The backup process checks disk space before it performs another backup to make sure the disk has double the disk space for a “full” backup, even when it does not need that much space. This is due to the “atomic” nature of the backup where the oldest backup snapshot will not be removed until the current backup snapshot is complete. If double the space is not available then this can force you delete the existing backup on the server and then initiate another backup. And if you are only storing 1 previous backup snapshot then this causes the only full backup to be deleted which means the current backup will have to be a full backup rather than optimized.

Starting in LMI 4.8.0 is an option called "Bypass disk checking" on the backup configuration page. Enabling this will skip the target disk free check which can allow a backup to  proceed and finish as long as there really is enough room to do an incremental update.

Prior to LMI 4.8.0 the alternative is to delete an existing backup (or more, depending on what's necessary) on the backup server and then initiate another backup.