Just started preclearing drives last night for my first unRAID machine, I bought 2 3tb WD Red's (one for parity, one for data) and I went so far as to buy one from amazon and one from newegg. Side note both arrived in great condition as well as had good packaging; amazon had the plastic end caps while newegg had the drive in this bubble wrap sleeve(pretty cool I thought). Anyway from everything I have read on the forums everybody recommended not getting 2 or more drives from the same vendor to reduce the potential of getting a bad batch. Now to some this my be overkill and they have bought large batches without having a bunch fail out once, others not so lucky. So what do people think a good waiting period is? Week, month, etc...

If I needed the drives badly, I would just purchase them if the price was right. I might buy an extra for an on location spare.

 

Otherwise, I would just wait it out until a good or slick deal is posted.

That's usually when I purchase a drive. i.e. here and there.

 

If I needed 10 drives immediately. I might buy two per vendor and jump around vendors or if the price was right, purchase all of what I need from my favorite vendor with an extra drive as an on location spare.

 

In the past with our hosting company, it was common for us to purchase batches of drives and have two fail within a week of one another.  We ended up purchasing less at a time until we had sufficient on location spares.  Then we purchased as we needed.  There wasn't a specific waiting period for us.

 

I did this recently, and my solution was to buy from vendor 1:

 

Your Order
----------
Invoice         E******
Date            13/08/2013 15:42

Product                              Quantity    Price
----------------------------------------------------------
4TB Seagate ST4000VN000 NAS 24x7 3.       1      £161.28
4TB Seagate ST4000DM000 Desktop HDD       1      £131.99
----------------------------------------------------------

 

and from vendor 2:

 

13/08/2013
Manufacturer	 Product	 Quicklinx	 Qty	 Price ex VAT	 Price inc VAT
Seagate	4TB NAS HDD SATA 6GB/s 64MB 3.5" Hard Drive	8QFH	1	 £127.49	 £152.98

 

So I ended up with 2 of the same drive from 2 e-tailers and 1 other drive from the same retailer, but a different model.

 

I am in a deal with vendor 1 where I get free next-day shipping, so I got 3 drives and only paid shipping on 1. All 3 drives were delivered on the same day, and I precleared all 3 at once!

I've read the same thing, 'don't buy several drives at once.'. Yet people and companies by in bulk all the time.  And, if a drive is good wouldn't that mean the drives in that batch should also have a higher possibility of being good?  So order between two vendors and get as many as you need, I say.  RMAing a drive isn't that much of a hassle.

Thanks everyone for the replies, everyone made good points and they were much appreciated! As the 4tb(new 3tb and old 1tb) of usuable space I have right now should hold me over for awhile I will look for deals as I want to have a precleared 3tb HDD on standby.

Speaking of good deals on 3TB drives => Newegg and Amazon have the superb 3TB WD Reds on sale TODAY ONLY at the best price I've seen for that drive [i paid over $200 each for my first few, and about $160 for the rest].

 

http://lime-technology.com/forum/index.php?topic=29970.0

 

[...] Anyway from everything I have read on the forums everybody recommended not getting 2 or more drives from the same vendor to reduce the potential of getting a bad batch. [...]

 

AFAIK the risk is not so much that of a *bad* batch but that of the *same* batch (from the same fab).

 

The reason is that of how MTBF calculation works. The given/advertised error rate is an overall statistical number and holds true for a single disk only.

The individual error a drive experiences however will be influenced by its operating mode/usage pattern and its environment.

So across all drives of the same model, errors events are spread across the error rate time window.

 

Theory is this, that drives from the same batch (and fab) running in the same operating environment and on same usage pattern are most likely to fail at the same time.

This is more likely to happen for drives of the same batch, as they share some more mechanical "features" in their "DNA"...as the parts they are build of are likely to be picked also from the same batch of supply.

Hence statistically spread error events are more likely to align at the same point in time for these drives.

 

I know of enterprise admins that, although buying in bulk, are mixing drives in their SAN arrays from different shipments,

because this effect had also happened to them before. In the case an error occurs, now hopefully not all drives would fail at

roughly the same time, giving some headroom window to the repair team.

 

One other point to consider is the staff handling the drives in the warehouse or retail store. I once witnessed a full stack of hard drives being dropped from top stock (about 12 foot drop) and they were just put on the shelf as if nothing happened.

Also with online purchases you have to worry about USP/FedEx and how they handle the package.