SlackwareThis Forum is for the discussion of Slackware Linux.
Notices
Welcome to LinuxQuestions.org, a friendly and active Linux Community.
You are currently viewing LQ as a guest. By joining our community you will have the ability to post topics, receive our newsletter, use the advanced search, subscribe to threads and access many other special features. Registration is quick, simple and absolutely free. Join our community today!
Note that registered members see fewer ads, and ContentLink is completely disabled once you log in.
If you have any problems with the registration process or your account login, please contact us. If you need to reset your password, click here.
Having a problem logging in? Please visit this page to clear all LQ-related cookies.
Get a virtual cloud desktop with the Linux distro that you want in less than five minutes with Shells! With over 10 pre-installed distros to choose from, the worry-free installation life is here! Whether you are a digital nomad or just looking for flexibility, Shells can put your Linux machine on the device that you want to use.
Exclusive for LQ members, get up to 45% off per month. Click here for more info.
Ok, severe brain damage from childhood has reared its ugly head. I can not for the life of me get slackware to install ona 20gig drive, nor redhat. However, mandrake does just fine. I have tried to name seperate partitions, etc, reset heads via fdisk, etc, and nothing seems to work. After the supposedly smooth install, it just hangs. Ive tried with and without lilo, etc. any ideas besides therapy would be greatly appreciated
Well I have a problem with my Slackware 8.0 install if I try to set a swap partition directly over a swap partition already created by Partition Magic. The system hang at this time If I try and i have to reboot. Maybe it's the same problem?
everything installs as normal at least as far as the standard routine goes, hwever when it reboots, it dies right after the bios flag. just sits there and does nothing, no messages, no errors, no lilo, nadda
This makes me think that lilo didn't get installed properly. A couple of things to check:
Make sure that in cfdisk that your primary partition is set as bootable.
Here's what I have in my /etc/lilo.conf
lba32 #you probably need this because you have a large drive
boot = /dev/hde #probably /dev/hda for you
...
image = /vmlinuz
root = /dev/hde2 #may be /dev/hda1 for you...depending on where your data is
i finally got it to work, heres how i did it anyway..
slack 8.2 beta
20 gig fujitsu drive [yes, i know...]
asus slot1 muddabored
256 meg ram
3dlabs oxygen dual head video [ uses permidia drivers, do NOT configure independantly]
i made 3 partitions: root 3gig
home 18.5 gig
rest as swap
from this point a did a standard everything type of install, auto , install everthjing, etc...
EXCEPT when it came time to install lilo, i had to install lilo on the MBR, and not the root drive partition. and had to manually enter in
/etc/fstab /dev/hda1, even tho program said it was allready done, i had to do it manually
dont know why, but this is the only way i could get it to work, i did it the day after the original post, and its been running a bot, an ircd and an ftp, all with no probs ever since
hope this helps.....
oh yeah, i did not have to pass the lba32 mode thingy to the kernal, but i think thats because the 8.2 beta uses the 2.4 versus the 2.2 kernel...
"oh yeah, i did not have to pass the lba32 mode thingy to the kernal, but i think thats because the 8.2 beta uses the 2.4 versus the 2.2 kernel..."
The lba32 is a lilo thingy, the kernel figures all this out on its own. I think that this option is installed by default with Slack-8 because I always have to replace it with "linear" to get my SCSI drives to boot.
I think Mandrake is using Grub instead of Lilo these days, which is probably why it works. Maybe I should learn Grub too.
It is probably possible to make Lilo work, but it will require a lot of reading to figure out where it is getting confused.
I know there are probably better ways to this this, but this is how I overcome these difficulties.
Partition the HardDrive so the first partion is 10 MB large, partition the rest of the system as I normally would. Mount that little 10 MB as /boot. Proceed with install, let slack create a boot floppy.. The first time after install, I have to boot with the floppy.
I copy the kernel to my /boot directory (the 10MB partition).
configure LILO so it installs to that partition and boots from that kernel. And voila, no more messing around with Large Drive options and LBA32 (that only works with Post 98 BIOS's, I think)
Originally posted by udtman
20 gig fujitsu drive [yes, i know...]
i made 3 partitions:
root 3gig
home 18.5 gig
rest as swap
Pardon my fussyness, but, how did you make 21.5 gigs (plus swap!) fit on a 20gig drive?
Earlier versions of slack had a version of lilo which didn't like to boot from partitions which were past block 1024. So if you ever put your root(boot) partition past that mark, that may have some impact.
Also, MBR is indeed the necessary place to put the lilo boot. If you only use the root block of the root partition, it won't boot without the aid of a boot floppy or some other sort of boot loader (NTloader etc)
Generally swap (on a machine which is expected to have some sort of load) should be at least twice the size of your physical RAM, some people use up to 3 times.
Also be aware that anything not under /home will be in the / partition in the scheme above. This includes /tmp, /usr, /opt, /var and many others which can become quite large, depending on how you use the system (ie /opt usually has kde, gnome, interbase and many others. /var is where most programs will log to, make .pid files and other such stuff. /tmp is as it suggests a system wide temporary holding place. X uses lots of tmp files, shell users who un-tar their 350MB tar balls etc /usr is where a lot of software is installed. The list goes on.)
If some program starts going crazy and generating 3MB logs every 10 seconds, or a user decides to unpzk that 350MB tarball, or your ftp server logs into it's own directory unders /usr/share/ftp/logs etc your root partition will fill up and anything else that wants to write to that partition will fail and your system could come to a screeching halt, or at least services will start failing.
It's best in a production system to make a separate partition for each of such directories, that way, if only one part fails, the rest of the system should be ok to go along it's merry way.
Of course if this is just a home use comptuer, and 24 hour availability of services isn't crucial, feel free to partition your drive as you like. you probably will anyway
Originally posted by tifkat
Earlier versions of slack had a version of lilo which didn't like to boot from partitions which were past block 1024.
The newer lilo's (including Slack-8's) will work past this as long the PC BIOS can do it in a standardized way. Most PCs made after about 1997 can. You may have to put "lba32" in the global section of the lilo.conf to enable this. (I think I read that the latest version of lilo does this by default)
Quote:
Also, MBR is indeed the necessary place to put the lilo boot. If you only use the root block of the root partition, it won't boot without the aid of a boot floppy or some other sort of boot loader (NTloader etc)
No, (well, and yes). You DO need a loader of some kind on the MBR, but the MBR is normally a bad place for lilo unless Linux is the only OS that will be on this computer because other OS's including the ones made by the Evil Empire tend to insist on having their own loader on the MBR. Fortunately the Micro$oft bootloaders can be set to boot to arbitrary primary partitions (on the primary drive only). So all you have to do is set the active partition in the Micro$oft MBR to boot to your Linux partition, then have lilo on your Linux partition to boot into everything else (including, back into your Micro$oft OS'es if you like). This arrangement works very well.
Quote:
Generally swap (on a machine which is expected to have some sort of load) should be at least twice the size of your physical RAM, some people use up to 3 times.
This is only a crude "rule-of-thumb", and a recommendation on how much RAM you should (or shouldn't) have as well as swap space. In reality, you need to make your best guess about how much total memory (RAM + Virtual) will be needed to run the applications and/or services you need on your computer. Then the rule-of-thumb says to purchase RAM and size swap for the ratios you describe. This rule-of-thumb breaks down when people really load up their machines with so much of today's cheap RAM, that they probably don't need any swap space at all. Another example is the people that are using old or embedded processors that only have 8 megs of RAM and are trying to do a lot. In this later case they may need swap space that is 10 times the 8 megs of memory, and have a machine that really "thrashes", but at least the apps/services will run!
Quote:
It's best in a production system to make a separate partition for each of such directories, that way, if only one part fails, the rest of the system should be ok to go along it's merry way.
I do agree with you here, and for all the reasons you described (and I deleted from this lengthy reply) and more. The key words are "in a production system", in such a system I would recommend either careful planning, or testing some prototypes to come up with what will work the best for your application.
Originally posted by wartstew
The newer lilo's (including Slack-8's) will work past this as long the PC BIOS can do it in a standardized way. Most PCs made
Agreed, but he didn't specify which version of slackware he was having the issue with, nor where he had his boot partition. He only mentioned Slack 8.1 B2 as his working solution.
Quote:
No, (well, and yes). You DO need a loader of some kind on the MBR, but the MBR is normally a bad place for lilo unless Linux is the only OS that will be on this computer because other OS's
Once again, agreed. Although lilo is better to use as the boot loader if you're dual booting Win9x + linux. You only have to worry about things messing up with a fdisk /mbr, and if you're doing that, you should know what you're doing, and if you don't then you're asking for trouble. Of course I am assuming (a very bad thing to do ) that win9x is install first. Installing after you install lilo would end up in a lost lilo.
I did say, or some other sort of boot loader (NTloader etc) so I believe I covered this point It seems however, that he has his box setup as Linux only, so I think we've both covered it well
Quote:
This is only a crude "rule-of-thumb", and a recommendation on how much RAM you should (or shouldn't) have as well as swap space. In reality, you need to make your best guess about how much total memory (RAM + Virtual) will be needed to run the applications and/or services you need on your computer.
Once again, I agree whole heartedly. Although to use more than 3 times your RAM as swap would indicate you need to get more RAM (or a new board which will take more RAM) And of course if you've got 4 GB of RAM 8-12GB of swap would be kind of pointless. Although there are people who are running stuff on multi-gigabyte virtual RAM. This would suck, but it's a lot cheaper than buying Multi-gigabyte RAM (still!) This is mainly research stuff though.
I made the point only to suggest that if he had already assigned 21.5GB of a 20 GB disk, he wouldn't have very much swap left, and should consider a better partitioning scheme.
Quote:
I do agree with you here, and... the key words are "in a production system"
Indeed, and with a home directory of 18GB and an ftp service running, do you think it safe to assume the guy is planning on offering FTP services? Of course he may not be, so it's up to him how to partition his disk Which was my original point So it seems we agree on all the important bits (at least )
LinuxQuestions.org is looking for people interested in writing
Editorials, Articles, Reviews, and more. If you'd like to contribute
content, let us know.