how much cache are you using with the raid controller?
you know that HDD test in 05 doesnt mean much right
how much cache are you using with the raid controller?
you know that HDD test in 05 doesnt mean much right
it's all in the cache mate it doesnt matter what you clocks are
your disk write cache is enabled. can you disable it or use a 1GB stick and compare also?
intels dont have onboard cache and the settings for the controllers is write through so no cache for write back![]()
i'm confused
i used write back cache via write though with irams on 1231ML
u cant run Write back Cache and Write through at the same time :S
im not talking about the cache memory that some SSD's have im talking about the cache that controller uses.
strange thing is the results were also extremely high and iram is even slower than SSDs
yes some things iram is quicker but some slower also. i've done benching with SSDs on raid controller and ICH and iram also.
can you explain what you mean by latency and what it impacts
random low k files read and write![]()
here is some good results with Areca 1880ix ( i hope now dinos will understand what im talking about)
http://service.futuremark.com/compare?pcmv=384262
Main test results
PCMarkSuite N/A
Memories Suite N/A
TV and Movies Suite N/A
Gaming Suite N/A
Music Suite N/A
Communications Suite N/A
Productivity Suite N/A
HDD Test Suite 157756
HDD Test Suite
HDD 1
HDD - Windows Defender 865.85 MB/s
HDD 2
HDD - gaming 615.9 MB/s
HDD 3
HDD - importing pictures to Windows Photo Gallery 825.67 MB/s
HDD 4
HDD - Windows Vista startup 796.93 MB/s
HDD 5
HDD - video editing using Windows Movie Maker 606.29 MB/s
HDD 6
HDD - Windows Media Center 1231.71 MB/s
HDD 7
HDD - adding music to Windows Media Player 406.86 MB/s
HDD 8
HDD - application loading 798.4 MB/s
hey mate
i just looked at one of my old 150K HDD score results and i am much stronger in some tests but also much weaker in others
here is one that i am interested to know more about
HDD - application loading
how do you get that exactly....are you saying that it's the new controller or what's the deal?
its just using my controller without the cache
no read ahead and everything that has to do with cache
do you still have the 1231ML to compare?
no i just have LSI models (like a maniac) its my first areca and i have to study it first and then run a full public test![]()
here is some results from SiSoftware Sandra Benchmarks.
i have for the moment the first place :P
steve i think i am finally working this controller out, youve always had me by a long way in the hdd tests, but i think i am finally getting there!
Wow Mr. Tilt - great numbers my friend!
Hello Boys,
I have a little problem with perfomace. If you look on graph (8k 75% Read and 100% Random test), you will see that Areca 1800ix have a poor perfomance compared with older controller.
NAVISIONSQLEVRO is our old SQL server ... HP DL580 G3 with Smart Array 6400, 512MB cache + battery.
NAVISION is a Supermicro server with Areca 1800ix-16 controller and ten disks Toshiba 2.5" 147GB SAS 2.0 15k. (RAID10, 64k strip size, battery, write back.)
I tried to change Volume Data Read Ahead normal to Conservative/Disbled. Random read/write perfomace didn't change. Only seq. read is lower.
IO Meter test:
Please help![]()
Mr. Vodokotlic - please send complete controller and storage array settings.
Sounds like yo've already started down the right path - you need to benchmark each alternative configuration and select the best for your application.
F/w used?
storport?
TCQ enabled when the volume was created?
ncq?
HDD read ahead?
volume read ahead (you have adjusted some already sounds like)?
write caching enabled?
what raid config?
Stripe size of arrays?
cluster size of partitions? ...
Others should chime in but my approach would be to allocate a lot of time to try test as many likely optimums as possable.
Last edited by SteveRo; 11-11-2010 at 08:59 AM.
Bookmarks