit the health check grade period, see edit Auto Scaling Group information.

See reason for health check failure
Check your server for the reason the health check has failed.
The following describes how to prevent the server from being terminated to find out why a health check failed:

1. Click Auto Scaling Group > [Settings] > [Process management].
2. Select and click the check box for the Terminate items and click [Pause].
3. Check the process management window and click [OK].
4. Access the server and check.