RH 7.1 install problems

karleberts · 07-28-2001, 09:26 PM

First of all, I would like to say that I had RH7.1 installed and running a while ago. Not anymore. I am trying to reinstall it, and everytime I do, a few minutes after it has actually started to install the system (after I answer all of the questions, and after it formats the filesystems) the error:
"install exited abnormally--received signal 15"
comes up. A few more messages come up about unmounting filesystems, killing progs, and such show up, and then it says:
"you may safely reboot your system"
What is this caused by, I am installing on an Emachine 633ids, on a second hard drive, this happens with any methods of installing the system.

I also have another problem, not with installing linux, but caused by my installing linux. I was trying to install it recently (it didn't work) and now whenever I try to start my computer (to winME--the default OS) the system stops at a black screen with an upper case "L" in the upper-left corner. Is there any way of fixing this without resetting windows?

DavidPhillips · 07-28-2001, 09:50 PM

If you are using graphical display press ctrl+alt+F1, or else use alt+F1 and check for errors, then use alt+F2 and alt+F3 etc, to check each console for errors during install, You should see some errors indicating your problem.

Your boot manager for windows has problably been replaced by lilo. No problem if you install lilo to boot windows during your linux install.

Sounds like lilo is trying to start and having problems.

The letters L I L O all indicate stages of the boot sequence as they appear. This would also indicate what your problem is. Which should go away when you solve the first problem.

karleberts · 07-28-2001, 10:07 PM

ok, there are not any errors when I hit alt-f1-3....but when I hit alt-f4 there are several. The last one on the page says:
out of memory killed program 10 (anaconda.real)
anaconda is the install program, right? Is this it?..and if so, how do I remedy this.....this error is on the last line, on the above lines there are messages about a dma error:
"<4> hdd: dma_intr: error=0x51 { DriveReady SeekComplete Error}
<4> hdd: dma_intr error=0x84 { DriveStatusError BadCRC }"
that message appears 5 times.......I don't know what any of this means though, or how to fix it.....anyone else know?

as for the LILO thing, that's kinda what I thought, that makes the linux installation question that much more urgent, because my starting windows depends on it....SOMEONE PLEASE HELP ME!!!

DavidPhillips · 07-28-2001, 10:41 PM

/dev/hdd is the secondary controller slave. What is on it?

If it is you cd then you may have a cd read error.

You may be able to disable dma in the bios.

How much memory do you have?

You should probably have your linux dard drive on the primary slave.
Windows on the primary master drive C:

Have you partitioned the linux drive?

karleberts · 07-28-2001, 11:03 PM

the secondary slave is the hard drive I am installing linux to. I installed it a while back with the specific intent of it being a linux drive. I have 64 megs of memory, which was enough last time I installed RH. I have windows on primary master, my cd on secondary master, and my linux drive on secondary slave. I think the floppy is hooked up to primary slave, but I'm not sure, yes I've partitioned it, it always gets past that part in the install process, it crashes as it is actually writing the programs on the disk.

DavidPhillips · 07-29-2001, 12:15 AM

The error you are getting is with /dev/hdd the linux drive.

Can you fdisk the drive, format it and read and write check it for errors?

Also the dma thing could be a bios setting, that may work in

windows but have problems in linux.

You will not have any luck with it if those hard drive errors are not resolved.

karleberts · 07-29-2001, 09:30 PM

I see that, and yes I can get to fdisk from a rescue image, but I don't know what good that would do. I don't think the DMA thing is a bios prob, because the drive isn't even seen in the BIOS. The bios stopped recognizing the drive when I formatted it for linux, and I haven't messed around in the BIOS after I had RH up and running before.

benwestgarth · 08-25-2001, 08:04 PM

If I am not mistaken, and I may well be, you do not have enough RAM to complete the install. I know that doesn't explain why it installed correctly the first time, but I am almost 100% sure that is what you're problem is as I have had this problem several times installing Redhat on older machines and more RAM always fixes the problem.

Good Luck

Ben

jorrjorr · 10-28-2001, 02:58 PM

I appear to have a similar problem. I've trashed my system so that it no longer boots. However,I was able to boot from the RedHat 7.1 CD and go into rescue mode. There, I fsck'd everything, and made a nice tarball of the stuff I want to keep.

However, whenever I try to copy the tarball to my JAZ disk, it runs for a while, then dies with
an "Out of Memory" and leaves me to reboot.

Is there a way to get the copy to use less memory? I don't care if it copies slowly, as long as it does it right and finishes the job. Should I be doing something silly like telling "dd" to copy one byte at a time?

zhenwu · 10-28-2001, 04:25 PM

Hey there,

do a man hdparm, or search this site for hdparm command help. I was getting exactly the same thing, and when I turned DMA support off (I believe you can feed that to the installer to turn it off on the CD drive, then disable it for the HD once you've finished the install.

There was a great link on this site regarding this issue and solved all my problems. That would be my best bet.

jorrjorr · 10-28-2001, 05:59 PM

No luck. I've tried disabling the DMA on the HD,
then I tried that together with turning off readahead. Still nada. But I've been at it all day, and it's time to get some sort of food. I'll check back tomorrow. (I wrote down the parameters for the SCSI JAZ drive, the IDE HD and the IDE CD, but I have a bus to catch... Otherwise I'd include them here.)

jorrjorr · 10-29-2001, 02:02 PM

As previously threatened (in my last post), the hideous details:
The original settings were:

Code:

hdparm /dev/hda
  multicount   = 0
  I/O support  = 0
  umaskirq     = 0
  using_dma    = 1  (I reset it to 0)
  keepsettings = 0
  nowerr       = 0
  readonly     = 0
  readahead    = 8
  geometry     = 1650/255/63 sectors=1060227 start=63
  model        = WDC AC313200D FwRev = J760A30K

hdparm /dev/cdrom
  model = Toshiba DVD-ROM SD-M1212 FwRev = 1L22

hdparm /dev/jaz
  geometry = 1021/64/32  sectors=2091050 start=0

Oh, and it's a Buslogic SCSI card (I forget which one, and I'm not sitting in front of that machine.)

jorrjorr · 11-01-2001, 10:08 AM

The problem: During a rescue operation, I needed to copy a HUGE
file from my hard disk to a removable JAZ disk. Unfortunately,
while booted up in Red Hat's rescue mode, memory management
seems to have some problems. Every attempt to copy this would
go for a while then run out of memory and force a reboot.

The solution: I hypothesized that if I could slow the machine down, I
might give it time to recover its memory. (I know I forget things
when I think too fast.)

So, how? The following is probably not the most efficient or
safest way to do things, but hey, it worked for me...

1. Break the file into chunks and copy a chunk at a time, with
delays between chunks.
2. The file (a bzipped tarball) was 689459745 bytes long.
3. dd (a file copying program) writes nulls when it hasn't got any
data. So, I couldn't write more data than was actually in the
original file. Otherwise I'd end up with nulls at the end.
4. That means, the blocksize times the number of blocks had to
exactly match the file size.
5. I found a web page with a factoring calculator, and learned that
the prime factors of 689459745 are 3, 5, 13, 19, 379 and 491.
6. Armed with that information, I opted for 379 blocks of 1819155
bytes each.
7. Finally, I wrote a little script:

Code:

          #!/bin/sh
          # SLOW DOWN! Copy slooooowly, and provide a running
          # progress report comparing the file sizes of the
          # two files in question. When done, compute the MD5
          # checksum for each file, for comparison.
          # 
          # NOTE: A 1-second delay wasn't enough. A 10-second
          # delay was, but it was also probably overkill.

          cd /mnt/sysimage/usr/share
          rm /mnt/jaz/home.tbz
          touch /mnt/jaz/home.tbz
          ls -al home.tar.bz2
          ls -al /mnt/jaz/home.tbz
          sleep 1

          for ((blk = 0; blk < 379; blk++))
          do
            dd if=home.tar.bz2   of=/mnt/jaz/home.tbz \
               bs=1819155        count=1 \
               seek=$blk         skip=$blk
            ls -al home.tar.bz2
            ls -al /mnt/jaz/home.tbz
            sleep 10
          done
          md5sum home.tar.bz2 /mnt/jaz/home.tbz

I hope someone else can get some use out of it.