I still believe that Gordy has an advantage on sequential write speed, but after some more tests and such, I am beginning to see your point about it...
I was referring to IO Meter tests, which I can post if you would like, that show his results peak at much lower queue depths on all tests. As for...
Thank you, knowing that Raid10 writing double actually uses up half of the 320 channel bandwidth explains a lot. So do you think thats the reason...
This isnt about competition, but real world performance. Knowing my drives are capable of going much faster than the computer / controller seems to...
Nevermind, I just saw your response to Gordy's post in another forum, I guess we will see when he posts his other results with longer duration, but I...
Joe, No comment on this? =)
Joe, Here is why I have gotten the impression that SATA is much better than U320 on write speed: gordy's Update results: DWritesPerSecD Write...
I see that makes a bit more sense then. And yeah, all arrays have been always set to Direct IO. I am running Raid10 on each array. (2x14xRaid10)...
Seems to me that considering I am using 28x15K disks to your 11x10K, that my performance is way low =) What would you suggest looking at? I did not...
Here they are: 73,38577,30000,899,1285.9,30,38577,0,0,648.6334,2.532585,0.02913819,638.6,0.04551623...
The 13MB/s is the speed I am getting per drive array so that # is still valid. The thing that still doesn't add up though is why I get 3x this speed...
Sorry I forgot to mention that I am using two identical HBAs and powervaults. 28 drives total. As for the write speed his sata blows me away on...
I did try the Update Test and scored 5MB/sec write speed which seemed really bad, but wasn't sure what that meant in context though. I know a friend...
Ok, well we are actually using SELECT INTO to create subsets of data from our main table so that the many updates do not issue table locks against...
quote: i seriously doubt your IOMeter 8K random write at low queue is going to be much better than 13MB/sec Well I agree, and thats the whole...
To clarify, we are running 2005 Enterprise. quote: in your case your IO Meter test was a single sequential write, in which case, you should not...
Hello all you SQL Performance gurus out there. I would really appreciate some insights to an issue I am having with write performance. System...
Separate names with a comma.
