Surely, there’s nothing more important to your company than the data your users access and store every day. For this reason alone, system admins would have evaluated and deployed a BDR solution in place, planned for the schedule of backups, configured them based on organizational RTO and RPO considerations and then spend a good amount of time monitoring the backups for alerts and notifications. So far, so good, But…
How do you know for sure that the data you backup can be restored without a glitch ?
Where is the assurance on the integrity of the backup, whether it is on-premise or in the cloud ?
Cone of Constraints
Data integrity is intricately intertwined with the technical sophistication with which the BDR product implements its error detection processes and incorporates fail-proof checks for corruption and subsequent steps for correction. However, all solutions are not equal on this count of proactively identifying data corruption preventing corrupted data from being transferred during the backup process.
Lets be clear that occurrence of data loss is inevitable due to the probability of either underlying storage hardware failures hosting backups or transfer of corrupted data blocks. There are huge trade-offs associated with competing priorities – performance, fault tolerance and time-to-market. Due diligence, therefore, has to be the prerogative of the system admin in carefully picking the right BDR solution that scores masterfully on all counts and drastically reduces the possibility of any data corruption.
Data Loss Circumvention
Traditionally, storage protection encompasses two main methods of protection, namely, Replication and RAID.
RAID (Redundant Array of Inexpensive Disks) is a storage virtualization scheme for dividing and duplicating data across multiple physical drives. There are a number of schemes ranging from RAID 0 to RAID 6. Raid 1 for instance provides data redundancy on a mirrored set of disks and RAID 6 incorporates double distributed parity to ensure no data loss for upto two failed drives.
Backup software vendors opt for replication, as a “backup for backups”, where one or more copies of the primary backup bundle is transferred and stored either locally or offsite for high availability and disaster recovery.
Enter Erasure Coding
A better alternative to Replication, in terms of costs, was made possible by the advent of a specific technology – Erasure Coding. It is similar to RAID technology, which generally adds parity as overhead. The difference being, while RAID works on the hardware level, erasure coding applies it to an object level. Basically, erasure code provides redundancy devoid of the overhead of replication. It divides any data stream (data blocks, file chunks) into “X” fragments and recode them into “X+Y” fragments. The key property of erasure codes is that the original object can be reconstructed from ANY of the X fragments. Hence integrity checks based on erasure coding tend to be more robust.
Vembu BDR v6.1 – Coming soon
Our patent pending VembuHIVE technology which is the foundation of the Vembu BDR v6.1 platform, possess inbuilt integrity checks using sophisticated erasure coding techniques that could detect and correct multiple errors at several breakpoints of failure- backup file level, disk image-level, chunk- level, backup-level, repository-level and client -level. This way, you can be assured that the quality of our backups have met and surpassed stringent expectations of data integrity and durability.
There are many exciting new developments with our upcoming version 6.1. Just go ahead and download the recent whitepaper to learn more about some of them.