Hi.

I just tried and installed a second drive, strictly following the FAQ. However, something must have gone wrong, because somehow (I still don't know how) the music partition (and probably more) of my first drive got damaged (damn, now I have to reload all my music). Here is what I did:

1. open empeg
   
2. remove drive cage (or whatever one might call it)
   
3. removed primary drive
   
4. attached new drive
   
5. installed car2-builder.upgrade
   
6. let drive stress test run for a while (1 hour or so), no errors
   
7. installed car2-developer-v1.03.upgrade
   
8. attached old drive again
   
9. rebooted empeg
   
   ---> empeg only found one drive (the old one), but no music partition, was even unable to find the root partition.
   
10. checked all cables again, no problem found here. Checked on the drives jumper settings
    
    ---> damn, I need two jumpers (one per drive) to get them to work together.
    
11. unattached secondary (new) drive
    
12. rebooted empeg
    
    ---> empeg didn't find the music partition, but provided a shell after exiting the player.
    
    ---> the music partition had the size of the new drive's music partition, I don't know how it came to this.
    ---> I tried re-partitioning the music partition only, but failed (it mounted the music partition ok, started the player, but it only resulted in emplode reporting:
    
    The download process failed in stage 2:
    
    Player error 0xfffffff8
    
13. installed car2-builder.upgrade on old (primary) drive


14. installed car2-developer-v1.03.upgrade, then car2-consumer-v1.03.upgrade, both with exactly the same result.
    

Now, any hints what might have gone wrong and how to fix it?

[Additional note: I now tried replacing the first drive completely, but the result was exactly the same as before (i.e. initializing the "new" harddisk (6GB IBM Travelstar that definately works perfectly in my notebook), then installing 1.03developer image resulted in the above mentioned error).]

cu,
sven
(proud owner of a MkII 12GB blue now green, #080000113)

Edited by smu on 11/07/01 12:26 PM.

Hi.

Now I'm replying to my own post. As for the new drive, I was able to solve the issue:
The player software seems to have a problem if /etc/mtab exists (I usually link it to /proc/mounts when using the shell, to make df and others work ok, and usually remove it again before exiting.). I know I had a similar problem before, where the player software would show even more problems. However, I really would like to know why the player software exits the connection to emplode after or during "controlling disk" stage if /etc/mtab is linked like I stated above.

An even more interesting question is how did /etc/mtab survive drive initialization using the drive builder image?
Answer: It looks as it didn't. I now get a stage 4 error 0xfffffffe with my old drive. DAMN, what's going on here?

cu,
sven
(proud owner of a MkII 12GB blue now green, #080000113)

[Additional note: I see you're close to resolving the problem yourself. I don't have any experience with a linux prompt, sorry.]

You did re-set the jumpers again when you changed configuration those couple of times? I remember some drives have 3 positions; single, master, slave. What are the exact models of drives you have?

To my opinion you took a very drastic method (builder.upgrade) to revive your initial drive... Hope you still got your MP3s backed-up somewhere?

Anyways, install the builder.upgrade with the first drive connected (and jumpered accordingly). That prepares it for use in the empeg. Capture the stress test log and post it. Do the same for your second drive. Then you could install the developer 1.03 upgrade with or without any drive connected. After that, jumper the drives again, attach them and boot the empeg. Capture the bootlog and post.

Good luck.

Edwin de Vaan
mk2 rev.7 Trillian # 080000263 6+20Gb blue/red

Edited by edwin on 11/07/01 01:02 PM.

You did re-set the jumpers again when you changed configuration those couple of times? I remember some drives have 3 positions; single, master, slave. What are the exact models of drives you have?

I currently only have the original drive (IBM Travelstar DARA-212000) installed and jumpered correctly. The second drive works well if installed as a single drive, but I currently don't have the appropriate jumpers to set the two drives up as master and slave (or even both to cable select)

To my opinion you took a very drastic method (builder.upgrade) to revive your initial drive... Hope you still got your MP3s backed-up somewhere?
What else should I have done? I was unable to get to a shell to fix it manually. And yes, I have my MP3s backed up somewhere else (backed it all up only two days ago - That was a close call).
Anyways, install the builder.upgrade with the first drive connected (and jumpered accordingly). That prepares it for use in the empeg.
Sure, I just applied it and am currently waiting for it to finish bulding the filesystem.

(...waiting some couple of minutes...)

Capture the stress test log and post it. Do the same for your second drive.

I skip that for the second drive, however the stress test doesn't show anything interesiting (see attachment), appart from the start of the log, just after mke2fs was finished:


done
Writing superblocks and filesystem accounting information: done

real 0m43.729s
user 0m0.280s
sys 0m1.970s
set_blocksize: b_count 1, dev ide0(3,4), block 1!
EXT2-fs error (device ide0(3,4)): ext2_check_descriptors: Block bitmap for group 64 not in group (block 2)!
EXT2-fs: group descriptors corrupted !
mount: wrong fs type, bad option, bad superblock on /dev/hda4,
or too many mounted file systems

What does that mean? I suppose it pretty much says that the filesystem (though just built) is corrupted, right? Does this mean the drive itself might be damaged?

Then you could install the developer 1.03 upgrade with or without any drive connected. After that, jumper the drives again, attach them and boot the empeg. Capture the bootlog and post.
Alright, ignoring the above error, I just rebooted my empeg (with the builder image still active) and the log (available at http://www.incase.de/empeg_builder_bootlog.txt) shows what happened (basically the disc was prepared again, this time without any error as it seems). After installing the 1.03developer image, the old disk works again. Problem solved.

Now I only need to find the right jumpers and I will be able to fill it all up again.

Anyway, thanks for the help.

cu,
sven
(proud owner of a MkII 12GB blue now green, #080000113)