Originally Posted by Nevcairiel
Because you only quote some arbitrary review sites, while others actually report their own real-world experience with the hardware?
Experience is always much more valuable then some numbers some biased person quotes on the interwebs.
My FlexRAID pool is about half 3TB RED drives, and the other half GREEN drives. While the greens are indeed rather slow, the REDs will usually average at 130MB/s, which is plenty fast for storage drives for me.
Speed should always compare same size drives, as the speed can vary between different number of platters, etc.
I guess the 12 hard drives I own and my experience does not count ? The fact I see exactly the same things that the review sites report makes me feel like I can't be the only one.
I am not saying your 3TB RED can't do 130MB/sec when it's new, it is fresh and empty, or it's using the faster part of the platter. I am saying it can not maintain that speed consistently long term. This is fact and normal for all HDD's. I feel like many WD owners and advocates are trying to take a best case scenario and pretend it's "normal" somehow downplaying the issue in support of their own preference and purchase decisions.
If we are going to do that then here's what I see with my Seagate 7200.14
Comparing a best case scenario to a worst case scenario just blurs the picture.
Here is the facts:
Seagate 7200.14 3TB ($99)
BEST CASE: 210MB/sec
AVERAGE SPEED: 160MB/sec
WORST CASE: 100MB/sec
WD RED 3TB ($140)
BEST CASE : 150MB/sec
AVERAGE : 100MB/sec
WORST CASE : 65MB/sec
At any point when you compare apples to apples and not apples to oranges you get a speed difference of nearly 50MB/sec or more. My point and my major issue with 5400RPM drives is when your drive is no longer new, and fresh, and empty then your speed will drop down rather significantly. For this reason- using benchmarks that a HDD can only achieve when it's empty and brand new and uses the fastest part of the platter as a basis for making a point is borderline deception if it's not just ignorance. I am not sure about you but nearly every drive in my server is full. Why would you buy a storage drive if your not going to use it and fill it up ? When I fill them up, I buy another. I keep buying storage for my media as I need it. I think this is how normal people do it. Leaving a HDD empty and buying more seems a waste of money and even poorer value IMO.
For this reason I always try to look at the big picture with HDD's. The big picture is that the speed of a WD 5400RPM drive is going to drop down below the speed capability of my LAN and impose a performance bottle neck, where the Seagate maintains enough performance even in the worst possible scenario to make this mostly a non issue. Please keep in mind there is a performance penalty imposed by many other things, including FLEXRAID that you and I both run. So if your drive does a theoretical speed in a pure benchmark it's very unlikely to hit that speed inside your flexraid array under normal typical usage.
The bottom line is that starting with a HDD that does 160MB/sec average means you end up with a worst case scenario after performance is lost, the drive is full and the other factors like software RAID's effects are factored in, that keeps you near your LAN capability. Most LAN in consumer set ups can do 100MB/sec consistently and the "theoretical" or "ideal performance" of Gigabit LAN is 125MB/sec. I very rarely get 125MB/sec on my LAN and my server and desktop are 3 feet from each other, with CAT6 and INTEL LAN NIC on both machines. I like to think my set up is above average and still I only see 100MB/sec to 120MB/sec. I have no doubt that a longer run, a crappier LAN NIC chipset, not CAT6, and slower HDD's inside software RAID would not yield the speeds I get, nor would they be close to the capability of gigabit LAN speed.
I am very happy with performance that is near the speed capability of my LAN (gigabit switch) I do not need 200MB/sec + speeds at all. But reality is that using a faster HDD is required to achieve this and the final results you get are going to be very different from the theoretically ideal benchmarks you might see. There is no way you can see 130MB/sec with your RED drives you say you do. That is faster than LAN will go for basic consumer hardware, and the RED drive will slow down to under 70MB/sec when it's old and filled up. Add into this the performance hit you get from FLEXRAID (which is still better performance than most other software RAID programs like Unraid) and your much more likely to actually be about half the speed your pretending.
I'm not pulling this stuff out of my ass. I have tested this, owned it, lived it. I believe I am 100% correct and would challenge you or anyone else to prove me wrong. I am mostly concerned with the final actual result, not ideal benchmarks. And I am highly concerned with maximizing value and my money spent. For both of these the clear winner is Seagate for me, and the clear loser across the board is the 5400RPM WD DRIVES like RED and GREEN. You may disagree but please give me enough respect not to suggest : "Because you only quote some arbitrary review sites, while others actually report their own real-world experience with the hardware?
Experience is always much more valuable then some numbers some biased person quotes on the interwebs."