Denis Gomes Franco
Regular Pleskian
Hey everyone.
Just wanted to share some recent and interesting findings. I'm using the Backup to Cloud Pro extension and saving a full server backup to OneDrive. As I was running low on free space, backups didn't run on one of my servers. "Low space" relatively speaking of course, because the server has ~26GB of free space.
So I gave this some thought as I didn't wanted to leave a lot of disk space wasted just to be able to run backups. Damnit, with 20 gigs I can host some more 40 domains...
And that's when I came across this: s3fs-fuse/s3fs-fuse, which can mount an S3 compatible storage service as a disk under Linux. And I decided to give it a go.
What I did was:
1. Activated a Object Storage instance in Vultr. They are just a tad cheaper ($5 for the first 250 GB, then $0.02 afterwards) than Amazon S3. You can use any other S3 compatible storage solution, though.
2. Followed the instructions in the Github page to mount the disk and set it to mount automatically on reboot.
3. Created a bucket and then a directory for each of my servers to store their respective backups.
4. Created a temporary directory for each of my servers as well.
5. Changed the location of backups as per Where are Plesk Backup Manager temporary files located?
6. Set Plesk to save backups in local storage instead of Onedrive
7. Disabled the "Start the backup only if your server has the sufficient amount of free disk space" option
Using a remote disk as a temporary directory might not give the best performance but it worked. The backup operation did not needed as much space in the local disk as before and could be completed just fine now. And since the object storage is connected to the server, I'll just save the finished backups there as well instead of using OneDrive.
And with this my first set of scheduled backups of the servers has been completed flawlessly.
Now I hope this tip can help someone out there
Just wanted to share some recent and interesting findings. I'm using the Backup to Cloud Pro extension and saving a full server backup to OneDrive. As I was running low on free space, backups didn't run on one of my servers. "Low space" relatively speaking of course, because the server has ~26GB of free space.
So I gave this some thought as I didn't wanted to leave a lot of disk space wasted just to be able to run backups. Damnit, with 20 gigs I can host some more 40 domains...
And that's when I came across this: s3fs-fuse/s3fs-fuse, which can mount an S3 compatible storage service as a disk under Linux. And I decided to give it a go.
What I did was:
1. Activated a Object Storage instance in Vultr. They are just a tad cheaper ($5 for the first 250 GB, then $0.02 afterwards) than Amazon S3. You can use any other S3 compatible storage solution, though.
2. Followed the instructions in the Github page to mount the disk and set it to mount automatically on reboot.
3. Created a bucket and then a directory for each of my servers to store their respective backups.
4. Created a temporary directory for each of my servers as well.
5. Changed the location of backups as per Where are Plesk Backup Manager temporary files located?
6. Set Plesk to save backups in local storage instead of Onedrive
7. Disabled the "Start the backup only if your server has the sufficient amount of free disk space" option
Using a remote disk as a temporary directory might not give the best performance but it worked. The backup operation did not needed as much space in the local disk as before and could be completed just fine now. And since the object storage is connected to the server, I'll just save the finished backups there as well instead of using OneDrive.
And with this my first set of scheduled backups of the servers has been completed flawlessly.
Now I hope this tip can help someone out there