Tech —

Big hard disks may be breaking the bathtub curve

1,240 10+TB hard disks installed, and not a single one has gone bad.

Big hard disks may be breaking the bathtub curve

Low-cost cloud backup and storage company Backblaze has published its latest set of hard disk reliability numbers for the second quarter of 2017. While the company has tended to stick with consumer-oriented hard disks, a good pricing deal has meant that it also now has several thousand enterprise-class disks, allowing for some large-scale comparisons to be drawn between the two kinds of storage. The company has also started to acquire larger disks with capacities of 10TB and 12TB.

The company is using two models of 8TB Seagate disk: one consumer, with a two-year warranty, and the other enterprise, with a five-year warranty. Last quarter, Backblaze noted some performance and power management advantages to the enterprise disks, but for the company's main use case, these were of somewhat marginal value. The performance does help with initial data migrations and ingest, but the performance benefit overall is limited due to the way Backblaze distributes data over so many spindles.

In aggregate, the company has now accumulated 3.7 million drive days for the consumer disks and 1.4 million for the enterprise ones. Over this usage, the annualized failure rates are 1.1 percent for the consumer disks and 1.2 percent for the enterprise ones. At least for now, then, the enterprise disks aren't doing anything to justify their longer warranty; their reliability is virtually identical. The focus now is on what happens to the consumer disks as they pass their two-year warranty period. Will they show the same reliability, or will deterioration become more apparent?

Traditionally, the expected pattern of hard disk failures has been the so-called "bathtub curve:" a spate of failures of new disks ("infant mortality"), as disks that were defective from the factory are shipped into the wild and rapidly stop working, followed by a long period of low failure rates, and then an uptick in failures as the disks pass their engineered lifetime.

The initial data from the 10TB and 12TB disks, however, has not shown that pattern. While the data so far is very limited, with 1,240 disks and 14,220 aggregate drive days accumulated so far, none of these disks (both Seagate models) have failed. The low level of usage means that the disks have been installed and formatted and not much beyond that, but true infant mortality—disks that immediately expire on their first use—hasn't become apparent. Of course, it's far too early to know if this trend will continue long-term, but for 1,240 disks to be bought and installed without a single one failing to spin up or having some other defect is nonetheless unusual.

Channel Ars Technica