ravton

Patron
  • Content Count

    66
  • Joined

  • Last visited

Community Reputation

0 Neutral

About ravton

  • Rank
    <a href='/patron.html'><b>StorageReview Patron</b></a>
  1. SCSI is probably best to avoid any hiccups in the data transfer. If you want SATA a more mature solution such as 3Ware or Adaptec may be a safer bet. Either way, set your stripe size larger. 1MB or greater for faster file streaming. Also, Windows 2003 Enterprise will use >4GB as file system cache even with normal Xeon processors.
  2. I fat-brained the vendor name. It's Securstar.
  3. Because anybody with physical access to the server can recover the EFS encryption key. Not good. Uh Huh. Check out SecurStor for protection from the boot prompt. Otherwise there are many vendors (including freebies) that provide container level encryption. You'll probably want to look at a "personal" encryption solution because most "enterprise" solutions have a back door. Keep in mind if you lose the password on a "personal" encrypted drive you're screwed. PGPDisk is a good one for container level encryption.
  4. ravton

    Need Advice On Stable Sata Raid5 Controller

    Like most vendors 3Ware pisses me off all the time but I don't think your spin on them is correct. Bullsh!t. This was only the very first 8500's that shipped and 3Ware REPLACED THE CONTROLLERS with no questions asked. How much more fixed can you get then that? I've never heard of the 66Mhz bus issue causing data loss. I have seen bad RAM cause what you describe but not the 66Mhz issue. Their firmware update fixed this for all of our systems (Intel based) since none needed the FREE 3Ware-supplied riser card. Either way, they fixed this sometime around the beginning of the year. I haven't worked with the 9000's yet, so no comment.[snip] Agreed! The P/SATA drives die at a considerably higher frequency then the SCSI drives, resulting in higher maint costs. I'm still waiting a RAID-6 (or other N+(2+) parity) SATA RAID controller to mitigate the risk with all these drives dieing. To sum up, I agree that 3Ware has had some hardware compatibility problems, but what I think they've done a really good job at is FIXING them. It's usually a hardware swap but they're very easy to deal with. Well, once you get past their TERRIBLE first level technical support. Actually I haven't talked to anybody since they completed the merger so I don't know if it's improved.
  5. ravton

    Raid 0:tips For Improving

    631MB/sec? That's 157MB/sec from each of the 4 drives sustained. And what card is doing that? Me think's this Troll is trying to fool somebody...
  6. ravton

    Smart - Imminent Failure !

    Western Digital has an advanced replacement program where they'll send you the replacement drive before you ship them the old one. There's no charge for this, they just need a credit card number in case you never return the failed drive. If it was my data I wouldn't wait. Back it up instantly. Stop reading these posts and back it up!
  7. ravton

    3ware Vocal Alert

    Once past the inept first-level guy (only a single person last time I checked), I've found 3Ware's technical support to be pretty good. Email or call them and ask. If under Linux it's trivial to setup a local mailbox that would alert via audio or visual means. If under Windows you can do the same thing but it would be harder. Perhaps a small smtp/pop server running local and then a popmail checker (for the audio alert)? A kludge but it'd work...
  8. That would be. http://linux-iscsi.sourceforge.net/ Nah, that's for a iSCSI initiator. A linux-based iSCSI target would (will?) be the cats meow.
  9. My hat goes off to early adopters. Beta testing a software iSCSI implementation? I hope you have good backups... Speaking of which, has anybody seen a mainstream supplier do the same? I looked for a linux implementation but haven't been paying attention. Now, THAT would be cool.
  10. ravton

    Fs: 640gb Raid Array

    !!SOLD!! Hmm... - Sure wish I could edit to make it easier to tell...
  11. ravton

    Raid 1+0 On Two Drives?

    This is correct and I've had the same issue with Compaq. I hope the people designing the RAID firmware know more then the people doing the UI Interface. Hmm, I wonder if it saves some dumb support calls. "I have two drives and it won't let me do RAID 1+0! Whats wrong???"
  12. I have a 640GB RAID-0 array that I'm finally done with. It's wicked fast with read numbers around 190MB/sec and writes at about 90MB/sec. It's so fast because it's made up of 8 WD800 drives which the 3Ware controller excels at distributing the load amonst. So the components are: 3Ware 8-port RAID controller 8 x WD800's 8 ATA cables It's 8 drives so please make sure you have the space for them. The smaller drives like these don't require as much power as the big drives and I've had no problem running them with a 350W power supply. If you're interested in fewer drives let me know. I can sell the rest off on eBay. Feel free to email me at ravton@yahoo.com
  13. The obvious followup question is: When will the WD740 article/benchmarks be updated with the new (MUCH) slower results? I seem to recall something like this before with WD. Is this right or am I recalling incorrectly? Does StorageReview update this or continue with the vendor-supplied drives?
  14. ravton

    Raid 1 On A Budget

    I think you need a primer on RAID. RAID 1 writing is about the same as non-raid since writes happen simulanteously. In general, reading is about the same as RAID 0, which is twice as fast as a single drive. As always there are exceptions (like two drives on a single PATA channel) or a stoopid RAID controller.
  15. ravton

    Raid 1 On A Budget

    Depends. Usually it just errors out and the drive goes offline, but like anything, if certain hardware failures occur then the system locks up. I've seen this happen with Software RAID, Adaptec RAID and 3Ware RAID. YMMV. If you're rebuilding with a new drive you need to do it manually through the built-in Windows Disk Management. If you just need to remirror, that is done automatically. It's very, very, very simple.