Popular Content

Showing content with the highest reputation since 10/12/13 in all areas

  1. 2 points
    Put them in a server and run some sort of SDS on top of it like Nexenta. Fun learning experience and gets you a cheap SAN.
  2. 2 points
    Note the three years in between though...they've been surprised by the interest in the platform I think. Now, if WD could just get those 2.5" Reds up to 2TB in a 9.5mm...
  3. 2 points
    Long-time watcher of StorageReview, but I registered just to be able to comment on this review. An excellent review, though your testing seems a bit high-end for the likely intended usage. I'd bet the majority of the target users for this SOHO device won't have a backbone that supports iSCSI or even dual-port aggregation. As you point out, 2-10 users in a casual / small office setting or for home use seems a likely audience. Such an audience would be much more likely to have an entry-level GbE switch as opposed to a managed backbone that costs 10x more. To that point, I've used the entire line of BlackArmor devices, and there are three critical issues common to them that seem to be repeated with the replacement Business Storage line... none of which are mentioned in the review, but they may not impact everyone so I'm not sure they necessarily bear mentioning up-front. 1) Performance. You obtained okay numbers in your testbed, but as summarized above, I doubt you'd see that infrastructure in the wild. I'd suggest you at least pair it with testing results from a cheapo GbE switch using a single LAN plug and simple Windows file sharing / disk mapping. Unless the BS line has markedly improved from the BA line, you'll see performance on the order of 15 MB/s read, 10 MB/s write. Horrendous for anything but backups, really, which is all I use my BA boxes for. Also, I recognize that there's a massive disparity of price points and target audience, but I get 110 MB/s--TEN TIMES the performance--from my Synology boxes, and 50-70 MB/s from my Drobos. And that's on a cost-conscious backbone of entry-level GbE switches using one LAN port per device and simple, iSCSI-less file sharing in Windows. There's no comparison at all. 2) Compatibility. Massively overpriced with disks, the BA and BS line are very reasonable when purchased diskless. I've used Buffalo, Seagate, Synology, and Drobo NAS boxes in small-business and personal settings, and diskless BA/BS boxes are far and away the cheapest way really of adding reliable (but not fast!) NAS storage in such contexts. But these NAS boxes only support Seagate disks. True, this is a Seagate device, but it seems as though someone had to intentionally code a rejection routine into the firmware, which is just kind of an obnoxious move. In addition, some of the compatibility notes for "certified drives" listed for the BA line are flat-out falsified--the diskless BA 400 will simply NOT work with the 1.5 TB desktop line of Seagate disks, period. 3) Risk. For those who know what they're doing, these are fairly easy boxes to deploy, and the web-based UI is second only to Synology's in my experience. But it's easy, far too easy, to make a catastrophic mistake. For example, if you set up a BA box using one LAN port, and then try to plug in a second LAN plug, it will not only not work, but it has a strong chance of corrupting the entire array, forcing you to not only lose all data and set everything up again, but in order to even begin to do so, you must eject each disk individually and reformat it using a separate computer. Otherwise it won't set itself up. Now, much of my comments above are from my experience with the older BA boxes, but I'd like to know if those issues have been resolved with the replacement BS line. Anyway, as always, I love seeing info on Storage Review.com so keep up the good work!
  4. 2 points
    You're almost right here. What's missing is that copying small files, even from the same directory, will automatically include some random access too. The files being read may be spread across the disk, they may be written different locations, filling up holes in the current file structure (what ever the OS see fit) and the MFT may be accessed. That's why multi-threaded copy for higher queue depths still improves throughput: the disk can arrange the accesses better through NCQ and can reduce access times. BTW: if the folders you're copying are often the same I'd look into incremental sync'ing with e.g. DirSync (nice freeware). Not sure it can increase QD, but it certainly saves time not to transfer untouched files again. And I'm not a fan of buying large SSDs for storage, that's often a waste of money (IMO). I'd rather use the SSD for temperary storage and as automatic cache. If you're concerned with many small files an SSD would be ideal. And if the SSD cache also buffers writes you may see a massive speed increase. The cache capacity would also be huge compared to the amount of storage required for small files MrS
  5. 1 point
    No, spindle speed and other drive architecture elements benefit you more than on drive cache. They don't aggregate like an SSD cache in front of your HDD storage.
  6. 1 point
    Client or enterprise? Either way, there's not really a technical reason why we don't have huge SSDs, it's more of a cost issue and what the market wants to buy.
  7. 1 point
    Seagate has all the information you could want on their site. You can come back here to read reviews of most of the products you listed.
  8. 1 point
    To be fair they're not claiming this would ship in mass quantities to end user stores. I concur with Brian in that it's probably SMR and by "support for archive workloads" they actually mean "don't try to use this for anything else than archive workloads". By "consistent enterprise-class performance" they probably mean "it's always slow". If it was any better than this they surely would have said so. That's not to say the drive wouldn't have its place.. but I think it may actually be good to limit their availability for now. BTW: it's no excuse, but Seagate has been soft-launchnig drives for years. Sometime retail availability took a full year! I see no reason to get any more tired of this than in the past years. MrS
  9. 1 point
    Sadly, reinstall. It's a massive pain if you have a lot of apps, but the end result is worthwhile.
  10. 1 point
    As Brian said, you're asking a rather broad question. What exactly are you asking with regards to performance issues? Did you have an application or actual proposed use in mind in sufficient detail to make this a question we can helpfully answer, or at least try to? Have you talked to any of the relevant vendors in this space-- newcomers like Tegile, Nimble, Tintri, Nutanix, etc.? Or established ones, such as Equallogic, EMC, Netapp, Fusion-io, Is commentary such as this sufficient...? http://www.techrepublic.com/blog/data-center/the-brewing-battle-between-hybrid-and-all-flash-storage/ The harddisks and the SSD's are usually organized separately when it comes to the actual hardware disk sets, it's not a straight RAID10, if that's what you're asking. Now how the firmware manages the various controllers involved with the harddisk arrays and the SSD arrays is the tricky part...
  11. 1 point
  12. 1 point
    Samsung is #1 in SSD drive shipments. I'd love to be crap if that's what it looks like, lol.
  13. 1 point
    Specifically, they have little ability to innovate going forward and I question their ability to build a proper support network. Note that the opinion is largely based on a US-centric slant, some brands have better adoption in Europe or Asia for instance. The decision to rebrand the M550 is a pretty clear case. They had no ability to create a new product with the delays in SandForce, so the best option was to copy Micron and hope to make money by selling a slower version for $10 less? Poor business sense and not something I'd invest my money in as an SSD buyer.
  14. 1 point
    I think eventually geometry will be layered so tight together sectors will suffer magnetism bleed. 0s next to 1s will flip, etc..... Arms and pivot points will be tight and accurate for only so long. Electric motors will get sloppy stepping after a while... Sent from my rooted HTC Supersonic using Tapatalk 2 Pro
  15. 1 point
    I'm pretty indifferent on specific drives, in such a small sample size it's hard to get too excited one way or another. Personal preference would be HGST Deskstar, but people find good success with Seagate NAS drives too, even though this isn't a NAS. You could get a Samsung EVO at under .50GB for the system drive and then just get a platter drive for your media too...another angle.
  16. 1 point
    When you swap your board you need to reactivate your Windows 7, I did this in January. My used copy was a System Builder edition you can buy in Germany without a PC. I just did a phone activation and everything worked fine.
  17. 1 point
    The two drives can be separated with standard sata connectors. We have an exploded image in the review and more shots in the forums thread. I was told today the MSRP is $129, but I don't see it anywhere in retail.
  18. 1 point
    I think they bumped it up recently... both are at the 1M value. Also if that HGST model is the best price for that capacity, I wouldn't go for anything else.... you really can't beat that performance or reliability for better price.
  19. 1 point
    I'd reach out to them. I'm not sure of the refurb process but it could just entail a quick drive check and wipe.
  20. 1 point
    Another vote for the 840 Evo. It's got a very well balanced controller and makes clever use of that TLC NAND. In the real world you don't have to worry about its write cycles. And you're not going to fill it with 4k random data repeatedly, since you said you're going to place OS and programs there. The price premium for that Plextor is insane IMO. I didn't look up the Kingston V300, but the 840 Evo likely performs better enough to warrent the price premium over the V300. MrS
  21. 1 point
    I like to test my desktop drives on Linux, because most vendors ignore it and target AS SSD, DiskMark, HD Tune and other Windows stuff. Running a find(1) with a cold buffer cache gives a good impression about random read access times. Likewise you can mount something -o sync, extract uncompressed tar archives and delete huge directory trees to get a peek on random write performance. If a drive looks bad there, it may even be the case, that the vendor wants it to look bad, so they can sell expensive enterprise drives with a slightly different firmware instead. The 8 GB flash caches take the issue to next level. While having almost no effect in practice, because they are to small. 8 GB is just big enough, to make drive look good in benchmarks against SSDs and impress with mostly irrelevant boot times. And that is what sells them. I'm accusing no-one here, but I'm very skeptical and suspicious. If something looks like a engineering marvel out of nowhere, it most likely isn't true. Remember SoftRAM 95?
  22. 1 point
    Also it may be cheaper to buy two UPS's (one for monitor, one for computer) than one big one. If you are borderline your existing UPS where removing the 50W of load of the monitor will make it work, then a second cheap UPS may be better than trying to buy an extremely expensive larger one.
  23. 1 point
    Fair point. We're working on next gen testing for these things and will take that into consideration. For now though at least the results are relative.
  24. 1 point
    I can see clearly in the HDD forum a thread with RE4 in it yet when i search your search engine says this No results found for 're4'. eh?
  25. 1 point