#SILICON IMAGE 3124 FULL OF REPORTS#
#SILICON IMAGE 3124 DRIVERS SILICON IMAGE#
#SILICON IMAGE 3124 SOFTWARE AND GETS#

The first question (and most important) is that I have some existing volumes defined in the array with a lot of data on them. I have a few questions centering around using a Silicon Images Sil3124 based RAID controller with Ubuntu 8.10. Silicon Image’s SiI 3124 is a 64-bit 133 MHz PCI-X to 4-Port Serial ATA controller that will be available in two pin-compatible versions: the SiI 3124-1, which operates at first-generation SATA speeds of 1.5 Gbps per port, and the SiI 3124-2, which operates at SATA speeds of 1.5 Gbps and also provides support for 3 Gbps signaling.Hi.

Silicon Image 3124 Drivers Silicon Image

The 2-port ones are a nativeSilicon Image Liens vers le site de Silicon Image Page daccueil (Page de t&233 l&233 chargement (Actualit&233 s relatives &224 Silicon Image 04/07/05 Nouveaux drivers Silicon image Sil3114 RAID et IDE 23/06/04 Contr&244 leurs RAID Sil3x12 : panneau de configuration Cat&233 gories de mat&233 riel associ&233 es &224 Silicon ImageSilicon Image SIL0649 ATA100 RAID Controller 2 independent IDE/ATA channels. There are some Sil PCIe cards out there. There is more than one version of the 3124, but Ive not heard errata about either version. Im running with two SIL3132 PCIe cards in my system with no problems. If I can get the card working properly in Ubuntu will I be able to use the existing volumes without losing any data?skeptic to silicon image.

Silicon Image 3124 Software And Gets

Silicon Image 3124 Full Of Reports

Forums are full of reports of recovery failing (read: wipe out the good drive left over when one drive fails in a raid-1) when a reboot happens during recovery or something similar.On the other hand, Linux' md driver (real software raid) is one of the finest pieces of Linux software and gets your behind out of the wheelbarrow all the time.In addition to the better chance of recovery it has more advantages such as not limiting you which drives you can smack into the array. The recovery in case of an error is complicated enough for one dedicated RAID system, but in the case of these things you have them twice, once in the BIOS, once in the driver at runtime. Please help me :)It doesn't directly answer your question, but unless you want to run raid-0 I cannot recommend that you use any of that onboard SATA raid stuff. Reading up only gets you so far.there are plenty of incomplete guides that don't take into account what could go wrong when you're following them through.

Even if you want to use this kind of RAID, using a random binary driver downloaded from a non-Linux vendor is increasing your risk.You should consult the libata documentation and see whether this chip is supported. I cannot copy the data off and rebuild new volumes for the data).You need to track down whether a driver for this is in the mainline kernel. I need to know if I can use these volumes in Linux as they are just now (i.e.

If you are using a bootable version of the SATARAID5 driver, it is already included in your Linux kernel. Before launching the SATARAID5 Manager, the driver must be installed and the daemon must be running in background. What will be 2.6.19).This () also seems suggest that the Sil3124 is supported by libATA.The User Guide () for the SATA RAID Manager says the following:On a Linux system the SATARAID5 Manager and daemon can be installed into any subdirectory.

Otherwise, you must manually start the daemon in a background mode by typing the following command:Finally, you can launch the SATARAID Manager by typing the following command:However, I can't find the package containing the deamon and RAID manager for Linux anywhere on the Silicon Image site. If the daemon has been incorporated into the startup scripts (typically included within the /etc/rc directory structure, it will start automatically when you boot your system. insmod si3XXXr5.ko (where “3XXX” refers to your specific SATA controller chip)Then, the daemon must be started.

It worked awhile ago (year an a half or so) and then bot timing issues crept into the picture - this seems to only be an issue wiht Intel chipsets. The five spindles in the external enclosure are put together in software, and the current Ubuntu 9.04 kernel supports it fine (both 32 bit and 64 bit).One gottcha seems to be booting with the multiplexer. I just don't know where to go with this.Ive got a couple of these running in various machines - the idea with the 3124 is its ability to multiplex one eSATA port and let you address multiple drives - in my case I have an enclosure that has 5 500GB spindles in it.Essentially its pretty decent as far as inexpensive interface goes. There is a note saying 'Any operating system not listed above is not supported.', but this is clearly in contradiction to the User Guide quote above.

And the performance is fine for having a 2TB mount point out there. Pain in the backside, but it works.Mdadm finds stuff just fine adn putting 5 spindles into a big repository has its advantages.Anyway, its inexpensive (not 'cheap') to implement. On boot the machine will got to busybox, and I havent found a workaround.That said, my simple solution is to disconnect the eSATA cable when re-booting, then sson as the boot starts plug it back in. Turning off APCI sems to have no effect.