Why do hard drives fail?

Hard disk drive servers lighting up in sequence

Posted on Friday, February 23, 2018

As the hardest working part of an array, hard drives are prone to failure. Modern systems are designed to limit the impact of hard drive failure, but there are still steps that the data center manager can take to further reduce these instances – if they know what the cause of these failures are.

Manufacturing defects

In every batch of hard drives, faulty units are inevitable. Typically drives with manufacturing defects fail within the first 18 months of operation.

Although very difficult (impossible) to detect in advance, manufacturing defects are typically covered by the OEM’s warranty. When designing and deploying arrays, ensure that drives are selected from a selection of manufacturing batches to avoid a common defect causing the entire array to fail.

Excessive heat

The high temperatures generated by servers within the data center are another cause of routine drive failure. Over time, exposure to excessive heat will affect circuitry and components – becoming the most common cause of disk failure.

Cooling is critical to the on-going health of your systems, and should be a significant factor in data center design in future. By increasing the efficiency of data center cooling, your business can help reduce instances of hard drive failure.

Mechanical failure

After three to five years, the most common cause of hard drive failure is mechanical failure. Hard drives, including flash drives, are not designed to last forever – eventually moving parts will wear out.

Again, these wear rates are hard to calculate. The manufacturer will supply mean time to failure (MTF) statistics, which provide some guidance for predicting breakdown – and your maintenance routines should use these as a baseline for planning replacement cycles.

Preparing for the worst

Hard disk drive failure is inevitable – and your business needs to plan accordingly. As hardware ages, your choice of maintenance provider becomes increasingly important – particularly for post-warranty assets.

Make sure that your support provider has guaranteed access to OEM-approved spares, or you may run into significant problems when the unavoidable does finally happen and your arrays are under threat.

To learn more about spares for post warranty storage arrays and how CDS can help, please give us a call.

More Articles

Recycling symbol in the middle of a motherboard

Is computer recycling a scam?

“Free” computer recycling services may actually place your company’s green credentials at risk.

4 servers on blue background

What is the future of IoT? Watch APAC says Dell EMC.

Dell EMC reports that IoT revenue in the APAC region has topped $1bn for the first time.