Page 2 of 5 FirstFirst 12345 LastLast
Results 16 to 30 of 71

Thread: raptor problems

  1. #16
    Joined
    Dec 2001
    Age
    71
    Posts
    64,554

    Re: raptor problems

    Definately not normal for a VT8237 RAID-0 setup. I'm still trying to find those past threads.

  2. #17
    Joined
    Dec 2002
    Location
    Tulsa, Ok
    Posts
    13,940

    Re: raptor problems

    We have seen a few HD-Tach results that look very similar to this. Not sure what or if the problem was resovled or if it is a issue with Hd-tach.

  3. #18
    Joined
    Aug 2002
    Posts
    50

    Re: raptor problems

    Take a look at this. http://www.simplisoftware.com/Forums...opic.php?t=221 The guy in this article fixed the problem by making the block size 16k. Would you guys recomend doing what he did?

  4. #19
    Joined
    Dec 2002
    Location
    Tulsa, Ok
    Posts
    13,940

    Re: raptor problems

    Have definitely seen the larger stripes show more sawtooth benchmarks.

  5. #20
    Joined
    Dec 2001
    Age
    71
    Posts
    64,554

    Re: raptor problems

    Quote Originally Posted by crocker84
    Take a look at this. http://www.simplisoftware.com/Forums...opic.php?t=221 The guy in this article fixed the problem by making the block size 16k. Would you guys recomend doing what he did?
    I would try it. Chances are you'll get better overall performance that way, instead of performance that favors larger files. Some experimentation is always a good idea.

  6. #21
    Joined
    Feb 2001
    Location
    USA, FL
    Posts
    3,597

    Re: raptor problems

    also try 8K stripe and adjust the PCI latency to its highest level allowed, usually 254 or 255....since the chipset occupies itself on the pci bus, it could be getting big hits when the bus is busy with other duties...so, increasing latency should help...the 8K will also improve ur performance, possibly shallowing out those valleys.

    also, look into updating the board bios to see if there is a later version to improve the raid chipset performance...

  7. #22
    Joined
    Dec 2001
    Age
    71
    Posts
    64,554

    Re: raptor problems

    Quote Originally Posted by Ace-a-Rue
    also try 8K stripe and adjust the PCI latency to its highest level allowed, usually 254 or 255....since the chipset occupies itself on the pci bus,
    VT8237 SATA isn't on the PCI bus.

  8. #23
    Joined
    Aug 2002
    Posts
    50

    Re: raptor problems

    Ya I didn't think it was, but I believe I am going to try the 16k. I just talked to a good friend of mine, he built a new computer this week, same drives as me with a msi neo nforce 3 board and has his block size set to 64k. In hdtach he is getting the same results, so i believe that it is just a combination of the raptors, windows and block size. When I get if formatted i will post results.

  9. #24
    Joined
    Feb 2001
    Location
    USA, FL
    Posts
    3,597

    Re: raptor problems

    Quote Originally Posted by BWM
    VT8237 SATA isn't on the PCI bus.

    where is it positioned?

  10. #25
    Joined
    Dec 2001
    Age
    71
    Posts
    64,554

    Re: raptor problems

    Quote Originally Posted by Ace-a-Rue
    where is it positioned?
    It isn't really on a bus at all, rather, it is an independent block of the southbridge, which then uses the Vmap link to the northbridge. The following link will pull up the block diagram of how the chip ties in with the rest of a system using the K8T800 chipset.

    http://www.viatech.com/en/images/pro...blkdiagram.jpg

  11. #26
    Joined
    Feb 2001
    Location
    USA, FL
    Posts
    3,597

    Re: raptor problems

    i did a little investigation and i see that the raid solution resides on the VT8237 and the 8237 also intergrates with the PCI bus....there is numerous issues, as i see, that might interfere with the raid chipset...it could be PCI bus, USB-2, ethernet, audio or other sata drives..

    my suggestion is disabling in the bios USB-2, audio, ethernet to see if that makes any difference...try pumping up the PCI latency...low latency could be complicating the communication part to the southbridge and intergrating raid communication with the PCI bus....

  12. #27
    Joined
    Feb 2001
    Location
    USA, FL
    Posts
    3,597

    Re: raptor problems

    Quote Originally Posted by BWM
    It isn't really on a bus at all, rather, it is an independent block of the southbridge, which then uses the Vmap link to the northbridge. The following link will pull up the block diagram of how the chip ties in with the rest of a system using the K8T800 chipset.

    http://www.viatech.com/en/images/pro...blkdiagram.jpg

    exactly, the PCI bus is intergrated with the 8237....it is all about timing, vis-a-vie, Latency!

  13. #28
    Joined
    Dec 2001
    Age
    71
    Posts
    64,554

    Re: raptor problems

    Quote Originally Posted by Ace-a-Rue
    i did a little investigation and i see that the raid solution resides on the VT8237 and the 8237 also intergrates with the PCI bus....there is numerous issues, as i see, that might interfere with the raid chipset...it could be PCI bus, USB-2, ethernet, audio or other sata drives..

    my suggestion is disabling in the bios USB-2, audio, ethernet to see if that makes any difference...try pumping up the PCI latency...low latency could be complicating the communication part to the southbridge and intergrating raid communication with the PCI bus....
    Vmap is a much higher bandwidth bus than PCI, thus SATA sharing Vmap bandwidth with PCI traffic isn't really a problem.

  14. #29
    Joined
    Feb 2001
    Location
    USA, FL
    Posts
    3,597

    Re: raptor problems

    but, it all has to be controlled by the southbridge....i think there is a bios issue with the southbridge...could be latency...i would look into both!

  15. #30
    Joined
    Dec 2001
    Age
    71
    Posts
    64,554

    Re: raptor problems

    Quote Originally Posted by Ace-a-Rue
    but, it all has to be controlled by the southbridge....i think there is a bios issue with the southbridge...could be latency...i would look into both!
    PCI latency is just that, the PCI bus, on which the SATA controller is not. I'm more inclined to believe it isn't a latency proble, rather, a background service related problem. Something is accessing the disk, the graph isn't random enough to convince me otherwise.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •