From mboxrd@z Thu Jan 1 00:00:00 1970 From: John Robinson Subject: Re: [Slightly OT] Cheap 4-port PCI-E SATA card? Date: Tue, 04 Jan 2011 14:03:36 +0000 Message-ID: <4D2328B8.4070608@anonymous.org.uk> References: <4D20E9F7.2080800@anonymous.org.uk> <4D216FA1.5010502@hardwarefreak.com> <4D21F2BB.3020905@anonymous.org.uk> <4D224B22.6070007@hardwarefreak.com> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8; format=flowed Content-Transfer-Encoding: QUOTED-PRINTABLE Return-path: In-Reply-To: <4D224B22.6070007@hardwarefreak.com> Sender: linux-raid-owner@vger.kernel.org To: Linux RAID List-Id: linux-raid.ids On 03/01/2011 22:18, Stan Hoeppner wrote: > John Robinson put forth on 1/3/2011 10:00 AM: >> On 03/01/2011 06:41, Stan Hoeppner wrote: [...] > When using PMPs one will always have less theoretical b/w per drive t= han > what that drive can push on paper with a streaming read. Considering > that over 90% of real world workloads are random IO heavy, not > streaming, it's unlikely you'll ever run out of b/w using a PMP based > setup. I haven't. That's probably true but I don't like making compromises if I can=20 possibly avoid it :-) >> I doubt my motherboard supports FIS PMPs. It's an Asus P5Q Pro, Inte= l >> P45+ICH10R, and I'm pretty sure the ICH10R doesn't support PMPs even= if >> the original spec said it would. > > I just looked it up, and the ICH10R does _NOT_ support PMP. Neither > does the onboard Silicon Image 5723 chip, which actually connects to = the > southbridge via a SATA II port (stupid). Must be connected to the Marvell 88SE6111's SATA port. Well, the spec=20 says 88SE6111 but my lspci shows an 88SE6121. The 88SE6121 ought to hav= e=20 two SATA ports, but I've found odd behaviour with the two non-ICH10R=20 ports on the board when testing - basically if I use the second port bu= t=20 not the first port, nothing appears, even with Asus' Drive Xpert=20 disabled. This is presumably the SilI 5723's contribution, hanging off=20 the Marvell chip's SATA port, and clearly no use for RAID where a drive= =20 might die but we want to carry on, so I'll be ignoring the on-board=20 Marvell/SilI stuff. > BTW, the P5Q Pro has 3 PCIe x1 slots and 2 PCIe x16/x8 slots, plus 2 = PCI > slots. You told us the only slot you have available is 1 PCIe x16/x8= =2E > What is consuming the 3 PCIe x1 slots? The PCIe x1 slot just North o= f > the top x16 slot should be free. I'm guessing the other 2 are blocke= d > by your GPU cooler, correct? Ah, I'm sorry, no I didn't say I only the PCIe x16/x8 slot available, o= r=20 at least that's not what I meant. I do need PCIe because both PCI slots= =20 are taken, but I do have all three PCIe x1 slots and one PCIe x8/x16=20 slot available. The graphics card is only single thickness and could=20 indeed be moved. I know I could easily put two $30 2-port cards in, but for neatness I'd= =20 prefer one card with 4 ports, which is what I started this thread to=20 look for. One 2-port card and a PMP is a reasonable solution that I=20 might yet go for. NB My PCI slots are occupied by a TV tuner card and a SCSI card which=20 serves my tape library, and in any case I want to avoid the limitation=20 of the PCI bus which would have half the bandwidth of one PCIe x1 v1=20 slot even if it wasn't sharing with two other highish-bandwidth devices= =2E [...] > I recommended the Marvell based card strictly because it does PCIe x1 > rev 2 for 500MB/s. I only use the Silicon Image based cards, but I u= se > more than one if I have more than 4 drives in an array. The cost com= es > out the same at $40 as the Marvell card, but the SI route requires 2 > PCIe slots, and you only have one. As noted I do have more PCIe slots, and I've just found a Lycom PE-115=20 PCIe x1 2-port SATA-III card for =C2=A320, which some third-party sites= say=20 has a Marvell 88SE912x chip, and is PCIe v2.0, so that'd be my=20 multi-card solution - two of them would give me the 4 ports I need to a= dd. [...] > Well, you've given us conflicting requirements that are impossible to > meet with any solution: > > 1. Low cost > 2. Full b/w per drive > 3. Only one PCIe slot availiable > > There is not a solution available to meet all of your criteria. My own research had got me that far, that's why I was asking here. If=20 this is unreasonable, please excuse the noise... >> Nevertheless, thank you very much for taking the time for such a >> considered reply. > > You're welcome. I think you're finding yourself in that "I want to h= ave > my cake, but eat it too" situation. You can't get what you want at t= he > price point you want. Don't we all want to have our cake and eat it sometimes? ;-) Thanks again for your thoughts. Cheers, John. -- To unsubscribe from this list: send the line "unsubscribe linux-raid" i= n the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html