It seems that there are issues with non-Amazon Linux AMIs. When a new volume is declared and attached to an instance using a CloudFormation template, we have something like this:
We found this way is only valid for Amazon Linux AMIs, the image is in charge of mounting the partition to the instance under the hood. But when we work with CentOS, Windows or Ubuntu AMIs that process is not being made, so we have a default EBS volume per instance and unused custom EBS volumes. The error was to set up the device name as /dev/xvda for all systems, and the solution is about changing it from xvda to sda1.
Hello folks,
It seems that there are issues with non-Amazon Linux AMIs. When a new volume is declared and attached to an instance using a CloudFormation template, we have something like this:
We found this way is only valid for Amazon Linux AMIs, the image is in charge of mounting the partition to the instance under the hood. But when we work with CentOS, Windows or Ubuntu AMIs that process is not being made, so we have a default EBS volume per instance and unused custom EBS volumes. The error was to set up the device name as
/dev/xvda
for all systems, and the solution is about changing it fromxvda
tosda1
.Cheers