Since our home and business data continually grows, let’s remember key things to keep HDD/SSD drives healthy for the longest possible time. In this article, you will find practical suggestions on how to keep HDD-based archives safe and powered down, as well as the most important parameters helping to check whether a drive is healthy.
Level 1 - Physical Conditions
First of all, you need to put your HDD or SSD in conditions which reduce negative external impact, such as vibration or overheating. Here are points to check:
#1 Clean the Drive and Its Environment as Often as It Gets Dusted
Since computers pass through themselves a significant airflow with dust, this dust covers all internal surfaces, including hard drives. Thermal exchange thus became worse and your drive will work at higher operational temperatures, shortening its lifespan.
It is also a good idea to use fine-meshed thin membrane covers on all the vent holes of your desktop PC – it helps to leave dust mostly outside.
#2 Avoid Static Accumulation
A Static field can accumulate on the surface, leading to static discharges. They actually can’t corrupt your drive's internal components, but remember about the drive’s controllers placed outside the “shell”. The static electricity is more dangerous at winter time when the air inside the room gets drier.
It is very easy to break some chip on the controller, so you just cannot access the data without special hardware or third-party assistance. To avoid such an inconvenience, always keep the PC case closed and use mobile racks for external drives.
If storing several HDDs\SSDs on the shelf - ensure they are placed inside static free individual boxes.
#3 If Using Shelf-Based Archives, Periodically Check Those Drives for Logical Errors
When using the HDD as a long-term archive, remember about its main component – the magnetic surface. There is no official “lifetime of the magnetic field”, but you most likely will have no problem with data reading after a dozen years of being offline. For example, there are sources claiming that an HDD can keep the data for 60 years!
Besides the fact that modern HDDs can preserve data for many years while being powered off, implement an additional practice to keep the data loss risk as low as possible: just connect those drives to the PC and run a simple full disk check (chkdsk command in Windows OS).
Electricity will refresh the drive magnetic field when the disk checking tool will find and correct possible logical errors, preventing them from converting to “bad blocks”. But remember that you need to check the entire drive (full scan), not quickly scan its metadata. It is a relatively long process, but obviously, data safety is more important.
But what about SSD?
Of course, SSD technology does not use magnetic fields – they are built using NAND flash chips. Since flash memory cells can also lose charge while being offline, the SSD has a special protection layer called the “floating gate”, which is electrically isolated. So it will not pass any external influence that could possibly change the cell's state.
#4 Be Сareful When Placing a Drive
Avoid placing external drives or the PC in a hot area with no air circulation – in such conditions your hardware will work at peak temperatures and shorten its lifespan. The same for vibration-loaded surfaces and rooms with a lot of dust. For example, most vendors don’t recommend using home-grade hard drives in server rooms, because of vibration-loaded racks for multiple servers. Such conditions require drives of an industrial level.
Level 2 - Using Practice
You can extend drive life by implementing the right usage habits, such as data utilisation and placement, as well as enabling advanced technologies (for example, TRIM for SSD drives). We do not include controversial advice, such as perform defragmentation regularly (modern OS do it automatically by itself) or store the data in a RAM disk, which is not straightforward and can lead to massive data loss.
#5 Reduce Write-Intensive Workloads, Especially for an SSD
Massive data writing tasks on an HDD shortens the life of mechanical parts and the magnetic surface, so you should not use general home-grade drives for video surveillance systems as an example. Hard drives vendors have special product lines for such heavy workloads – for example, WD Red with extended mean time between failures (MTBF) period. Server-grade disk drives are also more enduring for a larger price, of course.
If you run an SSD drive, remember that it has limited writes before the data starts becoming read-only. Though, it is less important for newer models with modern NAND flash memory and complex controllers that ensure the data is written to different cells every time. Since the seek time in an SSD is near zero, fragmentation is not a problem here.
#6 Enable Write Cache
Write cache is generally enabled by default, but some chipset vendors allow you to select whether the HDD controller will use write cache or not. Here is example for Intel server systems.
Write cache allows you to dramatically improve read/write performance of your disk subsystem by caching the data put by the OS in the RAM memory (wither system- or HDD controller-based). But this solution has a big disadvantage: if your server suddenly powers off, the data located in the cache will be lost. In case of server systems, you can avoid this by using battery-backed caching controllers.
In case of home or small office usage, you should implement UPS so the PC could continue working normally in case of power outage.
#7 Enable TRIM Functions for the SSD
Every write operation on an SSD is preceded by erasing the previous cell state (kind of a garbage collection), so multiple write operations lead to degraded overall drive performance. But you can use a special TRIM command that instructs an SSD controller to clear some unnecessary cells when the OS marked such data as “empty”. This approach helps to skip the “clean” operation for most of the writing requests and save a lot of drive performance.
TRIM is embedded to the basic set of ATA commands in the storage subsystem, but you need to ensure that your OS installation uses this feature. You can check TRIM support using the following console commands:
fsutil behavior query DisableDeleteNotify
If TRIM is enabled, you will get “DisableDeleteNotify = 0”.
system_profiler SPSerialATADataType | grep 'TRIM'
In a successful case you will get “TRIM Support: Yes” message.
#8 Always Leave a Dozen Percent of Free Space on the SSD
Since an SSD needs to first remove the old cell state before writing the new one, it needs free space so TRIM can prepare it for writing requests. But if this space is too small, you will face low SSD performance as if TRIM were not enabled at all.
So, always leave 15-20% of the drive space free. If you run large disks (terabytes), you can shorten such “reserved” space size.
#9 Install Monitoring App
Every hard drive has its own internal monitoring system called SMART. It has a lot of numeric parameters, each representing some drive’s characteristic. You only need to install an app allowing to collect and process such data. It is absolutely useless to just get the numbers (remember they are a lot), so it is better to use apps with a suggestions mechanism.
It is also a good idea to implement an email-based notification for critical conditions.
Now you have learned more about extending the hard drive lifespan. As a bottom line, we would suggest a few monitoring tools that could help with the prediction of an emergency breakdown of a disk: