[Rule-list] Testing 0.7.1 with 8 MB - almost!

Devon devon at tuxfan.homeip.net
Tue Feb 26 03:25:09 EET 2002


-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Monday 25 February 2002 06:55 pm, Martin Stricker wrote:
> I was able to almost install with RULE-installer 0.7.1!

I did eventually get it installed. It was a long slow process. To make a 
little extra room for the rpm transaction, I ended up removing all unused 
kernel modules. (reiserfs, msdos, vfat, fat, usb-storage, tulip, 
raid[0,1,5], scsi releated modules) It took many hours to finish 
installing the RPMs.

> The steps of the installer are really slow, getting to the next screen
> takes about 1.5 to 2 minutes. I'm using the low memory installation
> method as usual. Partitioning with fdisk and recycling the partitions
> from the last successful install on friday. The installer now asks to
> format swap (I'll loose all data - will this detroy the used swap?) and
> to use it immediately (really stupid, isn't it? ;-)) ). 

It only seems stupid. :) The installer doesn't know you already enabled 
the swap partition, and tries to run mkswap and swapon on that partition.

> Then it tells
> me "The kernel was unable to re-read the partition table on /dev/hda"
> yaddayadda please reboot yadda yadda. Continue... Now I get a strange
> error message scattered across the screen:
> Traceback (innermost last)
> File /usr/lib/anaconda/text.py, line 298,
> {probably missing some letters here} n partedExceptionWindow
> return buttonToAction[rc] KeyError:Übergehen
> Übergehen is the German word for continue or skip (I'm using German as
> install language). Again I get "Kernel is unable to re-read..." Moving
> forward. Parts of the error message stay on the screen until the end.

Hrmm, I thought I handled catching the Exceptions, must have missed one.

>         total   used   free shared buffers
> Mem:     5956   5448    508      0    2652
> Swap:  193528  39180 154348
> Total: 199484  44628 154856
> Nice swap amount... 

I wouldn't call that a 'nice' swap amount. The system is trying like 
crazy to keep enough memory available to run. If you ran top, you would 
see that kswapd is using close to 100% of the cpu.

> 5h20min: The progress bar for the first packet is at one-third when the
> installer finally crashed:
> install exited abnormally -- received signal 11
> Again the error message is scattered across the screen. I made a photo
> but have to find a scanner to attach it to a future mail. VT2 and VT3
> show no unusual information, but VT4 is full of these error messages:
> <4>Error -3 while decompressing!
> <4>c0269aa6(2280)->c0751000(4096)
> ^^^^^^^    ^    ^^^     ^^^^^^^^^
> These two lines together fill the whole screen. I jotted the last one,
> the second line does differ slightly in each repeat, the characters
> marked with a carret ^ are the same in every visible message.
>
> I don't know what caused this, maybe a memory/swap error, maybe
> something else.

My first attempt at 8M ended the same way, which is why I cleared out all 
the modules for the next attempt. I haven't looked into the error message 
at all, but I am guessing that the system finally ran out of resources. 
It just couldn't swap out enough to make the needed memory available.

> If the install had succeeded it might have needed more than 24 hours...
>
> Very good work, Devon! We're getting close! :-)

Maybe. ;)

I'll bring up some specific points in a new message.

- -D

- -- 

pgp key:  http://www.tuxfan.homeip.net:8080/pgpkey.txt

- --
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.6 (GNU/Linux)
Comment: For info see http://www.gnupg.org

iD8DBQE8euP1eMAUbzJhSVcRAhJfAKC0TSTGqvUu3VLu6hA02TOcc0lo6gCglx+s
bc7UNWVJelBp4MI8+I/S1AQ=
=rGyj
-----END PGP SIGNATURE-----

_______________________________________________
Rule-list mailing list
Rule-list at mail.freesoftware.fsf.org
http://mail.freesoftware.fsf.org/mailman/listinfo/rule-list



This full static mirror of the Run Up to Date Linux Everywhere Project mailing list, originally hosted at http://lists.hellug.gr/mailman/listinfo/rule-list, is kept online by Free Software popularizer, researcher and trainer Marco Fioretti. To know how you can support this archive, and Marco's work in general, please click here