• Content count

  • Joined

  • Last visited

Community Reputation

0 Neutral

About captainmidnight

  • Rank

Contact Methods

  • ICQ
  1. continuum and Brian: thanks much for your replies. (Sorry for the delay--while travelling this weekend, my laptop screen died...) I think that I am going to order the WD My Passport Ultra 2TB.
  2. I need a portable drive that I use for off site backup. In particular, once a month I will fetch this drive out of my bank's safe deposit box, take it home, back up a bulk internal drive (old WD caviar black that is still going strong...), and then return it to the safe deposit box. I need these features for sure: since my internal drive is 2 TB, I need this backup drive to be 2+ TB I am sick of slow backup times with my existing USB 2.0 WD portable drive, so USB 3.0 is a must After those requirements, in principle, I next value drive reliablity. (I say in principle, knowing full well that it is really hard to judge the reliability of a given line of hard drives even if you know the manufacturer). I next value drive performance. If flash drives were cheaper, I would get one, but I think for this application I am stuck with a hard drive. I really want a portable USB 3 hard drive with > 100 MB / s sustained speed across most of the drive. My backup will have a mix of file sizes to handle, but should be mostly sequential. Other drive features, like ruggedness, slimness, and looks are icing on the cake. My preliminary research has found a couple of possibilities: WD My Passport Ultra 2TB Seagate Backup Plus Slim 2TB Toshiba 2TB Canvio Basics Silicon Power 2TB Rugged Armor A80 But please open my mind to any drive that you are keen on. Most of the drives above are > 1 year old at this point, I think. Are there any new models in this segment that excite you guys? Anything from CES 2015 that I misssed? (For various resons, the announcements here, here, and here will not work for me.) I have found it really hard to find any good, recent (< 6 months old), survey reviews for this drive segment. A slight exception is The Wirecutter's recommendation, which discusses some of the drives above. Here too, if you know of any good survey reviews, please inform me. I am especially interested in reviews which include drive performance benchmarks.
  3. True for the former, not necessarily true for the latter if your RAID controller is smart enough to interleave reads on both disks, which most low-end controllers do not. And would it be a great bet that the Intel ESB2 SATA 3.0Gbps Controller is in the low-end category, with crappy RAID1 performance? I could not find anything at Intel concerning this controller that directly answers my question; some related links are What do you mean by "and take the performance hit"? The hit of using a single drive instead of a true RAID solution?
  4. Below are some links that are relevant to this discussion: 1) to get good write performance: "A 'cheat' to get better write performance is to enable the write back function, as opposed to write through. The system writes to the controller memory, the controller issues the write complete signal and deals with calculations and then the actual writes to the array." 2) this guy measured 3 disk raid5 to be almost as good as 2 disk raid0: 3) a thorough comparison: 4) warnings about raid5 writes: "Raid 5 has a very bad write performance if your OS issue non stripe size multiple IO because it has to read the old parity block of your stripe, use it to compute the new parity block, and write the stripe if you got a large "write back cache backed by battery" on your controller, you may lower the impact of this raid 5 write performance problem by aggregating IO from the cache but, in a multi-user usage, your controller cache can not guess if you are issuing sequential or direct io...and will issue many read-xor-write IO"
  5. I am going to building a new custom workstation based on this Supermicro barebones system: My system needs to be high performance. My main tasks (related to stock data processing) stress both the CPU (I will be sticking at least one, maybe two, xeon 5420's in this box) as well as mass storage (I routinely process ~100GB spread among, say, 5,000 or so files--so large file sequential reads/writes are what I mainly care about). Concerning mass storage, it seems that solid state drives are just not cost effective at this point. Plus I am too nervous to trust brand new technologies with my precious data. So I am left with plain old hard drives. I am thinking of the new single platter 320 GB WD drive (WD3200AAKS);Tpk=WD3200AAKS as it is almost as fast as the best SATA drives out there (except maybe for small random file operations; but all hard drives suck for this anyways). Please recommend a drive that you like better than this, however. I was originally going to buy 2 of these and configure them in RAID0 to get max large file performance. I do a daily backup of all my data, so this is not as dangerous as it sounds. Nevertheless, if a drive was to fail, it would be a pain to have to reinstall everything. So I am open to considering higher RAID levels, which I need your advice on. I think that RAID 1 is out because it gives me zero performance improvement. My understanding is that RAID1 not only performs (slightly) worse than a single drive for writes, it also performs no better than a single drive for reads. (This assumes that there is but a single read thread, which will typically be the case for me. However, if you have multiple read threads, such as in typical server applications, then RAID1 can actually have almost as good read performance as RAID0 because any decent controller will concurrently read different files from the individual drives, since they do not data integrity checking on reads.) Is this understanding correct? In theory, RAID5 looks very interesting. If a 3-disk RAID5 setup could get 90+% of both the read and write performance of a 2-disk RAID0 setup, then I would gladly pay 50% more for the extra drive. Unfortunately, the web research that I have done one RAID5 is highly confusing. I have come across completely contradictory reports, some claiming RAID5 performance being close to RAID0, others denying it, and others warning about it being highly dependent on your raid controller. Anyone know of some good modern benchmarks (for my box, or at least the controller that I describe below, would be superb). By the way, the motherboard on that Supermicro box uses a "Intel ESB2 SATA 3.0Gbps Controller". Questions: --is this actually a "fake hardware raid" controller? --anyone know how well it is regarded? In particular, is it well suited for RAID5 or pathetic? --while I will (unfortunately) initially be running windoze xp on it, I ultimately want to run linux (probably ubuntu) on it; will that be a problem? The supermicro link above claims that "RAID 0, 1, 5, 10 support (Windows only)" If this RAID stuff is ultimately impossible, then I may reluctantly just buy a single velociraptor (which is an utter ripoff at $300; if it was $150 i would not complain).
  6. If low-level tests are all that matter, why do sites struggle to devise meaningful high-level tests? I understand that low level results may be deceiving and that true high level results are to be preferred; that is an excellent analogy with procesor clock speeds, by the way. I should have cited this page from tomshardware review which focuses on what they call "Application" benchmarks (which I assume correspond with your high level tests). Notice how the WD4000KD is strictly in the middle of the pack in their first test (XP startup) and several notches down on the second (FileWrite). Actually, now that I am looking more closely at this, they seem to not have done that thorough of a job doing high level tests, so maybe their results are not too meaninful. Please put it back!
  7. Per the SR FAQ here, it's not a trivial matter to do so. Just thinking about all the design issues is making me cringe. 216998[/snapback] Reading that FAQ was interesting, and did not make me cringe at all. Of course, I say this as someone who has never worked for a disk drive manufacturer either! I am sure that the economics (the manufacturers want to force you to buy more drives, plus they are a low margin commodity industry with little luxury for technological innovation) are what are are supressing the vastly superior performance that you could get from internal RAID (e.g. n times data throughput where n is the number of platters). Still, as the FAQ itself notes, all of the technical issues can be addressed by straightforward use of current technology: Factor #1 (Head offset on a single actuator) is solvable simply by using dedicated actuators for each platter. (The probably more optimal solution of having multi-stage actuators is what would require substantial technology development; this could wait for the future). Factors #2 (Read channel bandwidth limitations) and #3 (Servo complexity) are likewise simply solved by having dedicated read channels and servo systems for each platter. (Again, the probably more optimal solution of having a single but more sophisticated channel chip and servo system is what would require substantial technology development; this too could wait for the future). Factor #4 (Interface bandwidth limitations) is a marginal issue with SATA 150 and is not an issue at all with SATA 300 which is now already a production technology. Even if SATA 150 is used, heck, I would be GLAD to have a drive which is saturating that interface compared to what we currently have! Note also that unlike SCSI, SATA is a point-point (not shared bus) architecture so that if you made a RAID of drives each of which is also an internal drive, then it should scale linearly. (At this point, you would need to worry about the data throughput of the RAID controller.) So, you can achieve internal RAID right now with very little new technology having to be developed. Yes, the solution of parallel everything (actuators, read channels, servos) is sub-optimal compared to what developing new custom technology might do, but it should still work perfectly fine as far as performance is concerned. Then the remaining question to my mind is what IS the actual cost of a hard drive's components? How much does each actuator cost? Each read channel chip? Each servo system? For the sake of argument, lets suppose that each of those 3 system costs $10. So, in a conventional drive, they total $30. Lets consider a prime candidate for an application of this technogy, the 74 GB Raptor which has 2 platters. Adding dedicated components to each drive just increased the component cost by an extra $30. Lets suppose the price increase that needs to be passed onto the consumer is double that, $60. Would you be willing to pay $60 extra for a Raptor with twice the data throughput (~140 MB/s)? I would! Or consider the 4 platter WD4000KD: if each platter had dedicated components and the total component cost is an extra (4 - 1) * $30 = $90 so the extra cost to the consumer is 2 * $90 = $180. Again, I would personally pay $180 extra (~$410 total if the YR's $228 price at Newegg is used, in order to get a drive that would offer me ~4 * 70 = 280 MB/s data throughput. Clearly, my argument above made lots of assumptions about component costs and what they lead to as final customer price increases, as well as on price/performance tradeoffs for customers. If you have better data, please post it. Also, if anyone who actually works for a drive manufacturer is reading, I would love your perspective.
  8. Eugene, Thanks for the prompt response: the benchmarks you posted have convinced me to buy the Western Digital Caviar SE16 WD4000KD. Did you just recently generate that KD versus YR webpage? I did a search earlier today on SR for the WD4000KD and did not find that page. Instead, I mostly found forum discussions that did not seem relevant for the questions that I had. The reason why I was worried about the SE16's performance, by the way, was because SR's review concludes that the YR is about one of the fastest SATA drives out there ("sweeps our high-level performance tests across the board") whereas other reviews that I read on the KD conclude that it is mediocre: "The WD4000 compares favorably to other WD drives, but remains behind its current competitor from Seagate (7200.8) when it comes to low level performance (except for the access time test.)" "A while ago we had taken a look at a couple 400GB hard drives from Hitachi and Seagate, and today we got to test run the Western Digital 400GB hard drive, WD4000KD. Although it doesn't support such technology like Native Command Queuing, it proved to keep up with the NCQ enabled Seagate drive. And as well managed to prove itself against the Hitachi drive also. It manages to easily beat out both other drives in the burst rate benchmarks. But as for the other HD Tach benchmarks it looks like there is no clear winner. The rest of the benchmarks also show impressive numbers despite the fact of no direct comparison to any other drive. If you happen to know your numbers well with these tests, you will know that the numbers compete with some of the best desktop drives currently available." [Note that this guy claims that it does not support NCQ, whereas your webpage has actual test results with it with NCQ enabled--did he just have some early eval version of the drive, or did he totally blow his review?] Personally, I would still buy the YR for its better reliability if it was not for its TLER feature. Oh, one more thing: I asked the MA salesman for WD why they have not revved the Raptor line in the last 1.5 years--it seems to me that it is now getting long in tooth, and is sorely overdue for 15Krpm speed etc. He said that their plans for the Raptor are to next introduce a 150 GB version of the drive around 2006/Feb, as their customers are demanding capacity increases more than rotational speed increases, altho 15 Krpm speeds are planned in future versions after that. I wish that they would be even more aggressive and do things like have each platter have an independent head (so that you would have intra-drive striping); I am amazed that no manufacturer currently does this.
  9. Does this apply if the drive is used on a RAID capable controller (nForce4), but not utilized in a RAID configuration? I would like to use one of these drives as a storage disk in a graphics workstation (Tyan K8E, Athlon X2 4400+), but I don't want to risk data corruption/integrity. Thanks. 216747[/snapback] I would like to know this too! I'm just dying to use this drive in a desktop based system, but NOT in a raid array! I wish there was some way to just manually disable this TLER feature! 216756[/snapback] Guys, I too am desperate to know if this drive can be used as a non-RAID drive, in spite of WD's warnings about the Time-Limited Error Recovery (TLER) making it non-suitable for ordinary single-drive setups. I actually called WD today to try and get an answer. It was really tough to find sales offices on their wesite; they really seem to try and hide the information. Anyways, I called their MA and CA sales offices and asked to speak to sales engineers. The MA office routed me to an ordinary saleman. He claimed that he did not think that the TLER feature would rule out this drive as a desktop single drive, which had me excited. The CA office routed me to a lady in their tech support group. She said that TLER -is- bad for a desktop drive because it may cause lots of spurious file not found or other error messages to appear. She said that she does not believe that TLER is configurable in these units. If true, what an oversight on their engineer's part. Their jumper settings for this drive certainly mention nothing about TLER: I also asked this lady if it is OK to use this drive in a non-RAID setup but if it is connected to a motherboard which has a RAID controller on it. (In my case, I have an Asus P4800 Deluxe motherboard; if I understand the specs correctly, SATA RAID on this MB is controlled by the ICH5R South Bridge.) She said that she was not sure, but did not think that it would work. If I can't get a satisfactory answer, I will have to fall back and order their consumer grade drive, the Western Digital Caviar SE16 WD4000KD, which I will probbly buy at Newegg (unless you guys have another recommendation for a better store or a better high performance and >= 400 GB drive--feel free to chime in.) By the way, here are some background references on TLER: (see the FAQ near the end; claims that normal drives can take up to several minutes performing error recovery) (see near the middle where they talk about Behavior in RAID system)