<solved> added a 5th OS and now having boot issues

[Nish]

I do solemnly swear that if somebody helps me fix this I won't ever play with another OS as long as live.

Not the usual boot issues where new OS takes over mbr.  This OS I added was to an existing partition on my first hard disk that had already had an OS I played with on it.  In my PCLOS grub boot that is hd0,5 (sda6) and I chainload.  I have had a lot of play OS's on there including other ones that use new grub.  When I install the other OS I told it not to put grub in mbr but put it on partition dev/sda6.

Everything seemed to go OK and on reboot I choose that partition in grub to boot but got an error 13.  Invalid or unsupported executable format.  Oh well I escaped and booted my PCLOS which is on sdb2.  It took a long time to boot and when I got in I looked at the structure of the files in the new OS boot and things seemed OK. Not sure why it wouldn't boot, I tried it again.  Same error.

And PCLOS took a long time so I went to verbose and saw the messages "waiting for sdb1 to appear, timeout one minute" ditto for sdb2. Then after awhile it gives up and said "could not find resume UUID..." goes by too fast for me to catch it but I assume the same message for the other disk.  Did the other installer change the UUIDs of my drives?   How do I find out and fix?
OK, ran blkid and there was no change in UUIDs compared to my fstab and grub files.  I did update fstab for that partition to be ext4
UPDATE: hey, you can see I am trying to work this out while having a NBD.  I found googling that resume id is the swap file and noted that the swap UUID did change, so hey, fixed grub with the new UUID for all the resumes and although I no longer have the resume id error, I am still unable to get my sdb1 and 2 to say "HI!" in a timely manner.

I have played around with other OS's on that partition before with no problem as long as I chainload is OK. And I think they all reformat the swap even though it is already linux swap and never had this happen.  The other OS's installer only formatted the sda6 partition (ext4) and the sdb1 (swap, they all do that swap is swap, right?)

here is the output of fdisk -l

Disk /dev/sda: 500.1 GB, 500107862016 bytes
255 heads, 63 sectors/track, 60801 cylinders, total 976773168 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x26132612

   Device Boot      Start         End      Blocks   Id  System
/dev/sda1   *          63   512007614   256003776    7  HPFS/NTFS  windows
/dev/sda2       512007676   976751999   232372162    f  W95 Ext'd (LBA)
/dev/sda5       512007678   720756224   104374273+  83  Linux  sabayon
/dev/sda6       720756288   976751999   127997856   83  Linux  debian squeeze - what I just installed

Disk /dev/sdb: 500.1 GB, 500107862016 bytes
255 heads, 63 sectors/track, 60801 cylinders, total 976773168 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x0001b338

   Device Boot      Start         End      Blocks   Id  System
/dev/sdb1              63    16868249     8434093+  82  Linux swap / Solaris
/dev/sdb2        16868250   836070794   409601272+  83  Linux  PCLOS
/dev/sdb3       836070795   976768064    70348635   83  Linux   Testing 64bit

This is just my external
Disk /dev/sdc: 320.1 GB, 320072933376 bytes
255 heads, 63 sectors/track, 38913 cylinders, total 625142448 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x44fdfe06

   Device Boot      Start         End      Blocks   Id  System
/dev/sdc1              63   307692944   153846441    7  HPFS/NTFS
/dev/sdc2       307692945   625137344   158722200   83  Linux

[djohnston]

Well, Debian is using GRUB2 now. But, you probably already know that. Without looking at GRUB right now, let's just concentrate on sdb1 and sdb2. Reformat the swap partition and do a filesystem check on sdb2. Boot from a live CD. Open a terminal and su to root. Might want to run fdisk-l just to be sure you're looking at the right partitions.

su
fdisk -l
mkdir /mnt/here
mount /dev/sdb2 /mnt/here
touch /mnt/here/forcefsck
umount /mnt/here
mkswap /dev/sdb1
exit
exit

Reboot to the hard drive boot menu and select PCLinuxOS 32 bit. Let the filesystem check run.

[Nish]

It may be a lost cause at this point.  I did start out troubleshooting this on my own - ok, always a bad idea. went googling and tried mkintrd (which was already there for the last kernel so a lot of good that did)  And I did boot up a liveCD to check the disk and the really scary thing was, it was seeing the two hard disks switched.  IOW, in PCLOS live the sda partitions were showing on sdb and vice versa.  I booted it twice.  I don't know what the heck happened, I made no changes to hardware - no jumper settings changed and no BIOS changes.  Was afraid to check disk or format or anything.

Decided not to go there and considered this install a write off, copied off my data to an external and put a fresh install of LXDE (which did see the partitions correctly) on the first disk.  Even made a swap drive on the first disk for it. 

Now, all I want to do is reformat that second disk.  I ask you, how do I ensure that this LXDE install is using the proper swap partition (the one on sda)?  Because I am going to blow away that whole second drive and start fresh with something or another.  Not sure if I will get used to LXDE so would be nice to put my PCLOS back on it once it is cleaned up and sees things the way they really are.

[djohnston]

I ask you, how do I ensure that this LXDE install is using the proper swap partition (the one on sda)?

cat /proc/swaps

[Nish]

It is showing both of them:
/dev/sda5                               partition   9101308   0   -1
/dev/sdb1                               partition   8434088   0   -2

I have checked in grub menu.lst here in new install and it has the UUID of the new (sda5) swap partition for all the resumes.

I have looked in fstab and it is showing both of them.  Is it sufficient to just comment out that bad one in fstab? Or do I need to do something elsewhere.

Thank you! 

[djohnston]

I have looked in fstab and it is showing both of them.  Is it sufficient to just comment out that bad one in fstab? Or do I need to do something elsewhere.

Thank you! 

Either comment the line in fstab for the unwanted swap partition, or remove it.

You're welcome. 

[Nish]

No hope for my poor battered going-on-its-second-year install, hmmm?     ancient in linux install years the way I mess them up.

I could try to boot it before I euthanize it

hehe

[djohnston]

No hope for my poor battered going-on-its-second-year install, hmmm?     ancient in linux install years the way I mess them up.

Dunno. What issue(s) are you having now?

[Nish]

I am in the doomed install now.  I reformatted the swap (sdb1 on the doomed drive), changed the UUID for resumes in its GRUB (I am loading this one from the new install of LXDE on the first drive). I set it to run the fsck on boot and it did.  OK, didn't get anything interesting except for root filesystem (sdb2) have 1.% noncontiguous blocks and the spare partition having 0.5% noncontiguous blocks.

It runs once I get it booted.  But I am getting that waiting for sdb1 and 2 to appear timing out.

Oh, there is no mention of swap in my fstab on this install.  Don't know if that makes any difference.  cat /proc/swps brings up neither swap drive
[root@AMD nish]# cat /proc/swaps
Filename                                Type            Size    Used    Priority
[root@AMD nish]#

[djohnston]

What were you posting from before? Results of:

# cat /etc/fstab
# fdisk -l

Any swap partition there?

[Nish]

When I had results for swap I was posting from a fresh install of LXDE version of PCLOS.  The second (no results for swap) is from this install that is screwed up (PCLOS KDE, old but so am I).

cat /etc/fstab returns:
# Entry for /dev/sdb2 :
UUID=8c42e187-fc24-4216-b576-d79d50381c59 / ext3 defaults 1 1
# Entry for /dev/sdb3 :
UUID=5c07fd18-3a4f-4c00-999e-65e78d0ec473 /64bit ext3 user 1 2
# Entry for /dev/sda6 :
UUID=ecc26673-0f33-4d67-a422-ca3244ce4acd /debian ext4 defaults 1 2
none /dev/pts devpts defaults 0 0
none /proc proc defaults 0 0

Code: [Select]

Disk /dev/sdb: 500.1 GB, 500107862016 bytes
255 heads, 63 sectors/track, 60801 cylinders, total 976773168 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x0001b338

   Device Boot      Start         End      Blocks   Id  System
/dev/sdb1              63    16868249     8434093+  82  Linux swap / Solaris
/dev/sdb2        16868250   836070794   409601272+  83  Linux
/dev/sdb3       836070795   976768064    70348635   83  Linux

Disk /dev/sda: 500.1 GB, 500107862016 bytes
255 heads, 63 sectors/track, 60801 cylinders, total 976773168 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x26132612

   Device Boot      Start         End      Blocks   Id  System
/dev/sda1   *          63   512007614   256003776    7  HPFS/NTFS
/dev/sda2       512010198   976768064   232378933+   5  Extended
/dev/sda5       512010240   530212863     9101312   82  Linux swap / Solaris
/dev/sda6       530214912   940910984   205348036+  83  Linux
/dev/sda7       940911048   976768064    17928508+  83  Linux

Disk /dev/sdc: 320.1 GB, 320072933376 bytes
255 heads, 63 sectors/track, 38913 cylinders, total 625142448 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x44fdfe06

   Device Boot      Start         End      Blocks   Id  System
/dev/sdc1              63   307692944   153846441    7  HPFS/NTFS
/dev/sdc2       307692945   625137344   158722200   83  Linux
[root@AMD nish]#


back from some more googling and found a post here in PCLOS - now I realize where I went wrong with the make a new mkinitrd  - I was supposed to rename the one in there to old first.  OK, got a little error after giving the command but I am sure it is nothing:
cp: cannot stat `/usr/share/icons/large/mandriva.png': No such file or directory

on reboot, things went much faster (too fast to catch it all) but I can see it is still hanging up on a handshake with sdb2 (this install).  There doesn't seem to be a problem with sdb1 (swap for this install) except, this install doesn't seem to notice where its swap is, darn  something is wrong but I think I am close

[root@AMD nish]# cat /proc/swaps
Filename                                Type            Size    Used    Priority
[root@AMD nish]#

There is also OP's advice regarding rename the swap with a label rather than a UUID? 

[djohnston]

I forgot to ask for the UUIDs. First, do:

# blkid /dev/sdb1

Add the following line to the end of fstab:

UUID=497fb500-4373-4932-9c0e-84b36eb00ef6 swap swap defaults 0 0

Susbtitute your sdb1 UUID for the one listed above. Save fstab and reboot.

[Old-Polack]

I am in the doomed install now.  I reformatted the swap (sdb1 on the doomed drive), changed the UUID for resumes in its GRUB (I am loading this one from the new install of LXDE on the first drive). I set it to run the fsck on boot and it did.  OK, didn't get anything interesting except for root filesystem (sdb2) have 1.% noncontiguous blocks and the spare partition having 0.5% noncontiguous blocks.

It runs once I get it booted.  But I am getting that waiting for sdb1 and 2 to appear timing out.

Oh, there is no mention of swap in my fstab on this install.  Don't know if that makes any difference.  cat /proc/swps brings up neither swap drive
[root@AMD nish]# cat /proc/swaps
Filename                                Type            Size    Used    Priority
[root@AMD nish]#

This is where you need to make a new initrd image for your present default kernel. Without that, the present initrd image is still looking for the old swap partition UUID, hence the wait before it searches the resume= line in /boot/grub/menu.lst and the swap entry in /etc/fstab.

[Nish]

OK, will do.  And just noticed something I thought I should check with you first.  My new swap on sda and the old one on sdb both have same labels - is that OK?  I never gave them labels I guess they normally call them swap anyhow.  Maybe that is what OP was talking about in the post I am using for a guide.  Anyway, OP, I did make a fresh initrd and things are better just not perfect yet

Code: [Select]

[root@AMD nish]# blkid
/dev/sdb2: LABEL="Linux" UUID="8c42e187-fc24-4216-b576-d79d50381c59" TYPE="ext3"
/dev/sdb1: LABEL="swap" UUID="96630cec-04b4-4526-a344-f578130defd9" TYPE="swap"
/dev/sda1: UUID="A0F0F425F0F40378" TYPE="ntfs"
/dev/sdb3: LABEL="Spare" UUID="5c07fd18-3a4f-4c00-999e-65e78d0ec473" SEC_TYPE="ext2" TYPE="ext3"
/dev/sda6: UUID="fe6c6834-c4a5-48e8-a6fd-b2460b5f83fd" TYPE="ext4" LABEL="PCLOS64"
/dev/sda7: UUID="0fdccad5-b3c4-4029-8351-a273d55e01f8" SEC_TYPE="ext2" TYPE="ext3"
/dev/sda5: LABEL="swap" UUID="5277c74f-ad1a-424a-91a7-1df2658dd2c0" TYPE="swap"
/dev/sdc1: LABEL="for_windows" UUID="3A8CC9623518C3AD" TYPE="ntfs"
/dev/sdc2: LABEL="for_linux" UUID="cc19e321-ce15-4d36-a2d3-62cef923636c" SEC_TYPE="ext2" TYPE="ext3"


[djohnston]

You still have two swap partitions?   Really not necessary. But, if you want to differentiate them with their own labels,

# mkswap -L LabelName /dev/sdXX

I guess you could make the label names SWAP1 and SWAP2, or whatever you choose.

If they do have their own labels, you can change the entry in fstab.

LABEL=SWAP1 swap swap defaults 0 0