• Please be aware: Kaspersky Anti-Virus has been deprecated
    With the upgrade to Plesk Obsidian 18.0.64, "Kaspersky Anti-Virus for Servers" will be automatically removed from the servers it is installed on. We recommend that you migrate to Sophos Anti-Virus for Servers.
  • The Horde webmail has been deprecated. Its complete removal is scheduled for April 2025. For details and recommended actions, see the Feature and Deprecation Plan.
  • We’re working on enhancing the Monitoring feature in Plesk, and we could really use your expertise! If you’re open to sharing your experiences with server and website monitoring or providing feedback, we’d love to have a one-hour online meeting with you.

Question How to limit the cpu load on restore tasks?

Bitpalast

Plesk addicted!
Plesk Guru
On a normal Linux server, CentOS 7.2, when restoring data from a backup in an external FTP storage, the cpu load temporarily reaches peaks that are causing severe delays on DNS resolution, httpd and PHP-FPM operations. Some delays are long enough to reach time out thresholds.

Can the cpu load that a restore process is using be limited by a setting, e.g. in panel.ini?
 
On a normal Linux server, CentOS 7.2, when restoring data from a backup in an external FTP storage, the cpu load temporarily reaches peaks that are causing severe delays on DNS resolution, httpd and PHP-FPM operations. Some delays are long enough to reach time out thresholds.

Can the cpu load that a restore process is using be limited by a setting, e.g. in panel.ini?
Hello Peter, have you tried to run them with low priority or to reduce the simultaneous backup processes ?
I have never noticed similar issue with this settings previously.

Reference :
High CPU wait, I/O and Load Average during backup
Global Backup Settings
 
It seems that developers have thought about the backup process, but not about the restore process. Because restore always consumes as much CPU power as it can get. Backup is not an issue, but restore is.
 
It seems that developers have thought about the backup process, but not about the restore process. Because restore always consumes as much CPU power as it can get. Backup is not an issue, but restore is.

You are right, but it may be interesting to identify what process use so much CPU, because restore process should use less resources than backup process in my mind. I will run restore from ftp to see if there is the same issue.
 
Yes, restoring from an ftp account is a very big problem. But I think that is not a mistake in itself.

What happens with a restore: The tar archives are copied from the FTP server to the local hard drive, unpacked and then copied from there to the target folder.

This process consumes not only a lot of CPU time, but also a lot of hard drives io's.
 
So probably we should formulate a user voice feature request for this? Limiting the cpu load on restore transactions?
 
So probably we should formulate a user voice feature request for this? Limiting the cpu load on restore transactions?

Sure, CPU and IO load/Priority!

Another way, backup to external SSH/SFTP storage without packing. The restoration could then run for example by rsync.

Or even better, grabbing the data and unpacking would run by command on a remote system.

Thats only a brain storm :D What could be better?
 
Last edited:
The backup restore process starts the deployer, which initiates the untar process. As you know, untar does not able to use multiprocessor CPUs and uses only one processor. Thus, the problem is either in the limit on the number of processors for your VPS, or in the incorrect operation of the disk subsystem, in particular the efficiency of I/O.
 
Back
Top