sechs

Member
  • Content count

    1516
  • Joined

  • Last visited

Community Reputation

0 Neutral

About sechs

  • Rank
    Member

Contact Methods

  • ICQ
    0
  1. Considering that the original cRaptor was parallel ATA drive with a bridge chip slapped onto it, I'd have to go with the latter....
  2. A shame? The reason that Western Digital left the market is because no one would buy their drives. The reason why? They were bad -- I mean B-A-D. Poor design and unreliable. You act as if there's separate research and development for enterpise drives like SCSI, and workstation and consumer drives like ATA. There isn't. The only real difference there is interfaces -- and that work is really done by a few companies that do the controller chips. And what's more, the drive manufacturers cross-license their technologies all over the place. So, any technology that is developed by one company can work its way into any drive by any manufacturer. Margins in the SCSI market are very good. But Western Digital doesn't know how to execute well, and has made its niche in selling crappy drives to folks who don't know any better. If you think that Western Digital is squeezed, it's because they have poor quality control and, at least, until recently, have been total technology laggards. I won't use their drives, and there are a lot folks who have a lot of drives run through their hands that won't either. That's the mountain that Western Digital has to climb.
  3. The reason that there isn't a bigger 10k SATA drive is because Western Digital can't figure out how to put the third platter in there. So, they have to wait for the density to bump.... and they generally lag in that department. That said, nobody worth their salt is going to put a Raptor into a serious production server. First of all, it's not optimised for servers. Second, they're not as good as comparable SCSI drives (Why doesn't Western Digital have a SCSI line anymore?). Finally, you should be using a SAS infrastructure anyway (whether the drives are SAS or SATA), so the Raptors will cost more in the long run.
  4. Considering that you need Win2k3 or newer to get StorPort, and SCSIport still exists, I don't see a lot of crappy consumer stuff making the leap anytime soon.
  5. I think that you've confused "taboo" with "having figured out how non-useful it is."
  6. What do you mean by, "knocking?"
  7. Infrant NASes have an FTP functionality, but I don't use it and cannot comment. I'd suggest looking on the Infrant website. They have all of the manuals up, as well as a forum where you can ask questions.
  8. Your RAID controller needs to expose each disk for third-party software to read SMART information. Generally, manufacturers have monitoring software which will do get status information for you.
  9. IBM/Hitachi SCSI drives are pretty much the noisiest monsters that you can put in your case. Use the performance database to look at drive noise. I haven't looked, but I'd guess that Seagate gets near the top in both the 10k ann 15k categories. If you're looking for cheap, you'll inevitably have to look at previous generation products.
  10. This happens pretty much anywhere that failure rates are relatively low or it's cheaper to replace than repair. That's pretty much everywhere in computer components.
  11. The RAID is handled by the controller chip on the card. You use the same MPT drivers irrelevant of what mode the card is in.
  12. Folks here are right. The operating system generally won't allow this. The issue is that, in order to abstract storage enough such that it's easy to a dress a wide range of devices easily, the operating system doesn't generally know if two devices can talk directly to one another. I/O transactions are between memory and device, not from device to device. Theoretically, SCSI would allow transfers from one drive to another. Keep in mind that the host bus adapter (controller) is simply another device on the bus. The operating system, however, would have to have intimate details about the devices involved. That means that each drive would need a driver. That would be a real pain.
  13. I'd posit that "for whatever reason" was cost. Since the higher-density platters aren't necessary to make 500GB in four platters, Seagate saved money by going with rejects from that high density process. Instead of wasting the platters, they made extra money with them!
  14. I usually suggest ReiserFS, but it does depend on how you will be loading the file system. As long as your distribution supports a file system, it should be able to read it.
  15. I've accidentially attempted something similar, and it didn't work. Presumably, it would be possible with some hacking of the BIOS, but I don't think it'd be worth the effort.