AVS › AVS Forum › Video Components › Home Theater Computers › Planning to rebuild my 20TB WHS Flexraid server: Information requested.
New Posts  All Forums:Forum Nav:

Planning to rebuild my 20TB WHS Flexraid server: Information requested. - Page 4

post #91 of 916
11/14/12 at 8:45pm
Thread Starter 
And for the record I'm not scared and still use green drives in my server.

I'd even buy another if I found a great deal.

I think to generally sum up my thinking : I want to believe its crap because I'm already using green drives and don't want to stop or change.

But assuming I can get a RED or non intellipark drive for same or lower cost/capacity I'll probably play it safe.
Reply
Reply

AVS Top Picks

post #92 of 916
11/14/12 at 8:47pm
Quote:
Originally Posted by Mfusick View Post

And for the record I'm not scared and still use green drives in my server.
I'd even buy another if I found a great deal.
I think to generally sum up my thinking : I want to believe its crap because I'm already using green drives and don't want to stop or change.
But assuming I can get a RED or non intellipark drive for same or lower cost/capacity I'll probably play it safe.

I think Reds are fine for the same cost but I wouldn't pay a penny more for one. But they also are new and relatively unproven at this point as well.

Would appreciate some actual data. I will stand by until someone posts something objective.
Reply
Reply
post #93 of 916
11/14/12 at 8:49pm
BTW just wanted you to see my post from the bottom of the last page. WD Black drives also park their heads --- at 12 second intervals. You seemed to have left that data point out in your "theory".

Oh, also forgot to mention another debunked theory --- the "batch" drive theory. There are actually studies that suggest that you should buy drives in batches (especially at the end of a drives manufacturing process or generation as that's where some/most of the manufacturing bugs may have been streamlined) to increase your likelihood of long life and reliability.

This was very controversial when I first proposed that avoiding buying 3-4 drives at the same time for fear that you would get the "same batch" should be avoided and was "hogwash".
Edited by assassin - 11/14/12 at 8:58pm
Reply
Reply
post #94 of 916
11/14/12 at 8:58pm
Thread Starter 
Quote:
Originally Posted by jim2100 View Post

Snapshot RAID is the way to go.

I thought snapraid has slower write speeds ???
Reply
Reply
post #95 of 916
11/14/12 at 9:02pm
Thread Starter 
Quote:
Originally Posted by assassin View Post

I think Reds are fine for the same cost but I wouldn't pay a penny more for one. But they also are new and relatively unproven at this point as well.
Would appreciate some actual data. I will stand by until someone posts something objective.

I'd second that and anything additional added.

It would be nice to jump on a Black Friday deal with a clean conscience if its a green.

I've had mixed luck but I'm not ready to give up yet.

Seagate 3TB are starting to dip down to 109$ and I've had good luck with those so far too.

I'm not at all brand particular. I've got Samsung, Hitachi, Seagate and WD currently running now... Half dozen one and 6 of the other if you ask me.
Reply
Reply
post #96 of 916
11/14/12 at 9:06pm
Thread Starter 
Quote:
Originally Posted by assassin View Post

BTW just wanted you to see my post from the bottom of the last page. WD Black drives also park their heads --- at 12 second intervals. You seemed to have left that data point out in your "theory".
Oh, also forgot to mention another debunked theory --- the "batch" drive theory. There are actually studies that suggest that you should buy drives in batches (especially at the end of a drives manufacturing process or generation as that's where some/most of the manufacturing bugs may have been streamlined) to increase your likelihood of long life and reliability.
This was very controversial when I first proposed that avoiding buying 3-4 drives at the same time for fear that you would get the "same batch" should be avoided and was "hogwash".

It's not my theory. I'm just asking ...

I'm not emotionally involved or knowledgable enough to really go any further on this subject. I'd love to hear more opinions though...

If you want to argue we can talk about SSDs.. wink.gif
Reply
Reply
post #97 of 916
11/14/12 at 9:07pm
Quote:
Originally Posted by Mfusick View Post

Half my drives are green but in shying away from more the more and more I read about it.
Quote:
Originally Posted by Mfusick View Post

So I did some digging and. It appears green drives are a poor choice (WD) if you have constant use or run software raid like Flexraid.

So will you at least admit that your "digging" turned up no actual usable data that led you to your current opinion that you are shying away from Green drives?
Reply
Reply
post #98 of 916
11/14/12 at 9:10pm
Thread Starter 
Hey I just noticed your edit and add of the "batches" theory ...

It's interesting and I generally believe in that.

I just don't know any reasonable way to accurately predict or maximize that.. So I just buy what I need when I need it.

I usually buy drives one or two at time because I'm a cheapo .. And also I tend to only need that at once.
Reply
Reply
post #99 of 916
11/14/12 at 9:13pm
Quote:
Originally Posted by Mfusick View Post

Hey I just noticed your edit and add of the "batches" theory ...
It's interesting and I generally believe in that.
I just don't know any reasonable way to accurately predict or maximize that.. So I just buy what I need when I need it.
I usually buy drives one or two at time because I'm a cheapo .. And also I tend to only need that at once.

http://www.avsforum.com/t/1420736/buying-hard-drives-from-the-same-batch#post_22230968
Reply
Reply
post #100 of 916
11/14/12 at 9:15pm
Thread Starter 
Quote:
Originally Posted by assassin View Post

So will you at least admit that your "digging" turned up no actual usable data that led you to your current opinion that you are shying away from Green drives?

No. I told you this all started directly from WD support email and recommendation/ explanation.

I'll forward it to you when I'm back home. Just email me and remind me.

And I didn't do any digging either.

(I'm in Hawaii till thanksgiving with only my iPhone. There isn't even cell service here in the jungle... But I have WIFI.. smile.gif )

I'll be back at resort tomorrow.

Why you up so late ? It must be late there. Lol...
Reply
Reply
post #101 of 916
11/14/12 at 9:20pm
BTW for those interested and with access here is one of the papers on batches...

Okay, sorry to get this off track. Stated opinions on things that have no data behind them (and actually often times have data that refutes what is being said) drive me absolutely crazy. The tech world is full of this garbage.

Disk drive vintage and its effect on reliability
Quote:
Continuous improvements in the
manufacturing processes could include tighter process
controls, close monitoring of process shifts, implementation of
better screening tests, better handling procedures, design
changes on major failure modes, applying firmware fixes as
necessary, higher cleanliness due to reduced particle
contamination from operators, tools, jigs/fixture etc. These
changes over time reduce the occurrence of failures in field
and hence over time improved the reliability of drives.
Reply
Reply
post #102 of 916
11/14/12 at 9:26pm
Thread Starter 
Quote:
Originally Posted by assassin View Post

http://www.avsforum.com/t/1420736/buying-hard-drives-from-the-same-batch#post_22230968

Before clicking it I'm pretty sure I read this before and remember it.

Thanks ill check it out.

You should know my general feeling about reliability by now.

I'm generally not scared. I'll take a risk. I value price and performance over perceived or suggested reliability because I think too many different factors effect end user reliability and its all just too much to predict with any reasonable accuracy.

I think "batches" fits into this. You don't know if a batch is good or bad - and even inside a single product line you can see variations in reliability.

So in end I choose to discard that from my decision making process because I can't predict or control any of it with any reasonable certainty. A MFG warranty or professional review does hold a bit of credibility for me but I just can't figure out how you could know if you got a good batch or a bad batch ???

Assuming you could - buying multiple from same "good" batch would be as good a strategy as buying from different times/sources to avoid all from same "bad batch".
Reply
Reply
post #103 of 916
11/14/12 at 9:33pm
Thread Starter 
Quote:
Originally Posted by assassin View Post

Okay, sorry to get this off track. Stated opinions on things that have no data behind them (and actually often times have data that refutes what is being said) drive me absolutely crazy. The tech world is full of this garbage.

Funny how I feel this way about Sandforce SSD drives in NOV2012 being bashed like its NOV2011. People seem to think every sandforce SSD is like a first generation OCZ line rushed to market a year ago.
Reply
Reply
post #104 of 916
11/14/12 at 9:41pm
Thread Starter 
Here's a good piece of info:
Quote:
Daniel Mauerhofer, Head of EMEA PR in Western Digital has sent us a comment regarding the Critical Design Flaw Found in WD Caviar Green HDDs article. "What you have described as a “critical design flaw” in a recent article is rather a misuse of a drive, which was not designed for the referenced system type. In summary, this issue is far from critical, does not impair normal functionality or place data at risk, can be mitigated with a utility for the small percentage of systems affected, and is a well understood industry standard.

This design feature is designed for power savings in systems that can take advantage of the Idle 3 mode. With the heads parked in Idle 3 mode, a typical WD Caviar Green drive with 4 platters saves an additional 400mW of power, which is roughly 10% of the power consumed in Idle 2 mode. This has been a standard feature of modern notebook drives.

Many OS’s, including Windows, have been modified to take advantage of this power savings mode (which has been standardized by the T10 and T13 committees) by eliminating a sync operation that occurred every 30 to 60 seconds regardless of host or user inactivity.

The systems in which you are referencing that are reporting the issue has not been modified appropriately, and for these systems, the Idle 3 feature should be turned off.

The WD Caviar Green drives are not recommended for and are not warranted for use in RAID environments utilizing Enterprise HBAs and/or expanders, nor in multi-bay chassis, as they are not designed for, nor tested in, these specific types of RAID applications. For all Business Critical RAID applications, please consider WD’s Enterprise Hard Drives, which are specifically designed with RAID-specific, time-limited error recovery (TLER), are tested extensively in 24x7 RAID applications, and include features like enhanced RAFF technology and thermal extended burn-in testing. The Idle 3 feature can be disabled for those systems that are unable to take advantage of this power savings mode. WD has always provided applications engineering assistance to system integrators to maximize system performance while also minimizing the amount of power consumed by our hard drives."

Not sure it's 100% right on the money but it basically says same thing WD support told me.
Reply
Reply
post #105 of 916
11/15/12 at 12:47am
Quote:
Originally Posted by assassin View Post

Also, there is also data that temperature has nothing to do with hard drive failure rates as long as you are within the range of operation (i.e. lower temperatures do not equate to lower failure rates) taken from Google's 100,000 drive data.

I'm not sure what you mean by "range of operation", but I think your statement is a bit misleading. I think many people will assume "range of operation" is the temperature range listed on the HDD spec sheet.

But if you look at the google chart, the failure rates increase significantly outside of the range 25 to 50C. Most HDDs are rated to work at a larger temperature range than that, but if you believe the google study, your best bet is to stay between 25 and 50C (or even better, 35 to 45C, but the improvement there is marginal).
Reply
Reply
post #106 of 916
11/15/12 at 3:05am
Quote:
Originally Posted by Mfusick View Post

Here's a good piece of info:
Not sure it's 100% right on the money but it basically says same thing WD support told me.

There is a difference between hardware and software raid and how they use their drives.

http://wiki.flexraid.com/2011/03/26/beginners-guide-to-flexraid-2-0/
Quote:
You may have heard that some disks have something called TLER disabled and can’t be used for RAID. No worries, not true with FlexRAID.

Again, this is still just opinion or theory. I still haven't seen any actual data.
Edited by assassin - 11/15/12 at 3:16am
Reply
Reply
post #107 of 916
11/15/12 at 3:07am
Quote:
Originally Posted by jim2100 View Post

I'm not sure what you mean by "range of operation", but I think your statement is a bit misleading. I think many people will assume "range of operation" is the temperature range listed on the HDD spec sheet.
But if you look at the google chart, the failure rates increase significantly outside of the range 25 to 50C. Most HDDs are rated to work at a larger temperature range than that, but if you believe the google study, your best bet is to stay between 25 and 50C (or even better, 35 to 45C, but the improvement there is marginal).

You are correct. 25-50c is a fairly wide margin which is what I meant. Thanks for correcting or clarifying my statement.
Edited by assassin - 11/15/12 at 3:18am
Reply
Reply
post #108 of 916
11/15/12 at 4:24am
Not to put this thread back on track, but I have been following this for a while and I am just wondering a few things based on the original post.

I haven't seen anyone recomend FreeNAS or unRAID, not sure why?

Since MB's were in question why hasn't anyone recomended one of the atom based ones that are <$100. Any of the D525 based cpu's are basically over kill for low use systems.
These ship with FreeNAS system when purchased from the OEM. But if running WHS is what you are looking to do, then I would go with the fore mentioned SuperMicro setup, as that would give you many options if you decided to go with using any VM's in the future.
Reply
Reply
post #109 of 916
11/15/12 at 4:35am
Quote:
Originally Posted by assassin View Post

... There are also reports that enterprise drives are no better than non-enterprise drives. And that 7200 RPM drives have a higher return and failure rate than non-7200 RPM drives. ...

I have read that as well, do you belive it? I don't. Enterprise class drives "seem" to be more reliable when stacked in drive cages, as they limit spindle vibration. I think the test's most OEM's do are on individual drives in very unrealistic evviroments to help make there data more positive. Googles drive pool is a great source of data, as you have mentioned.
Reply
Reply
post #110 of 916
11/15/12 at 11:08am
Thread Starter 
Quote:
Originally Posted by skybolt View Post

Not to put this thread back on track, but I have been following this for a while and I am just wondering a few things based on the original post.
I haven't seen anyone recomend FreeNAS or unRAID, not sure why?

Probably because I started this thread and I already own and use Flexraid.

The only negative most people seem to think about Flexraid is its not free. Since I already own it I guess it just wasn't a question if others would be good too.
Edited by Mfusick - 11/15/12 at 11:14am
Reply
Reply
post #111 of 916
11/15/12 at 11:24am
Thread Starter 
Quote:
Originally Posted by assassin View Post

There is a difference between hardware and software raid and how they use their drives.
http://wiki.flexraid.com/2011/03/26/beginners-guide-to-flexraid-2-0/
Again, this is still just opinion or theory. I still haven't seen any actual data.

Good info.

I'm not expert but I think you might be blending issues with green drives.

The quote you list about TLER is strictly about why green drives absolutely should not be used in a hardware RAID set up.

But the constant head parking in a 24/7 software raid set up is a totally different issue.

TLER issues will give you data errors in hardware raid.

But 24/7 server operations with green drives and head parking simply just exposes the green drive to an excessive number on off duty cycles limiting life expectancy.

There are rated for a million. So it might be a while before anyone sees an issue here.
But in the end I tend to think WD support email makes sense and a drive more appropriate might be a better choice.

I would not pay $50 more however. But I'd take a RED at same cost as I do think it would last longer in a server environment.

Head parking and intellipark is totally different issue with green drives than TLER and hardware RAID cards.

Both are issues with green drives but TLER only effects hardware RAID, and that's why the owner and creator of Flexraid is debunking that regarding his software raid.
Reply
Reply
post #112 of 916
11/15/12 at 11:29am
But somehow wd black drives don't get the same bad rap despite also parking heads. So they somehow get a pass?
Reply
Reply
post #113 of 916
11/15/12 at 11:40am
I mentioned Tler as just one example of differences between the two.
Reply
Reply
post #114 of 916
11/15/12 at 11:52am
Thread Starter 
Quote:
Originally Posted by assassin View Post

But somehow wd black drives don't get the same bad rap despite also parking heads. So they somehow get a pass?

I believe I'm mistaken on the black drive and mispoke.

WD listed two models in the support email and one was RED and the other was a model number I assumed to be black. But in hindsight I think it might have been one if the av gpr models or something.

I don't get my secure work server email on my iPhone so ill have to wait till I'm home to check it.

The difference is perhaps that black drives have a 5 year warranty, and are rated at double the cycles if a green drive - in addition to a longer delay on head parking limiting much more the issue.

I'm just guessing at all this ...
Reply
Reply
post #115 of 916
11/15/12 at 11:55am
Thread Starter 
Quote:
Originally Posted by assassin View Post

But somehow wd black drives don't get the same bad rap despite also parking heads. So they somehow get a pass?

I don't own any black drives and generally think they are over priced and over rated.

I guess I'm just not concerned about them either way.
Reply
Reply
post #116 of 916
11/15/12 at 11:58am
Thread Starter 
And last- while I'm not giving up on WD I at least favor other brands equally so I guess what I'm really looking to find out is what us a good low cost and moderate energy HDD for a 24/7 software Flexraid server.

If all things being equal -I'd prob continue to pass on WD green for another just to be safe.
Reply
Reply
post #117 of 916
11/15/12 at 1:16pm
Quote:
Originally Posted by Mfusick View Post

I don't own any black drives and generally think they are over priced and over rated.
I guess I'm just not concerned about them either way.
Quote:
Originally Posted by Mfusick View Post

And last- while I'm not giving up on WD I at least favor other brands equally so I guess what I'm really looking to find out is what us a good low cost and moderate energy HDD for a 24/7 software Flexraid server.
If all things being equal -I'd prob continue to pass on WD green for another just to be safe.

You are missing the point. The point is that the black drives (which many assume to be just fine for RAID, servers etc) have the same head parking issue as the Greens. Yet many people think that it "better" to use a black drive for some reason and that Green drives should be avoided. So this theory/myth gets perpetuated despite the black drives having the exact same issue. If this really was an issue, imo, then people would also be shying away from Black drives.

Its possible that Green drives have higher failure rates although there is absolutely no data to say that's its because of head parking. It could be cheaper production, less tolerance in design, cheaper parts, etc. But, again, I haven't seen any data to support this either. It could also be that the people buying Green drives are cheap and are more likely to post a bad review. I have had extremely good success with Green drives and have never posted a Newegg review of any kind --- either good or bad.

Sounds like the perfect excuse for WD to push their $$$ expensive enterprise drives or better yet come out with a completely new line of "server/nas" drives call Red. Again, I would love to see some data. Otherwise I think you are shying away from Green drives for no real reason at all.

Funny coming from the OCZ guy.
Reply
Reply
post #118 of 916
11/15/12 at 5:37pm
Thread Starter 
I get where your coming from. I do.

I'm just looking for a little more self assurance before I buy in 100% as I've had bad luck with Green drives so far.

I love the power saving -- and intellipark is a big part if that.

But for 2$ month in energy I'd much rather have a drive that lasts longer and performs better.

The performance issue isn't enough for me to choose a 7200tpm drive over a green drive. But the reliability issue added into the mix probably is.
Reply
Reply
post #119 of 916
11/15/12 at 5:52pm
Quote:
Originally Posted by Mfusick View Post

I get where your coming from. I do.
I'm just looking for a little more self assurance before I buy in 100% as I've had bad luck with Green drives so far.
I love the power saving -- and intellipark is a big part if that.
But for 2$ month in energy I'd much rather have a drive that lasts longer and performs better.
The performance issue isn't enough for me to choose a 7200tpm drive over a green drive. But the reliability issue added into the mix probably is.

Again, please post data that shows that green drives don't last as long and aren't as reliable. I have asked repeatedly for this and you keep making statements like these with absolutely nothing to back it up. I have shown you what is available that points to data that refutes your claim.

Again, MFusick post the data. Any data supporting things that you say. Otherwise you have had some failed drives and I have had almost none and my sample size is much much much larger than yours.

More data: http://www.behardware.com/articles/810-6/taux-pannes-composants.html
Edited by assassin - 11/15/12 at 6:02pm
Reply
Reply
post #120 of 916
11/15/12 at 7:06pm
Thread Starter 
I'm actually asking you and others to point out its ok to use a green drive in a 24/7 server with WHS2011 and Flexraid.

I want to believe it is ok. I'd like to use them without second guessing if I am limiting the lifespan.

I'm not sure why your trying to argue with me. You seem almost hostile and I'm only asking.

But I do appreciate your reply and I do want to believe you but you haven't convinced me yet.
Your data is old and suspect. It's far from case closed.
Reply
Reply
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Home Theater Computers

AVS Top Picks

AVS › AVS Forum › Video Components › Home Theater Computers › Planning to rebuild my 20TB WHS Flexraid server: Information requested.