It is possible that this behavior is due to the way Rapid Recovery gathers consistent snapshots for vFAT partitions. Due to a lack of native Linux function to freeze vFAT partitions, we set drop_caches to 3 which only affects clean cache and not dirty cache, during the snapshot process in order to cleanly gather the necessary disk information. During this process we do not call the sync operation however, which would write cached information to disk. This has the potential drawback of causing some load issues as mentioned.
In order to workaround this issue, once you have a successful base image backup of the /boot/efi partition, it is recommended to separate this partition to a different backup schedule, preferably once a night during off hours or you can remove this partition from the backup schedule all together. This will stop /proc/sys/vm/drop_caches from changing to "3".
NOTE: If you choose to remove this partition from the backup schedule after getting the base image, you may want to manually backup this partition after a kernel upgrades to get all of the latest changes to the /boot/efi partition, in case you need to do a BMR restore.
© 2019 Quest Software Inc. ALL RIGHTS RESERVED. Feedback Términos de uso Privacidad