hello
The problem I have is just tangentially related to Plesk but I would welcome any feedback from the community - hopefully someone have had similar problems and can provide solutions.
I manage a small amount of Plesk for Lunux on AWS EC2 instances and I use daily EBS snapshots as my main backup solution. I have to admit I have a serious problems with actually restoring from such backups - whether I attach restored volume to existing instance as root volume; or boot up a new instance from AMI created from EBS snapshot - these usually start as totally non responsive; show up as 0/2 checks passed and there is no SSH access or any other service available.
When attached as secondary drive to an instance, I can see and access the data on the volume from snapshot but obviously getting mySQL database dumps through file migration is tricky and time consuming.
My working theory is that, since root volumes are large (100+ GB in size), the snapshots are inconsistent due to the time it take to "snapshot" such volume. I personally restored successfully smaller instances with ca 30 GB in size from EBS backups with no problems, but the bigger the volume the higher the risk of getting unresponsive root volume in recovery.
I guess my question is.. has anyone had similar experiences and how did you resolve it? Is my little theory correct?
Is it possible to have Plesk setup with something like SYSTEM volume with all executables and DATA volume with web and email user data directories?
The problem I have is just tangentially related to Plesk but I would welcome any feedback from the community - hopefully someone have had similar problems and can provide solutions.
I manage a small amount of Plesk for Lunux on AWS EC2 instances and I use daily EBS snapshots as my main backup solution. I have to admit I have a serious problems with actually restoring from such backups - whether I attach restored volume to existing instance as root volume; or boot up a new instance from AMI created from EBS snapshot - these usually start as totally non responsive; show up as 0/2 checks passed and there is no SSH access or any other service available.
When attached as secondary drive to an instance, I can see and access the data on the volume from snapshot but obviously getting mySQL database dumps through file migration is tricky and time consuming.
My working theory is that, since root volumes are large (100+ GB in size), the snapshots are inconsistent due to the time it take to "snapshot" such volume. I personally restored successfully smaller instances with ca 30 GB in size from EBS backups with no problems, but the bigger the volume the higher the risk of getting unresponsive root volume in recovery.
I guess my question is.. has anyone had similar experiences and how did you resolve it? Is my little theory correct?
Is it possible to have Plesk setup with something like SYSTEM volume with all executables and DATA volume with web and email user data directories?