Upon examining the results from this article, Promise went to work on finding out why the FastTrak's sequential write scores were so poor. They were soon able to reproduce the issue. According to Promise:
"We found on certain motherboards we were generating excessive interrupts under Windows 2000. Our newest Family Driver 2.00 Build 6, which supports the FT66/FT100 and FT100TX2, remedies the excessive interrupts issue under Windows 2000."
Promise requested that we re-run the Sequential Write test on both a RAID 1 and RAID 01 configuration using the new drivers. Below are the new results along with the FastTrak's previous results and the MegaRAID IDE100's scores.
Page not found | StorageReview.com
Page not found
The requested page could not be found.
IOMeter Sequential Write
# IO's outstanding
RAID 1 Sequential Write (MB/sec)
RAID 01 Sequential Write (MB/sec)
FastTrak100 (old)
FastTrak100 (new)
MegaRAID IDE100
FastTrak100 (old)
FastTrak100 (new)
MegaRAID IDE100
1
1.00
27.36
27.82
1.02
37.66
38.57
4
1.00
27.33
27.82
1.02
38.03
39.05
16
1.00
27.33
27.82
1.02
37.95
39.03
64
1.00
27.32
27.82
1.02
37.89
38.98
256
0.97
27.02
27.51
1.00
37.51
38.61
It's undeniable that the new FastTrak drivers resolve the interrupt issue. We commend Promise's timely resolution of the issue.
