TeraCLOUD is built on Solaris ZFS storage arrays in a RAIDZ-3 (triple parity) configuration. Emphasizing safety over speed, this robust arrangement can withstand up to three simultaneous disk failures without suffering data loss.
In addition to protecting against physical drive corruption, data inconsistencies at read time are automatically repaired, avoiding silent data corruption (bit rot).
About ZFS snapshots
TeraCLOUD maintains a ZFS dataset for each user and regularly saves snapshots for each dataset in case of logical data corruption.
Snapshots are a ZFS feature that preserve the filesystem state at time they are created.
TeraCLOUD takes snapshots of account data every three hours, saving the most recent seven snapshots (one day's worth) and daily, saving the most recent 14 snapshots (two weeks' worth).
|Snapshots are created instantly and take up no space when created, but they begin to consume space as the filesystem contents change and data that no longer exist in the current state are preserved in the snapshots. Capacity used by snapshots is included in your total used TeraCLOUD capacity. If your available capacity runs low, the system will automatically delete your snapshots, but this does not affect the current state of your account data.|
How snapshots work:
In this way, data is protected from operator error, but snapshots do not constitute an offsite backup, so they cannot protect against a catastrophic scenario resulting in the total failure of the storage infrastructure.
Therefore, we recommend keeping a backup of your TeraCLOUD data in another location to ensure its safety.
Updated November 20, 2020