Slashdot Mirror

I'm considering migrating from Ubuntu to Debian stable for my NAS.

9.10 was pretty funny. It thought my 3114 PCI card was an nVidia RAID Array. It only took them a few weeks to get that fixed, but I still reverted to 9.04, and it's got me thinking about Debian instead.

Hmm... recompile my kernel...

No, I think it'll be easier if I just stick with Ubuntu 9.04. ;)

But knock yourself out. If 10.04 works again, that'd make me quite happy.

http://www.jetway.com.tw/jw/ipcboard_view.asp?productid=279&proname=J7F4K1G2ES-LF
http://www.siliconimage.com/products/product.aspx?pid=28

P.S. The PCI 4 port SATA card may be fixed now - but when I tried 9.10 in 2009, it detected it as an nVidia RAID Array (!) and obviously didn't function.

I had a very bad experience with my old nVidia nForce4 Chipset motherboard RAID chip while using it and just recently found that none of the old or new drivers work correctly when the Intel X25-M 80GB SSD is plugged into the motherboard causing my Windows OS to freeze during boot-up when the driver is initialized or the RAID capability just doesn't work at all. I even wrote up an entire account of this problem in a few threads, one on nVidia's forum and another on HardOCP Forum to warn users about trying to use Intel SSDs with their older nForce4 hardware that I linked to below.

The Silicon Image 3114 PCI to SATA 1 controller chip has serious issues also that caused it to drop my RAID-5 and destroy the 2 TB array. It has issues with PCI bus contention and also is incompatible with the Creative Labs X-Fi PCI sound card on the same bus causing audio stuttering and pops. A few people mentioned that the issue might be IRQ sharing but I tried the sound card in all different PCI slots with different IRQs and the problem was still there. Jet another bad experience with off-brand storage chips.

My current Asus P6T motherboard for Intel Core i7 with the JMicron JMB363 PCIe to SATA chip and JMicron JMB322 SATA 1 to 2 Port Multiplier chip are also having issues with the internal SATA ports where one of them is port-multiplied and if a hard drive and an LG Blu-Ray optical drive is connected at the same time to the internal ports the optical drive will randomly disappear and re-appear in the operating system.

So Marvell is not the one and only manufacturer of storage interconnect chips to have these problems. My experience is that pretty much all of them have issues to varying degrees driving users mad when they realize after purchasing the motherboard and trying to use these chips.

[H]ard|Forum > [H]ard|Ware > Data Storage Systems - The Hidden Cost of an SSD Upgrade (Intel SSD + nForce4 Chipset = No RAID! -> Upgrade)

[H]ard|Forum > [H]ard|Ware > Motherboards - nVidia nForce MediaShield 15.23, .25, .26 - Does Not Install!

NVIDIA Forums > nZone > Hardware > nForce Mobos > nVidia nForce MediaShield 15.23, .25, .26 - Does Not Install!, MediaShield installer does not install services, GUI, or files

OverclockersClub Forums Hardware Processors, Motherboards and Memory - Creative Labs Sound Blaster X-Fi and Silicon Image 3114 on DFI SLI-DR

o That is only a SINGLE CHANNEL IDE card (2 drives max)

o It's a bit pricey compared to a Silicon Image IDE-133 RAID-capable card:

http://www.siliconimage.com/products/product.aspx? id=31
http://computers.pricegrabber.com/storage-device-c ontrollers/m/19271025

o IDE is $dying. SATA and other tech is the way to plan for $future, and believe me I am currently fairly heavily invested in IDE drives.

Intel has specifications for SATA PHY interfaces, search for "SAPIS". Most of the SATA PHYs comply to this interface, though all the implementations of this that I have found are in the form of IP cores.

SiliconImage had some (using their proprietary SATALite interface, press release here: http://www.siliconimage.com/news/press/detailpress release.aspx?id=134) but they disappeared from their product list earlier this year. Back then, the datasheets were only available on-demand. This particular chip was often used with older VIA chipsets and some PATA+SATA RAID boards.

I think Marvell might also have some but all their SATA products appear to be on-demand.

As for latency, it is only measurable when issuing read commands. In that case, the controller has to wait for the "slowest" drive and this is worst when using an array of disparate drives: active head positions and transfer speeds are uneven so data cannot be expected to be typically late by one rotation at most. I wonder how much of this an NCQ controller fitted with NCQ drives could hide.

With all the gigabit serial links being deployed now, maybe next-gen low-cost FPGAs will start featuring some MGTs... they will need at least PCIE-x1 compliance sooner than later but probably would not feature more than four MGTs for a while.

I wonder what the pricing on 4VFX20 is... it is the closest 90nm equivalent to the 130nm XC2VP12 and I would be curious to see how much above/below $200 it might be. (The fact that it is not yet available for open distribution screams hefty premium though.)

I'm rather ignorant on this topic, so another question, my board basically has this built in. Can it do what i want it to?

Addonics makes such a bridge board here. Note their comment about compatibility. Also, AMS makes two styles here that use the 3611 chip according to their data sheets.