Kejtar
asked on
LILO does not install
I went through the download install of rh6.2 and both times, it neither prompted me for LILO install, nor did it on its own (my only other experience with setting up RH is on a multiple partition drive, where RH prompted me with a question, as to where did I want the LILO to go).
What do I do, and how do I insure that LILO gets installed?
Kejtar
What do I do, and how do I insure that LILO gets installed?
Kejtar
Have you tried running the text based install, rather than the graphic one? I don't know if there will be any difference.
Now that linux is installed, edit your /etc/lilo.conf file so that it contains the appropriate startup options, then run /sbin/lilo to write those changes into your master boot record.
If you need the syntax of lilo.conf, type "man lilo.conf"
Hope this is what you needed.
If you need the syntax of lilo.conf, type "man lilo.conf"
Hope this is what you needed.
ASKER
My problem is, that the computer comes up on reboot: Missing OS!!!! so how can I edit lilo.conf, if it doesn't let me boot it?
Kejtar
Kejtar
What does your lilo.conf look like right now?
Questions - can you tell me the following info please:
1. What your harddrive set up is (SCSI/IDE, dual/single, primary and secondary controller relationships if dual).
2. What brand of harddrive
3. Do you get ANY error messages during the install
4. What your BIOS setup is for the drive/s ?
Best wishes,
modulus
1. What your harddrive set up is (SCSI/IDE, dual/single, primary and secondary controller relationships if dual).
2. What brand of harddrive
3. Do you get ANY error messages during the install
4. What your BIOS setup is for the drive/s ?
Best wishes,
modulus
ASKER
I am reinstalling everything again.... so lilo.conf will have to wait...
It's a p100 ibm machine with built in controller, ( I will have to try to pull specs of their site), with a 1GB HD master single, and on secondary two small drives..... 16MB of RAM (what can I say, it's a test box...).
I will post outcome of workstation type installation shortly. If that goes through without a problem, I will attempt doing the custom install again.
Kejtar
It's a p100 ibm machine with built in controller, ( I will have to try to pull specs of their site), with a 1GB HD master single, and on secondary two small drives..... 16MB of RAM (what can I say, it's a test box...).
I will post outcome of workstation type installation shortly. If that goes through without a problem, I will attempt doing the custom install again.
Kejtar
Linux will only install the root partition (bootable) on a primary drive (ie hda or hdb). Did you try to install Redhat on the primary drive or a secondary drive? If it was secondary, then it would not be recognised.
ASKER
it looks like the booting is done out of hda5 (swap?!?!).... and yes, the disk is the single master on the primary channel.
Kejtar
Kejtar
What do mean by "through the download install"?
ASKER
it looks like the booting is done out of hda5 (swap?!?!).... and yes, the disk is the single master on the primary channel.
Kejtar
Kejtar
ASKER
by the download install I mean install of the ftp: I boot with the boot disk, initialize the nic and set the ftp server, and off I go...
Kejtar
Kejtar
It's possible that you have missed a file or something. This problem does not occur if you install off the cd version.
ASKER
well, the ftp install is just like the install from a cd. What happens is: there is a "CD" mounted as a network drive (or its contents to be precise), and you just access it.... Setup is the same for the rest of it though.
Kejtar
Kejtar
I've done ftp installs and remote mounted NFS installs before and not experienced this. I asked you a series of questions above about your hardware setup because I wondered whether you were not installing on a primary. I see from your references to hda5 that you must have IDE drives , so that rules out the possibility that you have AdaptecSCSI problems ( a frequent problem). However you've said nothing about your BIOS set up, whether or not you're using LBA, accessing your harddrive in linear mode etc. More information about all this is useful if you want help.
Also is this a completely fresh install? Are there any other operating systems on the machine? How did you choose to partition during the install?
Also is this a completely fresh install? Are there any other operating systems on the machine? How did you choose to partition during the install?
ASKER
sorry.... usually I give more information, but since I get only couple minutes at a time to play with the machine, and for most part it is stuck in install mode, I can't pull all the info off the bios.
Anyways: here is portion of it: it is a 1.2 GB IDE HD (make unknown at this point) on an ibm p100 16mb machine. I actually suspect a bad distribution corrupting the install. I know that there is nothing wrong with the machine, becase I was able to do similar install of freebsd on it, and it boots up fine..... Anyways, I will put up a sticky note on the monitor to look at the bios info if this subsequent install fails (yep I am doing it again). BTW, yes it is a clean install: harddrive is squeky clean (I actually deleted all partitions this time).
and during the setup I tried custom, where I configured my own partition info, and I also tried the workstation with autoconfig ( I am trying the workstation autoconfig again).
Kejtar
Anyways: here is portion of it: it is a 1.2 GB IDE HD (make unknown at this point) on an ibm p100 16mb machine. I actually suspect a bad distribution corrupting the install. I know that there is nothing wrong with the machine, becase I was able to do similar install of freebsd on it, and it boots up fine..... Anyways, I will put up a sticky note on the monitor to look at the bios info if this subsequent install fails (yep I am doing it again). BTW, yes it is a clean install: harddrive is squeky clean (I actually deleted all partitions this time).
and during the setup I tried custom, where I configured my own partition info, and I also tried the workstation with autoconfig ( I am trying the workstation autoconfig again).
Kejtar
You may well be right about the corrupted distro. Just one thing though - I note that you say that you are doing an "ftp" install and then you say that you have a cd or it's contents mounted as a network drive. Does that mean that you have:
1. copied the cd contents to the remote network drive and made the contents accesible through the ftp server on the machine
or
2. are actually doing an NFS install of a remote mounted harddrive
Just to make sure there's no simple confusion! Probably a stupid suggestion, but I wondered.
Also, why do it as method 1 rather than 2 if that is what you are doing? Just out of interest - is there some advantage?
1. copied the cd contents to the remote network drive and made the contents accesible through the ftp server on the machine
or
2. are actually doing an NFS install of a remote mounted harddrive
Just to make sure there's no simple confusion! Probably a stupid suggestion, but I wondered.
Also, why do it as method 1 rather than 2 if that is what you are doing? Just out of interest - is there some advantage?
You may well be right about the corrupted distro. Just one thing though - I note that you say that you are doing an "ftp" install and then you say that you have a cd or it's contents mounted as a network drive. Does that mean that you have:
1. copied the cd contents to the remote network drive and made the contents accesible through the ftp server on the machine
or
2. are actually doing an NFS install of a remote mounted harddrive
Just to make sure there's no simple confusion! Probably a stupid suggestion, but I wondered.
Also, why do it as method 1 rather than 2 if that is what you are doing? Just out of interest - is there some advantage?
1. copied the cd contents to the remote network drive and made the contents accesible through the ftp server on the machine
or
2. are actually doing an NFS install of a remote mounted harddrive
Just to make sure there's no simple confusion! Probably a stupid suggestion, but I wondered.
Also, why do it as method 1 rather than 2 if that is what you are doing? Just out of interest - is there some advantage?
ASKER
I am doing and actuall ftp install, and I just referred to the fact, that it should be just like a cd install, just a bit slower... sorry for confusion.
Kejtar
Kejtar
I think you ought to be aware that with Redhat GUI installs a bouncing mouse can cause some of your install setup screens to be skipped. I tried an install with a Microsoft Intellimouse and had a great deal of trouble until I changed the mouse. Whether Bill planned this attack on Linux is unknown. Is it possible the mouse is the problem? You may want to retry the install with another mouse.
ASKER
Mouse is not an issue, and I am not using an intellimouse anyways....
I will try again tonight and see what happens... My theory now is that the wrong partition is being set as an active.
To recap the computer install linux, but on reboot it says missing os, but it boots with a boot disk just fine.
Kejtar
I will try again tonight and see what happens... My theory now is that the wrong partition is being set as an active.
To recap the computer install linux, but on reboot it says missing os, but it boots with a boot disk just fine.
Kejtar
I have had this happen before when I chose the MBR for the boot loader. When I switched to use the first or second partition, it worked, when the MBR didn't. No explanation, but a possible cure.
ASKER
yeah, but I do not get the choice for some reason....
Kejtar
Kejtar
ASKER
victory!!!!!
My problem was that during the initial installation that has failed for me for other reasons, I chose all of the other options. During subsequent installs though, each time that I was asked whether I want the SCSI linear something enabled, I skipped it, which has caused the install not to give me choices as to where to place the LILO.
Now that I have fixed my problem.... let me ask this, and the first person to answer correctly will get the points:
How do I open up the machine to telnet access on the internal network (this is the only one up and running at the time).
Kejtar
My problem was that during the initial installation that has failed for me for other reasons, I chose all of the other options. During subsequent installs though, each time that I was asked whether I want the SCSI linear something enabled, I skipped it, which has caused the install not to give me choices as to where to place the LILO.
Now that I have fixed my problem.... let me ask this, and the first person to answer correctly will get the points:
How do I open up the machine to telnet access on the internal network (this is the only one up and running at the time).
Kejtar
ASKER CERTIFIED SOLUTION
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
ASKER
It is an IDE drive and that is why I did not choose the option enable SCSI linear something (I figured that this is not required). So it just beats me......
Anyways, thanks for the help...
Kejtar
Anyways, thanks for the help...
Kejtar
SCSI linear..hmmm. I'm going to have to find out more about this. AFAIK there is a choice of two modes of determining the layout of a hd : linear and cylinder/head/sector (3D). Which is configured depends on the bios and you can change this in some cases. According to RHat's installation manuals you only need to worry about linear addressing for IDE drives!
The lilo documentation warns that:
"_linear_ Generate linear sector addresses instead of sector/head/cylinder addresses. Linear addresses are translated at run time and do not depend on disk geometry. Note that boot disks may not be portable if linear is used, because the bios service to determine the disk geometry does not work reliably for floppy disks. When using linear with large disks, /sbin/lilo may generate references to inaccessible disk aread (see section 1.1.3), becuase 3D sector addresses are not known before boot time. Linear may conflict with compact."
I think that what this is saying is that there may be a storage of the kernel beyond the 1024 cylinder limit which would lead to a non-bootable system. So it's something you would be better without. I'm guessing that what you did was to turn linear off, because the RH installation manual says that it's on by default. But I'm out of my depth here. You can check by looking at /etc/lilo.conf to see if "linear" is mentioned in it. If it is then it could be that you did not turn of LBA (Linear Block Addressing) in your BIOS before installing and thus you are being forced to use it.
But I guess if it's working for you and your're not worried don't fix it!
The lilo documentation warns that:
"_linear_ Generate linear sector addresses instead of sector/head/cylinder addresses. Linear addresses are translated at run time and do not depend on disk geometry. Note that boot disks may not be portable if linear is used, because the bios service to determine the disk geometry does not work reliably for floppy disks. When using linear with large disks, /sbin/lilo may generate references to inaccessible disk aread (see section 1.1.3), becuase 3D sector addresses are not known before boot time. Linear may conflict with compact."
I think that what this is saying is that there may be a storage of the kernel beyond the 1024 cylinder limit which would lead to a non-bootable system. So it's something you would be better without. I'm guessing that what you did was to turn linear off, because the RH installation manual says that it's on by default. But I'm out of my depth here. You can check by looking at /etc/lilo.conf to see if "linear" is mentioned in it. If it is then it could be that you did not turn of LBA (Linear Block Addressing) in your BIOS before installing and thus you are being forced to use it.
But I guess if it's working for you and your're not worried don't fix it!
ASKER
Thanks, this makes a little more sense now. I will actually try to look into it more, because I want to know what happened I will post here if I find out more...
Kejtar
Kejtar