FC5 Hangs on startup after successful install
 

Hey all... thanks for reading.

After UNsuccessfully upgrading from RH9 to FC5, I'm starting fresh but haven't had any luck getting past the

Red Hat nash version 5.0.32 starting

screen. I've tried installing by changing various options (disabling SCSI & using an IDE drive and vice versa; LVM vs. ext3; MBR vs section 1; etc.) Nothing works. I'm reasonably familiar with the lower level linux stuff but not an expert. Does anyone have any suggestions for me?

I've disabled all optional modules during install and everything proceeds as normal (or so it appears) but it won't get past the aforementioned screen.

The box is a dual Xeon IBM M-Series, 1GB RAM w/ on-board SCSI.

Thanks in advance

Have u tried to login in single user mode.
Try and see if it works.

how do I start in single user mode?

1) At the graphical(GRUB) boot screen, press 'e' (for edit), scroll down to select the kernel, and press 'e' again.

2) Press the spacebar, type 'single' or '1' (FC allows 'S' and 's' as well), and press 'Enter'

3) Finally, press 'b' to boot, and you'll boot into Runlevel 1 (single user mode).

Cool... so to answer the aforementioned question, no it won't boot up into single user mode... now i get a wonderful:

Kernel panic - not syncing: Attempted to kill init!

<scratches head> huh. that can't be good. what next?

btw, grub has 2 boot options, one with _smp and the other without. both have the same result.

This has been asked a lot of times in these forums. Please search for it as there could be various reasons for this. Thanks
The one with _smp is for multiple processor or hyperthreading machines. The other is default. If you have a HT or multiple processor machine, you can safely remove the other one.

Why does my machine hang on fedora 5 just as it is about to turn to the logon page?

how do i setup samba

similar problem
 

I am having the same problem as djheath. I installed Fedora 5 and everything worked great. I did a yum update and restarted. But when it started up again, it hung at 'Red Hat nash version 5.0.32 starting'. I tried using grub to boot to single user mode. I edited the line and tried a 1 and a 5. Neither have any effect. The same thing happend, it just hangs. Can anyone help me?

some of the boot screen that hangs...
'
root (hd0,0)
FileSystem Type is ext2fs, partition type 0x83
kernel /vmlinuz-2.6.17-1.2157_FC5 ro root=dev/VolGroup00/LogVol02 rhgb quiet
...
'

also, if I select the other kernel versin under grub (the original kernel installed before the yum upgrade), everything works fine...

I never was able to get FC 5 to work... had to go to plan B which was RHEL 4 but still had considerable problems and eventually got it resolved. Kind of a bummer (http://www.ellenfeiss.net/commercials.php)

I am having the same (or at least similar) problem as both Jester and djheath.

In case hardware is an issue, I am running FC5 on an Asus A7V333 with an Athlon XP 1800+ CPU, 1GB RAM and two 320GB hard drives running in RAID 0 using the on-board Promise controller. The RAID array has a small (100MB) /boot partition, and the rest is allocated to a single LVM partition that contains 3 logical volumes.

After several unsuccessful attempts to get Ubuntu to install onto the RAID array, I decided to try Fedora, since it includes native support for the controller. I had already partitioned the drives using gparted as a part of the process of trying to get Ubuntu installed (3 primary partitions and an extended partition that contained a single logical partition), so I just reformatted all of the existing partitions rather than repartitioning for the Fedora install. The install went smoothly, and in short order I had a fully functional Fedora system up and running.

Then I made my first mistake. I was extremely tired at this point, having already spent the better part of a day trying to get Ubuntu working, so when I ran pup I just decided to update everything. Several hours later Fedora finally finished downloading all of the updates. Since apparently some of the things I had updated required a system reboot in order to be finalized (including, unbeknownst to me at the time, a kernel upgrade), I rebooted the system to complete the process.

GRUB loaded, auto-started the new kernel after a short delay and then...nothing. I saw the same Red Hat nash version 5.0.32 starting message that djheath mentioned in his first post, a string of Buffer I/O errors that I assume are caused by the kernel trying to access the array before loading dmraid, and that was it. The system was not frozen; tapping lock toggle keys would turn the appropriate LEDs on and off on the keyboard and I was able to reboot it using the infamous 3-finger salute, but it simply would not continue the boot process.

Thinking I may have hosed the partitioning process when using gparted and that something in the partition table was now causing problems, I decided to start of from scratch. First I rebuilt the array, in case gparted had somehow managed to wreck it. Then I reloaded the DVD and installed FC5 all over again. I also modified my previous partition scheme based on the default one anaconda wanted to use. Rather than use several primary partitions, I used the LVM partition scheme described above.

Again the install went smoothly. This time I paid much closer attention to what was being updated. Anything I didn't think needed updating, I left alone. I passed over the kernel update at first. I updated everything else that needed updating, then made sure that there were no issues getting the box rebooted and back running. Once that was complete I updated the kernel all on its own, to make sure no other updates were conflicting with it.

Once again, after updating the kernel, the box will not boot. I see the exact same thing I saw before. Red Hat nash version 5.0.32 starting followed by a series of Buffer I/O errors (although different logical blocks are mentioned since the array was rebuilt and the partition scheme changed) and then nothing.

The system boots fine using the kernel from the CD (vmlinuz-2.6.15-1.2054_FC5) but refuses to boot with the new one (vmlinuz-2.61.17-1.2157_FC5).

I don't think the Buffer I/O errors are related, because I see the same errors when booting into the original kernel. When booting to 2.61.15, the next thing I see after the I/O errors is:
This message does not appear when booting to 2.61.17. The cursor just continues to flash after the I/O error messages.

I disabled the quiet option on 2.61.17 and it shows the following after the I/O errors:
Again, after the final line, the boot process just seems to stall.

This is the only problem (so far) with configuring the system that has me completely stumped. There have been a few other hiccups on the road to getting some of the services running the way I want them to, but I've managed to get everything sorted out, save this. If anyone could shed any light on the issue, or even point me in a direction that would help me determine exactly what the issue is, it would be greatly appreciated.

That could have been me, who wrote this, except that you seem to have some more Linux experience.

I spent some 4 weeks now getting it running. Especially the kernel upgrade thing. Until there will be a solution I connected my HD to the IDE controller, not the Promise.

A very difficult thing was the Nvidia driver. Now I'm dealing with the c-media sound module.

Any other hints?

Regards.

B.

I have the same problem as Cartoonite... Boots fine on vmlinuz-2.6.15-1.2054_FC5 but hangs on the redhat nash part with the vmlinuz-2.6.17-1.2157_FC5.

dang it...

If I'm understanding this, you were experiencing the same no-boot issue after the kernel upgrade, but now that the HDD is connected to one of the IDE controllers the new kernel boots fine, is that correct?


Based on what's been posted so far, it seems that this is a symptom common to people attempting to upgrade the kernel on system booting from a RAID array. Is your system using a RAID array or is(are) the HDD(s) connected to an IDE controller? If RAID, what type of controller is it?

I'm not going to claim I have a solution, or even that I'm close to finding one, but the more information we have regarding the specific nature of the problem, the easier it should be to find one.

Edit: Another thought - has anyone tried upgrading the kernel on a system running 3 or more drives where there is a RAID array present but the system boots off of an IDE drive? Additionally, for bolognese and corefile (if applicable), what type or RAID array are/were you using, RAID1 (mirror) or RAID0 (stripe)?

This is precisely the problem I'm having. I used Stanton-Finley's tutorial on installing my nvidia drivers earlier today. That site recommended upgrading to the newest kernel first, so I did that. However, the newest kernel will not boot on my system.

My setup:

/dev/hda1 - common swap
/dev/hda2 - / (root Ubuntu 6.06.1)
/dev/hda3 - / (root Fedora Core 5)
/dev/hda5 - /home (Ubuntu)
/dev/hda6 - /home (Fedora)

/dev/sda1 - Win XP (NTFS)
/dev/sda5 - Storage (FAT32)

I'm using Ubuntu's GRUB. The /dev/sda drive is RAID 0 using Promise FastTrack 376 software drivers w/ onboard RAID controller (MSI)

The older kernel (the one from the install disks) on Fedora boots fine. Coincidentally, or maybe not, I'm also having a strange bootup problem with Ubuntu on cold starts only (after the system has been powered down for a few hours), never on warm reboots though.

I'd like to know what the trouble is with this newer kernel in Fedora.

Regards,

~Eric

EDIT: I've been reading about problem with the 2.6.17... kernel all over the place tonight. There are also numerous bug reports on Redhat's Bugzilla site about this very issue. I guess I'll just use my trusty ol' 2.6.15-1.2054_FC5 for now. ;)

Hi there,

Here is even more strange stuff to think about.

I messed up my whole installation. Unfortunately I lost my Grub installation and did not know what to do anymore, even being unable to boot into windows. So I had to install the whole thing again, without Windows, so you will find me some more frequently on this forum :study:

I tried to install a copy of Mandrake, because I had some positive experience with it a couple of years ago, the 8.1 version. But after installing, first boot...black screen with some lost coloured pixels...trouble with my nVidia "old" TNT2 (already a known bug). So back to Fedora again.

What happened this time:
I left the harddisk in the normal IDE controller, installed FC5 and everything went fine. Then I updated all the stuff that was avaiable, including the latstes kernel 2174. Guess what:( No boot. So I plugged the cable into the Raid controller and.......BINGO :scratch: I was able to boot.

Hope that I gave you enough info to figure things out. Don't rely on me...I'm only glad if I can be of any help.

I have a similar problem with "Red Hat nash version 5.0.32 starting" on FC5.
My system hangs but only for a minute or two. The IDE LED on the front case never turns off after this happens (never). First I had a single IDE drive. I found that by disabling SATA drives completely solved the problem. Did not matter because I had none.
Then I installed a SATA drive and enabled them and the problem returned.
The problem only shows up after upgrading kernel, 2.6.15-1.2054_FC5 to 2.6.17-1.2157 ( in my case ) the newer kernel does the same thing.

At least I do eventually boot.

I have tried various bios settings without success, AHCI - Emulated PATA -On Board Sata boot rom enabled and disabled.

ASUS A8R-MVP, Athlon64 3800+, ULI Sata. No raid dirves installed

As of a few days ago (maybe even just yesterday) there is a new version of the kernel available (2.6.17-1.2174_FC5) but I've found that this does not improve my situation at all. I experience the same issue with the latest kernel as I did after first upgrading to 2.6.17. I tried just leaving the system alone for a few minutes to see if it would eventually continue to boot, but after 4+ minutes of inactivity I decided it wasn't going anywhere and just rebooted into the original kernel again.

I've tried disabling the quiet option when booting into 2.6.15 to see if I can get a better idea of exactly what step is supposed to happen after the last message I see, but the screen gets flooded with SELinux messages and by the time they've slowed down the messages I need have been pushed way off-screen. I guess my next step will be to try temporarily disabling SELinux in an effort to remove those messages from the console so that the messages I'm looking for don't get pushed off the top.

Edit: I've been reading Red Hat's bugzilla and think I may have found a fix. I'll test it once I get home and post back with my results.

Even stranger: I destroyed my computer for some unknown reason, so I had to start a complete new install. I left my HD to the IDE (not raid) controller. When I upgraded the kernel, once again I wasn't able to get it booting. The solution (workaround?) was a move of the HD, this time from the IDE to the raid controller.

So far, so good.

I'm now up and running with 2.6.17-1.2174_FC5. Here is the fix that I found in Bugzilla, just be aware that it does involve installing some non-release packages from the development repository:

This will update mkinitrd to the latest development build (currently 5.1.8-1). It will also update a couple of lib packages as dependancies.

Once the updated packages are installed, you need to back up your current kernel image file and then create a new one with the newly installed mkinitrd.

Be aware that these commands are specific to the kernel version that you have installed so if you are not using 2.6.17-1.2174_FC5 you will need to modify them slightly to reflect your kernel build.

After that, you should be good to go. Reboot the system and choose the 2.6.17 kernel from your bootloader.

Credit for this fix goes to Chris Adams who originally posted it and Andreas O. whose steps I modified slightly to get the code listed above. The only difference is that he temporarily enabled the development repository by editing the repo file in /etc/yum.repos.d as opposed to using a command line switch.

With regard to your comment about waiting a few minutes for the boot
to continue... you might try waiting longer. I thought my reboot
was hung after upgrading to the latest FC5 kernel. On a whim I just left
it for awhile, and came back after 10-15 minutes to a booted system.
Subsequent boots didn't take so long, so I don't know what might have been
happening when it seemingly hung.

Thanks for the tip, Rob, although since my system is already running under the new kernel, it doesn't really apply to me.

Still, it is something for others who have not yet updated mkinitrd or who feel uncomfortable installing non-release packages on their systems.

This is my case.
I use FC 5 on 3 IDE drives. 2 drives make a RAID 0 array (strip) and the /boot partition is on one of this drives.
The 3rd drive is used as a storage.
The controller is external RAID 0/1 controller with mark SUNIX with Silicon Image chipset.

FC 5 works just fine. I updated the kernel and now have 2.6.17 with nvidia GeForce 2 MX 400 video card. My RAID array is not made by the controller. It is a MD raid array on 2 of the partitions of this drives.

I say again, FC 5 works fine. Just wanted to say my successful case.

Ok, I found out exactly where my 2 minute boot delay was happening. I booted with quiet mode off and I found the answer. I have 4 - sata ports with a 36GB WD Raptor connected to sata1. When the kernel starts probing the sata ports it probes all four, only 1 responds the other 3 take forever to timeout, see the qc timeout message below.

ata1: SATA max UDMA/133 cmd 0xFFFFC20000004900 ctl 0x0 bmdma 0x0 irq 193
ata2: SATA max UDMA/133 cmd 0xFFFFC20000004980 ctl 0x0 bmdma 0x0 irq 193
ata3: SATA max UDMA/133 cmd 0xFFFFC20000004A00 ctl 0x0 bmdma 0x0 irq 193
ata4: SATA max UDMA/133 cmd 0xFFFFC20000004A80 ctl 0x0 bmdma 0x0 irq 193
ata1: SATA link up 1.5 Gbps (SStatus 113)
ata1: dev 0 cfg 49:2f00 82:346b 83:7f21 84:4003 85:3469 86:3c01 87:4003 88:007f
ata1: dev 0 ATA-6, max UDMA/133, 72303840 sectors: LBA48
ata1(0): applying bridge limits
ata1: dev 0 configured for UDMA/100
scsi0 : ahci
ata2: SATA link down (SStatus 0)
ata2: qc timeout (cmd 0xec)
ata2: dev 0 failed to IDENTIFY (I/O error)
scsi1 : ahci
ata3: SATA link down (SStatus 0)
ata3: qc timeout (cmd 0xec)
ata3: dev 0 failed to IDENTIFY (I/O error)
scsi2 : ahci
ata4: SATA link down (SStatus 0)
ata4: qc timeout (cmd 0xec)
ata4: dev 0 failed to IDENTIFY (I/O error)
scsi3 : ahci
Vendor: ATA Model: WDC WD360GD-00FN Rev: 35.0
Type: Direct-Access ANSI SCSI revision: 05

I don't understand why it calls sata1 scsi3

Hope this helps

I finished my upgrade of FC5 yesterday. It installed without a hitch. Upon boot it froze at Start Service Local. I rebooted several times and can not get passed this. I eventually entered interactive mode and said No to every service. Right after that it freezes. Probably video related. I left it in that frozen state all night... still hung. I can not switch to other TTY1-6 either. Interestingly enough... sshd service starts, so I am able to ssh into my box from my windows PC and do everything from ls - yum update. I tried to look for a process that i could kill, but really don't know what i am doing.

Any help would be greatly appreciated. I have mythtv, and the mythbackend is running, and i really would like to get back in business!!!