Sign in with
Sign up | Sign in
Your question

CONROE + RAID = horrible performance

Last response: in CPUs
Share
July 5, 2006 3:51:30 PM

http://www.theinquirer.net/default.aspx?article=32818

Quote:
Conroe shows dodgy RAID performance anomalies

Overheads and slowdowns


By Theo Valich: Wednesday 05 July 2006, 16:24

PRELIMINARY CONROE results have been really impressive, but as time goes by and the platform gets tested in more and more environments, interesting things are popping up.
Right now, it seems that Intel has some serious problems when using RAID5 configuration. If you are using an Intel's own baby, D975XBX motherboard and put four drives in RAID5 array, an interesting overhead and a slowdown will occur on all upcoming Conroes.


Moderator Edit: To comply with copyrights
a c 471 à CPUs
a c 115 å Intel
July 5, 2006 4:02:03 PM

The fact that the 975XBX is having problems is not really a surprise. Revision 1.0 hardware generally has the tendency of being a bit unstable or provide lower performance than a more mature product. This "teething" problem is generally common throughout the industry. But it sucks for people who want to have a RAID setup with a 975XBX mobo.

Hopefully ATI's & nVidia's chipset won't have this issue. But I'm sure they'll have their own problems to deal with.
July 5, 2006 4:06:20 PM

thats why i never buy rev 1.0 products....

wait a little while for the problems to iron out.

But with CONROE....it seems so many people are so over hype for it...that they think nothing can possible be wrong in any way
Related resources
July 5, 2006 4:44:01 PM

I could care less... this is from theinquirer and therefore is a waste of everyone's time.
July 5, 2006 6:17:27 PM

For how long have you been using a 4 drive RAID5 configuration?
July 5, 2006 6:20:04 PM

What do you expect, the chip is not even released yet.
July 5, 2006 6:35:36 PM

Quote:
thats why i never buy rev 1.0 products....

wait a little while for the problems to iron out.

But with CONROE....it seems so many people are so over hype for it...that they think nothing can possible be wrong in any way


It is not CONROE's fault. It's the platform's.

It is like blaming an FX-62 in a computer for having problems with a key in its keyboard.
July 5, 2006 6:37:02 PM

Anyone who is using a 4 drive type 5 raid array for any productive reason would/should not be using an onboard contoller.

Id the on board contoller goes bad, you will have to replace the motherboard in order to access the information on the drives. A plug in card can be replaced and you are up and running in a few minutes.

This has nothing to do with the CPU and all to do with the on board raid controller how it utilizes resources.
July 5, 2006 6:55:49 PM

Quote:
This has nothing to do with the CPU and all to do with the on board raid controller how it utilizes resources.


/nod

Although the inquirer has some iffy information sometimes, I wouldn't doubt this could be true. But exactly like intelamduser said, its the controller not the proc. I'd like to see someone throw in a PCI based controller and find out if the system still had the problems; I doubt it.

Sounds like someone's trying to put the blame on Conroe to downplay its success. Not that I really care, just to anyone reading this thinking oh crap I don't want a Core2 now because it won't work with my raid array: Don't use the onboard controller or find a board with a different chipset.
July 5, 2006 7:18:21 PM

Consider the source of this rumor. No link, not a shred of evidence, published by a rag with zero credibility and posted here by one of AMD's most fanatical cheerleaders.

Yo, Lordp00p... wuzzup?
July 5, 2006 7:20:24 PM

Quote:
For how long have you been using a 4 drive RAID5 configuration?
Who doesn't use 4 drives in a RAID5 configuration? :wink:
July 5, 2006 7:21:38 PM

Quote:
For how long have you been using a 4 drive RAID5 configuration?
Who doesn't use 4 drives in a RAID5 configuration? :wink:

Newbies and wannabes :roll:
July 5, 2006 7:23:14 PM

FRUD TROHL

We meet again

the biggest INTELLIOT since the BLACK DRAGOOON in 1998 on toms old delpee forum


take your INTEL is god somewhere else
July 5, 2006 7:35:43 PM

They are not always wrong, sounds like you're echoing what someone else has said.
In the last couple of months they've been right with things they've reported, but you never hear that. It seems that they are only said to be wrong when it's about something that someone likes and it's bad news so it's wrong. That's simply moronic.
July 5, 2006 7:39:59 PM

What happens if you use 5 drives?

Edit: Attempted smart ass comment. No real explanation needed.
July 5, 2006 7:42:33 PM

Raid 5 has an overhead to begin with at then you go at take some time to process that, your going to have some slowdown and since your loosing clock cycles, your going to have a slower product.
July 5, 2006 7:47:21 PM

My CD drive tray doesn't open / close smooth..
It must be the FX-60 that's causing this :) 
July 5, 2006 8:06:21 PM

Pro AMD news is lies from the INQ
bad INTEL news is lies from the INQ


this is the war cry in the INTEL camp
July 5, 2006 8:08:50 PM

Do you even notice youre only talking to yourself?
July 5, 2006 8:10:56 PM

Trolling, Trolling, Trolling......

Not sure which is trolling more you or the inq?

The inq is a rag which just throws out junk in attempts to swing readers...

The real problem with them is they say they are a news source and they can't even find a way to somewhat validate their findings/rumors.

Identifying something as a rumor is one thing, but presenting information as fact while acting like a supposed authority, all while not validating or even bothering to understand the information is just unethical!!!

That is online trolling...

And you my friend just got hooked!
July 5, 2006 8:17:22 PM

PS..

Anything from the inq = useless...

Whether it is pro Intel/AMD/ANYONE.

If I were to write a blog making some not so grand psuedo announcements I could get relatively close to an accurate statement since it is not so hard.

So, in long, even an elephant can paint a picture that people will pay money for!

I have seen it. The B@B Circus for instance!
July 5, 2006 8:43:55 PM

I'm not disputing the report, just tired of bullshit bashing.
I'm Pro AMD and I'm Pro Intel. I'm Pro Performance I don't give a shit who's label is on it.
I love fast bikes any make and fast cars any make.
Anytime a piece of hardware comes out with no issues it was probably touched by a supreme being, because as long as man has a hand in making anything there WILL be problems.
Man has proved he can screw up anything.
July 5, 2006 8:56:21 PM

You're just giving me more reasons to buy a Conroe come July 27th.
July 5, 2006 9:14:06 PM

I hope you all do realize that anything that eats up bandwidth will have a negative affect on the platform, regardless. In other words, the more you have running the more you are going to slow your system down. Sure this is true for any system, but when you FSB is already starving the CPU, it just makes things even worse. I'm very glad Intel has a better cpu out, just need to keep pressure on them to get their act together and get the new platform out as well. CSI looks good, get it out the door!
July 5, 2006 9:14:13 PM

i am all for the CONROE

its a great chip

but like anything there will be issues here and there

sheeshh

people defend conroe like its the cure for cancer
July 5, 2006 9:23:25 PM

No lordpo,

They are defending the truth.....

They are defending a misrepresentation on yours and the inqs part.

Had you thought through what you had read maybe you would have come to the conclusion that any raid controller can cause conflict/performance issues.

Many have had issues with BIOS and with drivers conflicting with the OS or other cards in the system.

So in fact this problem that the trolls at the inq have attributed to the Conroe is in fact something that is easily duplicated in multiple environments AMD or Intel, new or old, SCSI, SATA or IDE!!
July 5, 2006 9:26:25 PM

Or maybe FSB is old, out-dated, obsolete?
July 5, 2006 9:33:58 PM

Just so you do not think I am blowing smoke:

Linkage:

Link1

Note: On this link the guy had a WORKING/performing RAID subsystem that then started to deteriorate during what should have been an upgrade:
Link2

In Linux:

Link3 redhat.com archives

University study (beware it is a pdf white paper):

Link4

There are many others as well!!
July 5, 2006 9:39:03 PM

Also, for future reference (take it or leave it):

The hardware/software (drivers) items that I have run into (causing the most issues overall) have been the following in order of total screwups by manufacturers:

Modems
Sound Cards
SCSI/ANY Raid controllers (mostly SCSI for me)

Now this is just "MY" compilation of errors over the past umpteen years!

Anyone care to add to the list to inform Lordpo???
July 5, 2006 9:52:54 PM

Quote:
Pro AMD news is lies from the INQ
bad INTEL news is lies from the INQ


this is the war cry in the INTEL camp


Popegoldx I presume.
July 5, 2006 10:10:02 PM

well there is only ONE POPE

wonderful to run into you SPUD

I hope i can last here long enough b4 i am banned to strike up meaningful tech talk with you.
July 5, 2006 10:10:52 PM

I agree with you. Never buy the Frist new chip or your going to have problem. everything to unknown problems with the cpu or bios. Take a look at the Intel duo with the usb port. Yes pluging in anything in the usb port drain power even if the was power running from a wall socket and the laptop was running on battery. It took 3 months for Mirocrap to fix the bug.

Morons buy new chips frist. Smart people wait and watch. Then buy chips. We dont need the inquierer to tell us this.
July 5, 2006 10:14:43 PM

Quote:
well there is only ONE POPE

wonderful to run into you SPUD

I hope i can last here long enough b4 i am banned to strike up meaningful tech talk with you.


It's all up to you just as it always has been. But I am looking forward to a new mature pope!
July 5, 2006 11:01:02 PM

Atoi:

I applaud the early adopters!! It seems without them YOU would never receive a stable product!

I have in fact been known to be an early adopter. In fact on occasion it has served me well! I do understand the risk involved, but also understand the potential gains. Not to mention the road to those gains are often paved with GEEK GOLD!! Nothing like finding a workaround/fix for something! ;) 
July 6, 2006 12:33:18 PM

Quote:
They are not always wrong, sounds like you're echoing what someone else has said.
In the last couple of months they've been right with things they've reported, but you never hear that. It seems that they are only said to be wrong when it's about something that someone likes and it's bad news so it's wrong. That's simply moronic.


I bet you read People magazine and watch entertainment tonight too? lol. You are the people who let trash journalism happen by reading it and giving it your viewership. Don't complain when you realize all journalism becomes trash because of people like you. But I doubt you would even notice.
!