|
BoyBlunder posted:Just picked up 2x 3TB WD Reds at $109.99 a piece! Went back to grab the link for you guys, and it was back to $132 Back at 109! http://www.amazon.com/WD-Red-NAS-Hard-Drive/dp/B008JJLW4M/ref=sr_1_1?ie=UTF8&qid=1417155705&sr=8-1&keywords=wd+red
|
#
?
Nov 28, 2014 07:21
|
|
|
|
| # ? May 31, 2024 15:41 |
|
|
semicolonsrock posted:So if we have an old computer and put FreeNAS on it, and then attach external hard drives to it, will those be read/write accessible? Just want to make sure that this makes sense. Yes, why wouldn't they be? What does the position (internal/external) of the hdd have to do with being able to read/write on it?
|
|
#
?
Nov 28, 2014 16:21
|
|
|
Alright, round two. I'm thinking of grabbing a Qnap TS-212P and a pair of those WD Red 3TB drives, along with an individual Crashplan subscription. Is there any reason to get the 212E instead of the 212P? It looks to be $30 cheaper in exchange for cutting the RAM down to 256MB which strikes me as a bad tradeoff. Realistically 3 TB is overkill and I'd settle for 2 normally but 2 TB Reds are priced only $15 cheaper so again, not the tradeoff I want to make for $30 all told. Anyone with experience with Qnap apps and the Crashplan app in particular? Or is it easier to just setup the drive as shared and run CrashPlan from my Windows machine? e: Forgot to mention as folks might be interest, Crashplan has a BF discount for new users here.
|
|
#
?
Nov 28, 2014 17:18
|
|
|
I just purchased a Synology 415+ for watching movies and shoving Steam games onto. I bought three 3TB WD Reds to go along with it. I'll make a post on my thoughts when I get it Monday.
|
|
#
?
Nov 28, 2014 21:39
|
|
|
Those WD 3TB drives are back up in price on Amazon's site, but there's a Black Friday deal of some sort about to start on the 4 TB ones. I'll update this post with the price when it goes live in 10 minutes. e: Knocked another $10 or so off, down to $154.99 for a 4TB drive. They are flying though, 20% claimed already. Mo_Steel fucked around with this message at 00:41 on Nov 29, 2014 |
|
#
?
Nov 29, 2014 00:29
|
|
|
Anyone have any opinions on Seagate NAS 3TB/4TB drives? http://www.amazon.com/dp/B00D1GYNU8/ref=psdc_1254762011_t2_B008JJLW4M How come they are cheaper but no one talks about them?
|
|
#
?
Nov 29, 2014 03:36
|
|
|
I am selling 2x 2TB WD Reds on SA-Mart right now if anyone's interested! http://forums.somethingawful.com/showthread.php?threadid=3684596
|
|
#
?
Nov 29, 2014 03:58
|
|
|
Strong Sauce posted:Anyone have any opinions on Seagate NAS 3TB/4TB drives? http://www.amazon.com/dp/B00D1GYNU8/ref=psdc_1254762011_t2_B008JJLW4M How come they are cheaper but no one talks about them?
|
|
#
?
Nov 29, 2014 14:01
|
|
|
Josh Lyman posted:The Backblaze tests have turned off internet nerds to Seagate drives, despite being a flawed test. WD Red or die. How was it a flawed test? Did they end up using SMART data or something to pull or flag bad drives ahead of time?
|
|
#
?
Nov 29, 2014 14:32
|
|
|
Methylethylaldehyde posted:How was it a flawed test? Did they end up using SMART data or something to pull or flag bad drives ahead of time? Almost everything backblaze presents is wrong for a variety of reasons. When I look at the data, there are two big points that stand out to me. First, backblaze freely admits that it would buy external HDD, crack them open, and stick them in their systems. External HDDs are the bottom of the barrel, often reworked drives, that are intended to be on for maybe a couple hours a week. Second, they run these consumer/client HDDs in an enterprise environment, where these drives are subject to enterprise workloads and are powered on 24/7. Lets look at ST1500DL003, the Seagate Barracuda with the supposed 120% AFR, according to backblaze. The spec sheet clearly states that it can provide 700k MTBF @ 2400 power on hours per year. No poo poo these are going to fail at a spectacular rate when run outside their spec. And lets not forget we have no idea of the vibration characteristics of their storage pods. My guess is that the vibration is so high that is is shaking the drives to bits.
|
|
#
?
Nov 29, 2014 15:57
|
|
|
watwat posted:Almost everything backblaze presents is wrong for a variety of reasons. When I look at the data, there are two big points that stand out to me. First, backblaze freely admits that it would buy external HDD, crack them open, and stick them in their systems. External HDDs are the bottom of the barrel, often reworked drives, that are intended to be on for maybe a couple hours a week. Second, they run these consumer/client HDDs in an enterprise environment, where these drives are subject to enterprise workloads and are powered on 24/7. Lets look at ST1500DL003, the Seagate Barracuda with the supposed 120% AFR, according to backblaze. The spec sheet clearly states that it can provide 700k MTBF @ 2400 power on hours per year. No poo poo these are going to fail at a spectacular rate when run outside their spec. And lets not forget we have no idea of the vibration characteristics of their storage pods. My guess is that the vibration is so high that is is shaking the drives to bits. These might be valid points if they weren't doing the same thing for WD drives.
|
|
#
?
Nov 29, 2014 16:09
|
|
|
DNova posted:These might be valid points if they weren't doing the same thing for WD drives. You can't say that though. They don't specify which drives and at what quantity were farmed vs which ones were bought raw. I realize that somewhat contradicts my original argument, but that's the point. Backblaze provides this data where the sourcing is unknown and the quantities are unknown. What we do know is that client drives are running in an enterprise environment. If you want to use that is a point that HGST and WD are more robust than Seagate drives, you can but you would be wrong. Drives are tailored to their intended environment. Got instance, the inclusion of RV sensors (that the HGST drives have, and the Seagate drives do not) makes a massive difference.
|
|
#
?
Nov 29, 2014 16:26
|
|
|
watwat posted:You can't say that though. They don't specify which drives and at what quantity were farmed vs which ones were bought raw. I realize that somewhat contradicts my original argument, Yeah it does, entirely. I'm not saying to take their reports very seriously, though.
|
|
#
?
Nov 29, 2014 16:28
|
|
|
DNova posted:These might be valid points if they weren't doing the same thing for WD drives. Can you even get a WD Red that isn't a bare drive? I won't write it off as horribly broken, but it's not really designed to be an actual test case. I stay away from Seagate but it's because their RMA process is more painful than WD's, and they're the only ones where I've had to RMA the 'same' drive three times because they sent me two busted ones in a row.
|
|
#
?
Nov 29, 2014 16:30
|
|
|
It's less Backblaze misleading people and more people pulling the numbers out of context and reading way too much in them. Which is tempting, I'll admit. The title of the original blog post, "What hard drive should I buy?", is misleading, sure. But read the articles and you'll find they're very careful to give you no information on that topic whatsoever. It doesn't matter if they think cracking open externals is viable business to them if you stop looking at this as a "test" and stop assuming that reading those numbers will/should lead to a useful conclusion for you.
|
|
#
?
Nov 29, 2014 16:30
|
|
|
watwat posted:Almost everything backblaze presents is wrong for a variety of reasons. When I look at the data, there are two big points that stand out to me. First, backblaze freely admits that it would buy external HDD, crack them open, and stick them in their systems. External HDDs are the bottom of the barrel, often reworked drives, that are intended to be on for maybe a couple hours a week. Second, they run these consumer/client HDDs in an enterprise environment, where these drives are subject to enterprise workloads and are powered on 24/7. Lets look at ST1500DL003, the Seagate Barracuda with the supposed 120% AFR, according to backblaze. The spec sheet clearly states that it can provide 700k MTBF @ 2400 power on hours per year. No poo poo these are going to fail at a spectacular rate when run outside their spec. And lets not forget we have no idea of the vibration characteristics of their storage pods. My guess is that the vibration is so high that is is shaking the drives to bits. When you're testing how long drives last in environment X, the fact that the drive was intended for environment Y doesn't mean anything other than as an explanation of why the drive performed as it did. The drive still performed as the test described in environment X. If you are looking for drives for an environment similar to environment X, this is useful information.
|
|
#
?
Nov 29, 2014 16:32
|
|
|
IOwnCalculus posted:Can you even get a WD Red that isn't a bare drive? I won't write it off as horribly broken, but it's not really designed to be an actual test case. "The same thing" meaning using drives outside of their intended environments. They aren't only using Reds, which you could argue are intended for such use. They're also using Greens, for example.
|
|
#
?
Nov 29, 2014 16:32
|
|
|
Thermopyle posted:When you're testing how long drives last in environment X, the fact that the drive was intended for environment Y doesn't mean anything other than as an explanation of why the drive performed as it did. Valid point. My gripe is that the internet has seemingly taken this data as the definitive "which drive is better no matter the use case." Naturally, HGST will say "look at how good our drives are backblaze says so" and Seagate will say "backblaze is crap and here is why" which further feeds the public misconceptions. All drive suppliers meet their MTBF specifications when run in their intended environments.
|
|
#
?
Nov 29, 2014 17:16
|
|
|
watwat posted:Valid point. My gripe is that the internet has seemingly taken this data as the definitive "which drive is better no matter the use case." Naturally, HGST will say "look at how good our drives are backblaze says so" and Seagate will say "backblaze is crap and here is why" which further feeds the public misconceptions. Yeah. The internet is good at taking stuff and running it into the ground. watwat posted:All drive suppliers meet their MTBF specifications when run in their intended environments. Do we know this?
|
|
#
?
Nov 29, 2014 17:29
|
|
|
Thermopyle posted:Do we know this?
|
|
#
?
Nov 29, 2014 18:26
|
|
|
Thermopyle posted:Do we know this? Yes* I work in HDD QA at a storage vendor, so I have to be careful about what I say, but yes they do make their MTBF numbers. The caveat here is that there are games the suppliers play to make that number. MTBF is usually derived from a short term test on a small population of drives. What gets missed are the so-called "latent" failures, which aren't necessarily wear out failures but are inherent flaws in material or firmware, like lube that degrades much earlier than intended. Suppliers will typically discount these failure modes to show that they still meet their specs. What also isn't accounted for is drive workload. Our field data shows that temperature has less of an effect on failures, but workload makes a huge difference. Heavy drive workload will bring out these latent failures much much sooner. We do see drives that have come out of the Thailand flood period to be really piss poor, but beyond that all the suppliers have pretty much been at the same point reliability wise. Once we get into tech like HAMR, bit patterned media, SMR and tech like that, this will change.
|
|
#
?
Nov 29, 2014 18:50
|
|
|
So is the conclusion that we can put the cheapest 4TB NAS drives we can find in a home storage server, or that we should should stick with WD Reds. :P
|
|
#
?
Nov 29, 2014 20:11
|
|
|
Is there a rack that will accommodate this 2U SuperMicro server for less than $200? Looking for around 6U of space. The server is 26" deep and most racks I'm seeing are only about 24" deep or $500+.
|
|
#
?
Nov 29, 2014 23:17
|
|
|
ashgromnies posted:Is there a rack that will accommodate this 2U SuperMicro server for less than $200? Looking for around 6U of space. The server is 26" deep and most racks I'm seeing are only about 24" deep or $500+. People practically give them away on craigslist.
|
|
#
?
Nov 29, 2014 23:18
|
|
|
ashgromnies posted:Is there a rack that will accommodate this 2U SuperMicro server for less than $200? Looking for around 6U of space. The server is 26" deep and most racks I'm seeing are only about 24" deep or $500+. Doors not included but it's $150 ~ http://dallas.craigslist.org/dal/sop/4760182789.html 
|
|
#
?
Nov 29, 2014 23:38
|
|
|
Quixzlizx posted:So is the conclusion that we can put the cheapest 4TB NAS drives we can find in a home storage server, or that we should should stick with WD Reds. :P *stick with HGST drives *avoid Seagate
|
|
#
?
Nov 29, 2014 23:43
|
|
|
If i'm picking up 1-2 drives for home storage, and am considering both WD reds and HGST's, which ranks higher, and is the WD black warranty worth shelling out a few bucks for? basically i want to get a shitton of storage now and not worry about replacements or failures for the next few years.
|
|
#
?
Nov 30, 2014 01:20
|
|
|
HGST or Western Digital will both work well, just be careful to note the warranty and barring that being great, you can try your credit card company's extended warranty policies. Thing is, hard drives are expected to at least double in value within the next two years (which didn't happen due to the flooding in Thailand these past few years - they've just recovered to where they were during the Great Hard Drive Advancement Recession) and I suspect the drives you buy now will hardly hold up to the 8TB monsters for $170 in 2017. It seems far away, but that's when your warranty will have run out and your drives start failing.
|
|
#
?
Nov 30, 2014 01:38
|
|
|
Well, 100-200 spent to last four years or so is a pretty good deal. Its the in-between failures I'd want to mitigate. I just need to replace the current one that's starting to go, and worst case scenario I just get a backup drive for the enclosure.
|
|
#
?
Nov 30, 2014 01:42
|
|
|
I do love ZFS. Replacing some super-high-hour-count Samsung 1.5TB drives with fresh 3TB Reds, and one of them survived only one of the three resilvers needed (raidz1). ZFS was kind enough to tell me exactly which files were corrupted (maybe a couple hundred megs at most - only 40 photos so far) and is letting me resilver the array with another fresh drive in its place so that I can restore them from my backups when it's done.
|
|
#
?
Nov 30, 2014 04:51
|
|
|
TheParadigm posted:I just need to replace the current one that's starting to go, and worst case scenario I just get a backup drive for the enclosure. Don't lose sight of the fact that RAID 5/6/Z1/Z2/SHR/SHR2 is there for availability purposes. If you lose a drive (or 2 depending on your parity), you can still access your data. None of these are backups. Always maintain (preferably offsite) backups of data you care about no matter is is on a parity setup or not. No number of parity drives is going to help if someone breaks into your place and steals your computer, it is caught in a fire, etc. I use Crashplan. That being said I personally view Reds and HGST drives as about on par quality wise - I'd get whatever is cheaper if I were buying.
|
|
#
?
Nov 30, 2014 05:50
|
|
|
Don Lapre posted:People practically give them away on craigslist. Sadly not in Detroit  Would a music equipment rack without rails/a back/sides work okay? I feel like the tension might be too high on the joint but they manage to hold decently hefty effects processors and the like... Something like http://www.amazon.com/On-Stage-Stands-RS7030-OnStage-Stand/dp/B000CD1R84/ref=sr_1_2?ie=UTF8&qid=1417327647&sr=8-2&keywords=Music+rack if I mounted in the bottom spaces?
|
|
#
?
Nov 30, 2014 07:06
|
|
|
I'm building a Plex box and trying to figure out my OS options, specifically for storage purposes. The server is going to be an i3-4150 and running 3-4 WD Reds. I'd like a storage option with parity so I can sustain at least one failure at a time. I was originally going to use a Windows base, with storage pools but the write performance looks to be laughably bad for this. So I'm considering a few options: - Windows software RAID5. Should have stronger write performance from what I've seen. - FlexRAID. Seems pretty cool, but also money I don't want to spend. Wondering what my options are for a *nix OS? I need: - Plex and transcoding - SABnzbd/couchpotato/sickbeard - Torrent client - Some sort of parity storage option Past that, I don't really want to spend hours loving around because I'm getting old and grumpy and lack the endless hours to tinker I once had (hence original windows considerations). Any thoughts?
|
|
#
?
Nov 30, 2014 16:32
|
|
|
Thinking of getting one of these: QNAP HS-210 http://www.amazon.co.uk/QNAP-HS-210-Fanless-Stylish-Attached/dp/B00GZN4UCK#productDetails Is it dumb trying to get a silent NAS? It will either be under my bed or on the TV rack at the end of my bed... It will either have WD Greens or Reds depending on whats quieter.. I need to not be able to hear it all when I'm sleeping. It's going to be teamed up with a fanless i3 NUC that will be running 24/7 OpenElec with sabnzbd+sickbeard. Don't need a ton of space, it will be storing music, movies and TV but the movies and TV will be removed once I'm done watching them... the files will be extracted on the ssd in the NUC and then moved to the NAS and I'll need to access it all over the network on my Win7 PC, preferably with folder shells as well as in the browser and also while I'm outside my network. I'd put a 1TB ssd in there to make sure it's silent but I'm not sure if it will be enough space and it's going to drive the price way up. The NUC is already going to be costing me ~£400 if I go through with this then another £400 or so for this NAS+SSD. it's going to end up costing more than my main PC did. Is this a dumb idea or is there a better/cheaper alternative? I need to keep it as silent as possible (fan and mech. hdd vibrations) so it doesn't disturb my sleep and hopefully uses as little power as possible
|
|
#
?
Nov 30, 2014 17:33
|
|
|
Walked posted:I'm building a Plex box and trying to figure out my OS options, specifically for storage purposes.
|
|
#
?
Nov 30, 2014 17:55
|
|
|
uhhhhahhhhohahhh posted:Thinking of getting one of these: QNAP HS-210 http://www.amazon.co.uk/QNAP-HS-210-Fanless-Stylish-Attached/dp/B00GZN4UCK#productDetails Put it in a closet.
|
|
#
?
Nov 30, 2014 18:16
|
|
|
Don't know if I can run a cable in there and can't really be chopping holes in anything. It might also overheat in there??
|
|
#
?
Nov 30, 2014 19:34
|
|
|
Walked posted:- Windows software RAID5. Should have stronger write performance from what I've seen. Combat Pretzel fucked around with this message at 20:33 on Nov 30, 2014 |
|
#
?
Nov 30, 2014 20:31
|
|
|
Combat Pretzel posted:For the love of god, don't do this. The NTFS journalling doesn't communicate with the LVM, and whenever the box restarts unexpectedly, it starts to rebuild your whole array. I'd rather take the write hit of Storage Spaces, which journals what slabs are dirty (independent of the filesystem), than have my array poo poo the bed performance wise and add wear to the disks whenever the power grid is wonky during summer. This is the info I'm here in search of. Thanks! Probably go with Xpenology, I just want to be sure it's pretty self-sustaining with minimal maintenance.
|
|
#
?
Nov 30, 2014 20:43
|
|
|
|
| # ? May 31, 2024 15:41 |
|
|
I ran a Windows (disk management) mirror for a while and it was rebuilding all the loving time. I think going to sleep/hibernate was an 'unexpected' restart as far as it was concerned. Phases of the moon or time of day might also be enough to throw it off. Haven't had any issues with Storage Spaces but I'm not using it very heavily and it's all getting backed up to my NAS anyway.
|
|
#
?
Nov 30, 2014 20:49
|
|
