The storage fills up with things nobody will ever restore: WordPress plugin caches, logs, the site’s own backups kept inside the account. How much do you save by leaving out what doesn’t need a backup (caches, logs, old archives and dumps, plugin backups), and where does that list come from?
Why it matters to whoever runs the server
- 35% less read is 35% less work for the server’s disk during the backup; 31% less written is space left on the storage, night after night.
- Email stays off the list: what you save is what nobody restores, not customer data. And the admin edits the list per job.
See the data as a table
| No exclusions | With the default list | |
|---|---|---|
| Files | 157,827 | 142,162 (−9.9%) |
| Data read | 6,269 MB | 4,064 MB (−35.2%) |
| Written to storage | 2,955 MB | 2,052 MB (−30.6%) |
Method
Where the list comes from: 29 patterns recommended by JetBackup, 16 additions of ours (Magento folders, LiteSpeed cache, trash, panel sockets) and 1 fixed JetBackup exclusion for CloudLinux. That’s 46 patterns, editable by the administrator in each job, as in JetBackup.
One deliberate difference: JetBackup removes /mail from the account folder because it copies email separately. Salvo copies email together with the account, so email is not on the list.
A production cPanel server. The same 6 WordPress accounts of about 1 GB from Lab 09, 3 at a time, first backup, measured without exclusions and with the default list on the same day. The comparison holds for bytes; time isn’t comparable, because without exclusions the accounts ran in different groups. With the list, the 6 accounts took 49 s.
What we decided
The list comes on in every new cPanel account job; the administrator edits it per job: remove, add or reset to the default. Jobs for plain folders (Linux without a panel) get nothing by default.