As workflows evolve and larger amounts of data are handled by even small businesses, the importance of data integrity becomes ever more critical. Hard drive capacities are now reaching 14TB in the standard 3.5" hard drive form factor, which means more chances for unrecoverable errors (UREs) during the RAID rebuild process, should you have to contend with a failing hard drive in your server or workstation's RAID array. During the RAID rebuild process, after inserting a new hard drive, the drive is put under high load for an extended period of time, which makes it all the more likely that these insidious UREs will assert themselves and lead to data corruption, or total data loss of the entire array.
So what can you do about it? Well, back up your arrays, of course. A backup is always the first line of defense against data loss. However, a second item to consider is whether your server or workstation's memory is error-correcting (ECC). ServeTheHome tells us more about what ECC RAM is and why it matters in the enterprise:
For desktop users, as we read above, ECC is not imperative and is not as widely supported as on server platforms (Intel Xeon and AMD Epyc, respectively). But even for your small or medium business, ECC RAM is crucial. Uncorrected on-the-fly errors are written to disk, which means the corrupt bits are stored, permanently, until they're accessed, leading to unpredictable and undesirable results for the customer and the business owner.
Any off-the-shelf tower or rackmount server from the likes of Dell, Lenovo, or HP should be configured with ECC RAM by default. If not, you most certainly want to configure it as such yourself, during the ordering process. Geek Housecalls can help you in auditing your servers for data integrity compliance, as well as help you configure and build your own servers for deployment.
Contact us today for peace of mind about your on-premise data!