Why do I get to be on there twice (ie, 1st and second) The higher result is just my array with an added drive. The controller is an Areca 1220 and its a 800GB RAID5 array if you want more detail for the table
It's not supposed to be a competative benchmark, more a collection of how different setups perform. Hence the reason for your two differents setups (even if it is only another drive) being on there.
gotcha well the first one was 4x160 and the latter 5x160 (hopefully the next one will be 6x74GB raptors )
Awhile ago(talking a few years back) i read where after some time the nf7-s's SATA controller would bork/corrupt the data on SATA drives..thus render the data useless. Now i know many moon's have passed and it could have been a driver issue and its all been sorted(i think?) but i just want to be reassured before i go spending £100+ on 2 new HDD's for the rig.
Poor Raid 5 Preformance ? Here's my general spec's: AMD Sepron 3400+ 2gb Ram WinXP SP2 LSI MegaRaid i4 (PCI - 4 channel / 8 Drive PATA raid controller) 3x Seagate ST3400632A (in a RAID5 array) --- As you can see, not the preformance I was expecting from this array. Can you help me fix this? (StorageReview.com Post... indepth spec's and benchmarks)
I saw your post on SR yesterday and couldn't think of anything to add on top of the advice you have already had. Your controller seems to be a bit of a bottleneck (or more accurately put, one hell of a bottleneck). I would suggest you use RAID10 (with another drive) as the controller doesnt seem to have the grunt you need for decent RAID5 performance. Regardless of the individual performance of the drives attached, there will always be a limit to the throughput of the controller. The onboard processor can only do so many XOR calculations per second. Even the uber Areca 1280 will run out of steam at a certain point, regardless of how many drives are attached. If that isnt an option, you could always run RAID0 with your onboard controller, but that is risky with 3 relatively unreliable drives. The other option is to just run them seperately. They will be quicker on their own than they are the way you have them currently configured, and you wont lose any space to parity.
Of course right after I post this the StorageReview's Forums go down When SR's Forum's come backup you can see that I even tried a Raid 0 array on this controller card and got poor preformance... awe heck, why wait I'll repost the Raid0 preformance here: As you can see... not good preformance at all This controller can't be this bad, can it? Is there something I am doing wrong? You can see once the SR post comes back online but I have checked all of the following: -Jumpers (tried CS and Master and Slave Settings) -Device Manager (All IDE devices are set to use DMA if available) -BIOS (All Bios settings are for DMA) -Cables & Length (I've tried both 32inch [rounded] cables as well as 18inch [standard] IDE cables) -Raid Controller Bios (I've tried every setting there is to try, and the best preformance I can get is 74.5MB/s Burst & 54.8MB/s Avg RAID5) Any and all help would be greatly appreciated. Thanks, -BassKozz
My old Seagate 120gb ATA drive My nice new Samsung SP 250gb Very interesting comparason from old and new hard drive that should help ATA people decide if SATA speeds are worth it or not
All three of my Seagate HD's are brand spankin new... granted they are PATA drives, I still shouldn't be seeing such poor preformance should I? Also how much does a difference (in preformance) does Burst Speed Make? My Avg speed isn't that bad (roughly ~10MB/s less then your SATA drive) but my burst speed can't touch yours. Thanks, -BassKozz
You're seeing slow burst rates as a result of both your controller not being able to handle high throughput and using PCI (which will max out at around 100-120MB/sec regardless of the speed of the controller)
