Re: SATA RAID 1 - which controller to pick?

From: Revo (rmon90223_at_hotmail.com)
Date: 07/02/04

  • Next message: Me: "notebook and BIOS" 
    Date: 1 Jul 2004 21:00:33 -0700

    They would not even let me see their i/o benchmark figures, let alone
    release their source code for a driver, god forbid!

    BTW, I restarted testing the Adaptec 1210SA on freaking Windows.
    Because I could not get the Redhat driver working, and the Suse driver
    doesn't seem to work at all. Their Windows XP driver works perfectly,
    of course, after a full day and a half of IOMeter and no mishaps. But
    we thought against putting a Windows XP server in production, so we're
    testing Windows 2000 Server with their newest driver now.

    I noticed this started happening in the last 2 days of testing on this
    1210SA card.. when I do a warm-boot (ctrl-alt-del) or hardware reset
    button, the 1210SA appears to hang during the bootup. If I do a total
    power off/on, then it initializes.

    Do you get this problem?? I don't recall it happening originally
    during my Redhat testing phase.

    v8625@hotmail.com (v8625) wrote in message news:<7c6d0c8e.0407010715.4c8c4982@posting.google.com>...
    > Not surprisingly, Adaptec provided the following answer on whether
    > they can provide either another prebuilt module or source code for
    > building my own:
    > "...the only prebuilt modules we have are posted on our website, also
    > we do not have source code for this as it is proprietary information."
    > Anyway, I always ask, just in case.
    > As for SCSI, I weaned myself off that back when fast IDE drives
    > appeared (ATA/100 and ATA/133) - they did RAID0 and RAID1 - good
    > enough for most solutions. SCSI is really an overkill for small
    > environments - any SCSI solution is still significantly more expensive
    > than SATA. And I don't expect it to come down any time soon. In fact,
    > I think SATA would continue to add speed and features (improved
    > signaling, native command queuing and other SCSI-like functionality)
    > and thus chip away at SCSI market from the low end.
    > Anyway, I am thinking if I have to reinstall the system - which is
    > what Adaptec guys originally suggested - I might as well get a new
    > Athlon-64 board with 4 on-board SATA inputs. Then again, I am not too
    > sure if the applications would work on 64-bit athlons, so maybe I'll
    > just fall back to 2.4.20-8 and save myself some time and effort (oh
    > yes, and about $750.00).
    >
    >
    >
    > rmon90223@hotmail.com (Revo) wrote in message news:<d285a739.0407010023.63f04026@posting.google.com>...
    > > Yep, I was looking for the same thing to compile into my own kernel!!
    > > Unfortunately, no source is released by Adaptec. So I had to fall back
    > > to using the 2.4.20-8 drive for that kernel release. I don't really
    > > care as long as thethe only prebuilt modules we have are posted on our
    > website, also we do not have source code for this as it is
    > proprietary
    > information.
    > drives don't crash. Since a newer kernel is just
    > > mostly for system patch issues. And right now, this SATA driver just
    > > isn't there yet.
    > >
    > > On the other hand!! I found out that if you buy the Adaptec RAID 2120S
    > > SCSI controller, and use a 2x SCSI drives mirror, it works PERFECTLY
    > > on the new Adaptec RAID 2120S Redhat 9.0 2.4.20-8 driver! The Bonnie
    > > I/O tester is still looping non-stop after 48 hours of continuous disk
    > > I/O testing. However, the price of that controller is much higher, as
    > > are the prices of the SCSI drives. It accomplishes our goal, but at a
    > > much higher price.
    > >
    > > It seems SATA is still too new to implement in production level
    > > environments.
    > >

  • Next message: Me: "notebook and BIOS"