On Tue, Oct 17, 2000 at 12:31:42AM -0700, Andre Hedrick wrote:
>
> DiskPerf /dev/hde
> Device: Maxtor 91536H2 Serial Number: N200PGDD
> LBA 0 PIO Read Test = 5.32 MB/Sec (46.98 Seconds)
> LBA 1 PIO Read Test = 5.32 MB/Sec (46.99 Seconds)
> Outer Diameter Sequential PIO Read Test = 5.29 MB/Sec (47.24 Seconds)
> Inner Diameter Sequential PIO Read Test = 5.29 MB/Sec (47.28 Seconds)
> LBA 1 PIO Write Test = 5.22 MB/Sec (47.86 Seconds)
> Outer Diameter Sequential PIO Write Test = 5.21 MB/Sec (48.02 Seconds)
> Inner Diameter Sequential PIO Write Test = 5.22 MB/Sec (47.89 Seconds)
> LBA 0 DMA Read Test = 48.84 MB/Sec (5.12 Seconds)
> LBA 1 DMA Read Test = 49.07 MB/Sec (5.09 Seconds)
> Outer Diameter Sequential DMA Read Test = 26.53 MB/Sec (9.42 Seconds)
> Inner Diameter Sequential DMA Read Test = 14.68 MB/Sec (17.03 Seconds)
> LBA 1 DMA Write Test = 44.94 MB/Sec (5.56 Seconds)
> Outer Diameter Sequential DMA Write Test = 14.72 MB/Sec (16.98 Seconds)
> Inner Diameter Sequential DMA Write Test = 14.39 MB/Sec (17.38 Seconds)
>
> These are all 4000 IO's at 128 blocks each or a total of 512,000 blocks.
> Not to bad........
>
> Now we can profile drives to super-charge the elevator!!
I doubt this will ever happen. We'll see IDE drives reorder the
requests inside their caches sooner. The elevator will always be more
important to get right in terms of fairness to processes than bare
throughput.
> One catch, these tests require a native ATA-passthough
Do they? I believe I can do the same with a simple script ...
> and that will not be included until 2.5.........
But you still can publish it, right?
-- Vojtech Pavlik SuSE Labs - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Mon Oct 23 2000 - 21:00:10 EST