Here is the syslog.

 

don't really know what is happening. It says "command not found" when I try the ethtool command.

The SKGE is listed when I run the LSMOD command.

Ifconfig, just reports back 127.0.0.1, that's a loopback address, looks like it cant see the network card at all!!

 

Tom, can you take a look at this 4.3beta5 syslog?  Compare it with his earlier 4.2 syslog.  It appears that his DFI Lanparty nForce4 board is not compatible with the 2.6.24 kernel, but it was fine with the 2.6.22 kernel.  The trouble starts when the ACPI module begins processing.  Plus, it does not find his network chipset, and only 1 of his 4 SATA drives, and that one has serious errors.  If it had processed his ACPI hardware correctly, perhaps it would have been able to process the rest of his hardware correctly.  Perhaps it needs some special boot parameter, something like NOAPIC?

 

The troubling line in the syslog is:

 

Apr  4 12:47:11 Projectx kernel: ata1: EH in ADMA mode, notifier 0x1 notifier_error 0x0 gen_ctl 0x1581000 status 0x1540 next cpb count 0x0 next cpb idx 0x0

 

You might try editing your 'syslinux.cfg' file on the Flash, changing

 

append initrd=bzroot rootdelay=10

to

append initrd=bzroot rootdelay=10 sata_nv.adma=0

 

There have been a lot of problems reported with linux/nforce4.  For example:

 

http://groups.google.com/group/linux.kernel/browse_thread/thread/431a202a615ca000/9add0153612875a2?lnk=st&q=EH+in+ADMA+mode#9add0153612875a2

 

and,

 

http://groups.google.com/group/linux.kernel/browse_thread/thread/2012371f5e2a5895/0ddcb5f510c9dacb?lnk=st&q=EH+in+ADMA+mode#0ddcb5f510c9dacb

 

 

Well, I got into the web interface on 4.3 by using a different network card that I had lying around.

 

It made no difference, it is the exact same problems I had in 4.2. Works fine without the parity, as soon as I use a parity drive it hangs after about 10 minutes.

 

So I am going to try a reiserfsck using the rebuild tree option, and if that doesn't work, I willl wipe the whole lot and start again.

 

Thanks for all your help Rob, Weebotech and limetech.

You might try editing your 'syslinux.cfg' file on the Flash, changing

 

append initrd=bzroot rootdelay=10

to

append initrd=bzroot rootdelay=10 sata_nv.adma=0

 

Would you try Tom's suggestion, then capture a syslog and attach it?  I think both of us would like to see what happens with that parameter added.

 

ok I will try it later on today and get back to you!!

Ok, Here is the syslog with the changes to the syslinux.cfg.

 

I used 4.3 beta 5 to test.  I hope this helps you out.

 

It didn't make any difference to my problem though, it still hung about 10 minutes after enabling the parity drive.

 

I also tried reiserfsck with the rebuild tree option on both disks, but no joy there either.

It didn't recognize that option, try this instead:

 

append initrd=bzroot rootdelay=10 sata_nv.adma_enabled=0

 

ok, try 2 with the new option. Same result.

It should have recognized Tom's first suggestion (sata_nv.adma=0), but didn't.  Another similar thread:  http://ubuntuforums.org/showthread.php?t=471255.

 

Just to clarify, you can boot the server without a problem, and it can sit there for at least 15 minutes without a problem, and Web page looks good, temps look normal, but 10 minutes after you click the Start button, it hangs with no further refreshes possible on the Web management page, and no new messages on the console?  Do remember to start a tail running on the console so you can see the very last message it is able to display.

 

I prefer this form of the tail command:  tail  -f  --lines=99  /var/log/syslog

so that I can use Shift-PgUp and Shift-PgDn to page up and down through the last 4 pages of the syslog.

 

OK, I got disk diagnostics tools from Western Digital, ran it and find one disk was given errors. Ran the repair and it is working fine now with parity enabled.

 

Is there no tool like chkdsk or scandisk in Linux? It would have saved a lot of effort!!

 

I am back using 4.24 because it works best for me.

 

Sorry that it just turned out to be a disk fault.

Is there no tool like chkdsk or scandisk in Linux? It would have saved a lot of effort!!

 

There is - and you ran it - reiserfsck !

 

But it sounds like the WD diag tool found something else to fix.  If you don't mind some curiosity, do you recall what it identified as wrong?

 

Tools like chkdsk, scandisk, fsck, and reiserfsck are file system specific, and only fix problems at the file system level.  WD tools essentially work at the physical level, not the file system level.