Electro-mechanical hard disks are designed to spin continuously. For most 3.5” form factor disks, rotational speed is 5400, 7200 or 10,000 revolutions per minute. If the disk is used in a blade or tower server, for example, it will get cooled by the host’s system fan and will hopefully have a steady supply of clean power. Operating in an ambient temperature, such a disk (whether standalone or RAID) can run for several years without interruption.
However, there is one risk factor which a lot of IT admins forget about. As the disk(s) is running, because it uses an “air bearing”, some external air is inducted. This air is filtered by a tiny filter known as a barometric or breather filter. In addition to this, due to the effects of internal component wear and tear, tiny debris from the platters can also start to accumulate inside the disk chamber. For the most part, even with debris accumulating inside the disk, the read-write process can continue as normal. That is until, some poor IT person gets assigned the task of physically moving the server or migrating its data as they can be in for a nasty surprise.
Take last week, for example where a company in Dublin got into a spot of bother with their old Dell PowerEdge server running Windows Server 2008. Their IT administrator was tasked with the job of decommissioning it. The server was running fine, but was slow and no longer meeting the organisation’s requirements. He turned the system off and carried it back to his basement office with the intention of doing a complete backup. However, back at his office, he switched it on again, only to be greeted with the hue of a Windows Server 2008 “blue screen of death” informing him about an “Unmountable_Boot_Volume”. He removed the disk (Hitachi HDT721010SLA360) and slaved it onto another PC. No dice. In Computer Management, the disk was showing up as “unformatted”. This was the last thing he wanted. So, if this disk was spinning fine for the last 12 years, why did it pick the most inopportune time to kick the bucket?
Well, when you move an old hard disk which has been in-situ for years, the dust and debris collected by its air filter can get displaced. This can result in particulate matter getting strewn across and platters and collecting under the disk-heads, making the drive unreadable.
Drive Rescue took the disk into our clean-room where we removed the head disk assembly and cleaned the disk platters using a process which merits another blog post. We were able to recover 98% of their data.
Lesson: the benefit of in-situ backups…
Servers can be located in the most uncomfortable places such as under staircases or in cramped comms rooms. The temptation for the IT admin to move an old server and perform a full disk backup in a more congenial environment can be quite strong. However, before moving the server anywhere or removing its disks, it would be prudent to a use a disk replication tool such a Macrium Reflect to copy the server’s volume onto another medium. This should be performed while the server is in-situ. This way, you can prevent any nasty surprises and need the call a data recovery service!
Drive Rescue are based in Dublin. Ireland. We offer a full server data recovery service. This includes Windows Server 2003, Windows Server 2008, Windows Server 2012, Windows Server 2016 and Windows Server 2019. Our service covers both standalone disks (S-ATA, SAS) and RAID (0, 1, 5,6,10)