HDD replacement if there is no like-for-like

[Warwick_Tech]

Hi Guys,

 

How do you deal with drive failures when you lack the spare parts like for like?

 

I'm lucky enough to have some spare HDD's for my server farm but supplies are running low and they've all gone EOL -

 

Is it simply a case of finding the modern 'equivalent'? anyone have any companies they use to source specific parts?

 

So far we have:

* 1TB SEAGATE ST1000NM0023

* 256GB TOSHIBA THNSNJ25 (Now confirmed EOL)

* 128GB TOSHIBA THNSNJ12 (Now confirmed EOL)

* 4TB SEAGATE ST4000NM0025 (These were replacements for an older model)

[Chris_Cook]

It depends on the number of disks in the array, and how critical the workload is.

 

For small arrays or critical loads, I'd try and push for a new set of disks and switch them 1 by 1, or re-build from scratch.

 

for large arrays or non-critical loads, then I've had some luck with getting a more recent disk from the same manufacturer that's as close as possible. The disk size needs to match or be bigger. Disk access speed should be close.

 

If the new disks are much bigger, then you can often increase the overall array size by going from 1Tb to 4TB for example. You only get the new size when all disks have been changed.

 

What you need to consider is how likely the rest of the disks are going to fail in the near future. If the disks are properly EOL, then the risk is quite high. Even 'new' disks may have sat on a shelf for years. Lubricants and capacitors can all degrade over time as well as usage. Boxes get dropped and put back on shelves etc.

 

You also need to check with your raid controller as to what it supports for both size and speed.