hard drives: WD Red vs WD Red PRO

[rogerh]

What fascinates me about the Red Pro drives is that they are recommended for use in a group of "six to sixteen", the implication being that they would somehow work less well if used without at least five companions. This is somewhat counter-intuitive.

 

[jgreco]

They're programmed to communicate wirelessly and start losing bits once you introduce that 17th drive.

 

[rogerh]

They're programmed to communicate wirelessly and start losing bits once you introduce that 17th drive.

The scary thing is that this is not technically impossible; though there might be a bit of scandal for a week or two if discovered.

 

[jgreco]

That's the terrifying thing about technology today. It's perfectly conceivable that you could do all sorts of things, possibly even within firmware on the drives.

 

[DrKK]

The problem is that in three years, the cost for a disk has usually fallen through the floor. If you're paying top dollar for that drive today, it is likely to cost less than half that in three years.

I agree completely. Well said sir. You demonstrate something interesting to me: the fundamental flaccidity of the logic underpinning the decision to buy tech that is fancier/more expensive than strictly needed, concurrent with the failure to recognize that the buyer is essentially motivated by (I am going to coin some French terms right now) choses brillantes as opposed to choses utiles.

One problem, I think, with technical guys like us, is they simply *WANT* "better" stuff for its own sake ("choses brillantes"). They worry about justifying that later. It doesn't really matter, at all, if they need it. That's why everyone has a Xeon ($250+) here when 85% of them could have an i3 ($129) or, even, a G3220 ($45), and literally not be able to measure the difference during normal, organic usage. That's why everyone has hot swap, when they'll probably never swap anything. That's also why a lot of people have SATA DOMs, L2ARC's, ZIL's, and everything else. (Of course, some people do need those things, and they are must-haves in some contexts. But in many cases they are just spinning rims on your tires.)

And so on and so on and so on. Typically, when queried on the point, the person overbuying their tech (or anything else) will have some kind of a posteriori and ad hoc argument about "convenience", the "value of his time", "future proofing" (that's my favorite), or (especially) "resale value". But the truth is, they chose to buy the fancier tech a priori, and the justifications are dishonest in nature as they did not fundamentally inform the tech buying decision.

In your journey through tech, nothing will actually empower you more than the hundreds (or thousands) of dollars you will save every year by asking yourself hard questions about what you truly need, and what you need to stay "in the game" and intellectually engaged in tech.

On the other hand, common consumers overbuying their tech, effectively, subsidize me, since it takes some of the profit pressure off of the pricing points for other equipment. For example, you can bet the WD Red is $10 or $20 cheaper than it would otherwise be because they've got suckers and commercial enterprises buying the WD Red pro. :)

 

[jgreco]

@DrKK while I agree with much of what you've said, I'll also raise the point that part of the consideration there needs to be the anticipated lifecycle of the equipment. You can argue, perhaps, that this is a variation on "future proofing", but I think this is a more complicated issue than at first glance.

For example, in the mid-2000's, during the "upgrade to 1000baseT", I purchased lots of 1G gear at a relatively low cost (Dell 5324's, largely) which turned out to have a ~10 year lifecycle in service here. I've been upgrading them for the last year or two. When purchased, they were perfectly suited to the environment here - with no extras. We had lots of individual servers and didn't tend to move gobs of traffic around the network; many of the servers were 100Mbps. As time went on, though, we ended up needing to bodge around certain issues, such as putting in PoE midspans to power gear over the network. As the network mix changed (virtualization, network storage, remote desktop/consoles, TV-over-IP, etc) it became clear that the old 1G switches were extremely limiting. So when it came time to upgrade, after ten years of kinda-suffering with the 1G switches, I opted to pop for mid-level gear instead. The edge switches are nice 1G PoE switches with layer 3 routing, full filtering capabilities, and 4x10G uplinks. That's already paid off in simplifying things that used to either require a PoE midspan or a local power supply, and the routing capability, wow, very nice. For the core switches I bumped up to 10G with 40G uplink. All of a sudden, VM's fly when migrating them around the cluster. The question is, how will this pan out over the next ten years? My crystal ball is cloudy, but I think it was definitely a better choice to get the stuff that had a shot at lasting a longer time because it had fewer weaknesses to begin with.

In that same way, I think we need to be careful because the lifecycle of a NAS could easily be 5-10 years. You do want to make some reasonable accommodation to make sure your device isn't totally useless in a year or two.

 

[DrKK]

In that same way, I think we need to be careful because the lifecycle of a NAS could easily be 5-10 years. You do want to make some reasonable accommodation to make sure your device isn't totally useless in a year or two.

I'm quite sure you're right.

But the way people like you see things, and reason things, is not at all like the way the average consumer (and the average consumer drives the market---I mean, someone bought all those Billy Ray Cyrus albums, right?) does. I guess I'm saying that I have little faith in the average consumer (even, tech consumer) to make a correct analysis of these factors.

 

[joeschmuck]

I'm a cheap bastard so I went with likely the only AMD system out here. It's fast and cheap. I purchased it with future proofing in mind as well because I'd rather not purchase replacement hardware other than hard drives and cooling fans. I think if I had to purchase a system again, I'd go for a Supermicro board and and i3 cpu because I like to tinker and use jails. I never know what I might want to push my system to do.

I agree that DrKK has a point that many people purchased more powerful hardware than they will ever use however I believe it give those folks a peace of mind with respect to compatibility and hopeful future proofing.

As for the original thread, Red v.s Reds Pro... That is up to the persons perspective purchasing the drives. Do they need or desire 7200 RPM with faster IOPs (even if they don't utilize them) and then the warranty period being longer vs. the cost. In my opinion sometimes it's worth paying for the longer warranty, after all if you are buying a Red over a Green then in reality, for a FreeNAS based system, you are choosing warranty vs. price because a Green drive runs fine in FreeNAS. I will always take a hard look at the pros and cons, then make a choice that makes me happy.

For me, a 7200 RPM drive is not for a home NAS system such as mine. I will never see a difference in drive performance for what I do and as previously mentioned, the heat added heat has to be dealt with, and the extra noise may be an issue for some folks.

As for wireless hard drives, I heard the WD Red Pro+ has what is called "Red-Tooth" which is based on a 16GHz carrier @ .5 mWatts and has an operational range of .5, 1.0, or 1.5 meters (jumper selectable). It can transfer data between up to 127 drives (the 128 spot is the controller of a single drive). The way I understand it, one drive is all that is needed to be connected via SATA to the host computer, any additional hard drives not connected via SATA when powered up will search for another Red-Tooth drive and wirelessley connect to it. The drives can transfer data between each other at a very high rate of speed whereas the single SATA alone is limited to 6Gb/s. There are two main benefits with this technology, 1) You can expand the number of hard drives beyond the physical number of SATA ports, and 2) (the good part) when transferring information from a drive or set of drives, all the information is split across all physical SATA ports so it can transfer data at up to the maximum speed of each SATA port, well except there is a limit of 630 MB/s (whatever that is in Gbps) according to the spec sheet. That is FAST!. Just to give an example: If you have four 10TB hard drives, all connected with a SATA port to your motherboard and you request to transfer a 5GB file from hard drive #2, all the hard drives will communicate and pass the data over all four SATA ports to your computer thus in effect maximizing all the SATA port bandwidth. I mean there are still some physical hard drive limitations such as IOPs and moving the head around, but you get the point, and I cannot defend WD on what they are doing but it sounds pretty cool. I've attached the spec sheet for the drive, it's pretty interesting and I can't wait for it come hit the streets, although I wouldn't purchase it just yet, I have no need for something like this and I'm sure it will be very expensive and there could be some unforeseen bugs. I've attached a specs sheet that my ex-brother-in-law gave me, it doesn't really detail anything about how the system works. BTW, he works for WD so I guess it's a good thing we are on speaking terms after I divorced his sister.

 

[jgreco]

I'm quite sure you're right.

But the way people like you see things, and reason things, is not at all like the way the average consumer (and the average consumer drives the market---I mean, someone bought all those Billy Ray Cyrus albums, right?) does. I guess I'm saying that I have little faith in the average consumer (even, tech consumer) to make a correct analysis of these factors.

Fair enough. I'm sure we've seen a number of home users who are creating monster-scale systems suitable for corporate use when they only really needed something half as big.

 

[jgreco]

I've attached the spec sheet for the drive, it's pretty interesting and I can't wait for it come hit the streets, although I wouldn't purchase it just yet, I have no need for something like this and I'm sure it will be very expensive and there could be some unforeseen bugs. I've attached a specs sheet that my ex-brother-in-law gave me, it doesn't really detail anything about how the system works. BTW, he works for WD so I guess it's a good thing we are on speaking terms after I divorced his sister.

You sure that's cleared for public consumption? And p.s. yes it sounds bat$#!+ crazy.

 

[rogerh]

You sure that's cleared for public consumption? And p.s. yes it sounds bat$#!+ crazy.

I don't know, at least it doesn't claim they're going to specify their new drives in TiB; that really would be a dead give-away that it was made up.

 

[jgreco]

It's about six months off for an April Fool's joke.

 

[Ericloewe]

I'm a cheap bastard so I went with likely the only AMD system out here. It's fast and cheap. I purchased it with future proofing in mind as well because I'd rather not purchase replacement hardware other than hard drives and cooling fans. I think if I had to purchase a system again, I'd go for a Supermicro board and and i3 cpu because I like to tinker and use jails. I never know what I might want to push my system to do.

I agree that DrKK has a point that many people purchased more powerful hardware than they will ever use however I believe it give those folks a peace of mind with respect to compatibility and hopeful future proofing.

As for the original thread, Red v.s Reds Pro... That is up to the persons perspective purchasing the drives. Do they need or desire 7200 RPM with faster IOPs (even if they don't utilize them) and then the warranty period being longer vs. the cost. In my opinion sometimes it's worth paying for the longer warranty, after all if you are buying a Red over a Green then in reality, for a FreeNAS based system, you are choosing warranty vs. price because a Green drive runs fine in FreeNAS. I will always take a hard look at the pros and cons, then make a choice that makes me happy.

For me, a 7200 RPM drive is not for a home NAS system such as mine. I will never see a difference in drive performance for what I do and as previously mentioned, the heat added heat has to be dealt with, and the extra noise may be an issue for some folks.

As for wireless hard drives, I heard the WD Red Pro+ has what is called "Red-Tooth" which is based on a 16GHz carrier @ .5 mWatts and has an operational range of .5, 1.0, or 1.5 meters (jumper selectable). It can transfer data between up to 127 drives (the 128 spot is the controller of a single drive). The way I understand it, one drive is all that is needed to be connected via SATA to the host computer, any additional hard drives not connected via SATA when powered up will search for another Red-Tooth drive and wirelessley connect to it. The drives can transfer data between each other at a very high rate of speed whereas the single SATA alone is limited to 6Gb/s. There are two main benefits with this technology, 1) You can expand the number of hard drives beyond the physical number of SATA ports, and 2) (the good part) when transferring information from a drive or set of drives, all the information is split across all physical SATA ports so it can transfer data at up to the maximum speed of each SATA port, well except there is a limit of 630 MB/s (whatever that is in Gbps) according to the spec sheet. That is FAST!. Just to give an example: If you have four 10TB hard drives, all connected with a SATA port to your motherboard and you request to transfer a 5GB file from hard drive #2, all the hard drives will communicate and pass the data over all four SATA ports to your computer thus in effect maximizing all the SATA port bandwidth. I mean there are still some physical hard drive limitations such as IOPs and moving the head around, but you get the point, and I cannot defend WD on what they are doing but it sounds pretty cool. I've attached the spec sheet for the drive, it's pretty interesting and I can't wait for it come hit the streets, although I wouldn't purchase it just yet, I have no need for something like this and I'm sure it will be very expensive and there could be some unforeseen bugs. I've attached a specs sheet that my ex-brother-in-law gave me, it doesn't really detail anything about how the system works. BTW, he works for WD so I guess it's a good thing we are on speaking terms after I divorced his sister.

Careful, that looks good enough to pass a not-too-casual inspection. People are going to blame you for something, for sure. :D

It's about six months off for an April Fool's joke.

A wasted opportunity, @joeschmuck

 

[joeschmuck]

Just thought I'd liven up this thread a bit.

 

[jgreco]

A wasted opportunity, @joeschmuck

I did note that it was dated Nov 2015. So the thing is, on one hand I can see them doing something like this because technology has gotten sufficiently cheap that the drive manufacturers have suddenly discovered that you can offload some abstraction responsibilities to the HDD (see: Seagate Kinetic). The concept of an ethernet connected hard drive that stores objects is ... intriguing, after all. So I guess that tackling more conventional storage with something like this ... it could happen.

 

[joeschmuck]

I think it would be fantastic if something like this could be done but I see a few issues such as how a computer would rejoin data split across multiple SATA ports, then what if you wanted some sort of isolation between the drives so they wouldn't connect to each other automatically. In the future I actually could see something like this, after all there is distributed processing and it could be a form of that. Eh, it was just something I made up in the moment. I'm sure someone will take a look at it and make money off the idea and I'll be high and dry, what's new. Well you do have quantum computing being developed, not in the same sense as a normal computer yet but it will get there, hopefully in my lifetime.

 

[jgreco]

The technology is loosely there already; there's a lot of experience with mesh networks and IoT style ad-hoc networks. The real problem is that we haven't really done a great job of figuring out how to cope with resources that come and go, because for something like storage, that's a big thing. Redundancy of data and then also just creating an environment where that's not likely to happen. Look at the Kinetic drives. I just have these nightmares where Seagate makes the next "logical" move by sticking them into a clamshell with an external brick power supply, and selling them with custom software loaded onto your PC to act as "network backup" drives. The only significant thing missing there is some software.

 

[TheDubiousDubber]

Fair enough. I'm sure we've seen a number of home users who are creating monster-scale systems suitable for corporate use when they only really needed something half as big. :)

Indeed. I will probably never utilize the power of a dual xeon, in fact rarely would i utilize the entirety of one. Though I couldn't argue with the price of free. Sometime people build around the spare parts they have lying around. The energy cost is higher, but will likely never exceed the cost of a newer system over what I paid for a chassis to fit the hardware I already had.

 

[jgreco]

That's a rational evalulation of capex+opex. What we are talking about is irrational choices.

 

[rogerh]

That's a rational evalulation of capex+opex. What we are talking about is irrational choices.

Is "because it's shiny" not rational then! Please don't tell the car industry.