Here's the story.

We purchased a Dell PowerEdge 1850, Dual Xeon processors, 1 146GB HD, 3GB RAM etc. I installed FC4 on the box.

1. During the installation we chose automatic HD partitioning, then deleted the logical volume allocations, and created new swap and file systems partitions. When we finished the setup, and the install actually tried to start, it failed. On the 2nd time through, we made the same HD partition choices and then setup completed. We fed it discs and all was well. Also during the install, Setup did not recognize my monitor, I believe this is because it's on a KVM switch. As such, setup loaded a generic driver for the "unknown vendor" monitor.

2. Once the install completed, and we logged in, I used Up2Date to get the updates for all the packages EXCEPT the kernel. (we're on kernel 2.6.11-1 by the way) The Up2Date downloads would hang on certain packages, I had to restart the app four or five times before it got through all the packages. That was kind of annoying, but I didn't think too much of it, because it eventually DID get all the updates (except the kernel, because I didn't want to update the kernel)

3. Things worked fine for a few days. We installed Webmin. No problems. I installed Retrospect Backup client, to back up the system, the install was fine. The next day, my server is unavailable. I check the PowerEdge and the amber error light is flashing.

I had to do a hard power down and reboot. The server appeared to boot normally (one error during boot is: FATAL: Error inserting acpi_cpufreq (/lib/modules/2.6.11-1.1369_FC4/kernel/arch/i386/kernel/cpu/cpufreq/acpi-cpufreq.ko): no such device) but I think i've been getting that since day 1 (Bad things happened on day 3) Except when it was supposed to get to the login screen, I get an error on my monitor that says:

ATTENTION
SIGNAL FREQUENCY
IS OUT OF RANGE
FH24KHz FV37Hz
PLEASE CHANGE
SIGNAL TIMING

If any of you have any idea if Fedora 4 (or this version of the core) has issues with the PowerEdge 1850, please let me know.

If any of you have any idea if the monitor problem is easily solved, please let me know.

If any of you have any idea what I can do to solve the acpi_cpufreq error, please let me know.

Thanks a lot!

David

hi there

what u can try is boot into safe mode and then change /etc/inittab file to bot into 3 runlevel

then ur system will boot fine and then u can make changes with xorgconfig or xorgsetup to make things fine for u
also i do not think the KVM switch will create any problem for u as it working fine for me

regards

First of all, thanks for reading my original post and responding. Secondly, thanks for reading this one.

I've read some about the xorg.conf file. I've tried to edit it and allow the frequencies that it requires, but sometimes still on boot it halts where the login screen should be with the Signal Timing error. The OK news is that I can still get in through SSH and Webmin to troubleshoot, but as I have NO idea what to fix, it feels a little hollow.

The machine actually booted successfully once, we got in and chose a different monitor, which told us it was going to change the xorg file. Indeed it did, as I've shown below, but when we reboot it STILL comes up with the Signal Timing error.

EXCERPT OF xorg.conf

Section "Monitor"
Identifier "Monitor0"
VendorName "Monitor Vendor"
ModelName "Gateway VX700"
HorizSync 30.0 - 86.0
VertRefresh 50.0 - 120.0
Option "dpms"


I have a feeling it's a core problem, or a dual processor problem. Sometimes we can get the 'up' kernel to boot, and it boots fine. once. Sometimes we'be been able to get the 'smp' kernel to boot just fine. But only once. Whenever we try to reboot, it comes BACK up with the signal timing error. Any more ideas?

If we can't figure this out here at work, we will most likely blow it away and move to Redhat...

Sorry but i have read and the error message speak about an error in a module for the 2.6.13 kernel ,
(/lib/modules/2.6.13-1.1532_FC4/kernel/arch/i386/kernel/cpu/cpufreq/acpi-cpufreq.ko)
are you sure that you haven't upgradet the kernel image, it seems so but you havent the modules ???

I donn't know if this will help but...
When we changed from a CRT to an LCD we would get the same error everytime XWindows started.
The problem was that our LCD could only handel 1024x768 at 75Hz max.

The fix was to adjust the xorg.conf to

HorizSync 30.0 - 60.0
VertRefresh 50.0 - 75.0

And in Section "Screen" SubSection "Display" Modes
we removed "1280x1024"
leaving "1024x768" "800x600" "640x480"

This fixed our LCD issues.

Good eye, I actually copied and pasted that error messages from another post, but my error is the same, with the 2.6.11 kernel. I edited my original post to reflect that. Sorry.

I have had the same problem with the same environment.

After a lot of searches in google I've found a pdf document mentioning that there are some known bugs/problems in xorg driver for ati radeon 7000-M board: these bugs freeze completely the poweredge due to an heavy extra load by xorg server.

Unfortunately I could not find any official xorg driver for ati radeon 7000-M, the only solution I've found was to install the drivers for RedHatEnterpriseLinux4 provided by dell, you can find them here: http://support.dell.com/
- in the "Popular Support Tools" give your server info in "Select product" and than select "Drivers & Download" in "Select a tool", than click on "GO"
- in the next form select "Red Hat Enterprise Linux 4", specify your preferred language and click on "find download".
- you will next find the link "Video drivers", click on it and you will finally find the ati radeon 7000M driver link. ;-)
- than follow the download/installation instructions;

I suggest you to install the xorg-x11-rhel4_radeon_dell_server-6.8.1-23.DELL.3.EL.i386.rpm package.

Note: although the driver is compiled for RedHatEnterpriseLinux4 it works very well on Fedora core 4.

gundam00,

This sounds very promising! Right now Dell is going to be replacing parts for me. I called them and we troubleshot together, and eventually the tech decided to replace the mainboard, the backplane, the riser and the processors. Those are on the way, but they're on back order. I might try your solution just for posterity.

Thanks very much for your help, and out of curiosity, do you have a link to that PDF document detailing the known issues/bugs with ATI radeon driver and FC4? I would very much like to read it.

I appreciate your reply. Have a good one.

Here is the link:
http://support.dell.com/support/edoc...t/MC774A01.pdf

Search in the document for the keyword:
- "radeon": you will find the first advice;
- "xorg": you will find some extra tips for xorg configuration;

The documentation is not properly aimed on PowerEdge and Fedora, but clearly expresses the problem with ATI radeon 7000-M and early version of xorg/xfree86.

Note: I don't know if the same problem is replicated on other RH/XFree86/Xorg installations.

I am happy you appreciate my post, I hope it will be useful too. ;-)

Good luck.

gundam00,

Your advice worked GREAT on the Red Hat 4 ES install that I had. However, I decided to blow it all away and try it with Fedora 4. It didn't work. So now I've REINSTALLED Red Hat Enterprise 4 ES on the 1850. If this doesn't work (which it should, because it worked earlier today), I'm going to chalk it up to a hardware malfunction and just spit.

Thanks for your help, it proved useful once, and I'm hoping it will prove useful again. This project needs to move forward and this server is severely holding it up.

Have you also tried the configurations suggested in the pdf doc?

Yes, we had edited the xorg.conf file according to Dell's instructions, too. Still didn't work with Fedora.