Read_Icculus

After booting it is reported to be mounted, (according to mount) but going to /mnt/external and ls-ing shows nothing, there's nothing there according to df as well, and trying a variety of other pokings about, like "file:///mnt/external" in mozilla, shows nothing, no folders/files. After doing "umount /mnt/external ; mount -t ext2 /dev/sda1 /mnt/external" - ls /mnt/external, and df see the mounted drive's data. I guess I could add the "umount;mount" stuff in a boot script, but I wonder what's going on.

I was fiddling about and after booting I saw "mount" listed external as being mounted, so I did "umount /mnt/external", and then "mount -t ext2 /dev/sda1 /mnt/external", and then it's mounted. So I guess fstab isn't mounting it right? /dev/hda5 / ext3 defaults 1 1 /dev/hdc /mnt/cdrom auto umask=0,user,iocharset=iso8859-1,codepage=850,noauto,ro,exec,users 0 0 /dev/hdd /mnt/cdrom2 auto umask=0,user,iocharset=iso8859-1,codepage=850,noauto,ro,exec,users 0 0 /dev/sr0 /mnt/cdrom3 auto umask=0,user,iocharset=iso8859-1,codepage=850,noauto,ro,exec,users 0 0 none /mnt/external supermount dev=/dev/sda1,fs=ext2:vfat,--,umask=0,iocharset=iso8859-1,sync,codepage=850 0 0 none /mnt/floppy supermount dev=/dev/fd0,fs=ext2:vfat,--,umask=0,iocharset=iso8859-1,sync,codepage=850 0 0 /dev/hda1 /mnt/windows vfat umask=0,iocharset=iso8859-1,codepage=850 0 0 none /proc proc defaults 0 0 /dev/hda6 swap swap defaults 0 0 Disk /dev/hda: 120.0 GB, 120000000000 bytes 16 heads, 63 sectors/track, 232514 cylinders Units = cylinders of 1008 * 512 = 516096 bytes Device Boot Start End Blocks Id System /dev/hda1 * 1 40641 20482843+ c W95 FAT32 (LBA) /dev/hda2 40642 232514 96703992 5 Extended /dev/hda5 40642 231189 96036160+ 83 Linux /dev/hda6 231190 232514 667768+ 82 Linux swap Disk /dev/sda: 120.0 GB, 120034123776 bytes 255 heads, 63 sectors/track, 14593 cylinders Units = cylinders of 16065 * 512 = 8225280 bytes Device Boot Start End Blocks Id System /dev/sda1 * 1 14593 117218241 83 Linux I checked "fdisk -l" output before I got it mounted and it was the same as this. Also just ran into a problem with mounted DVDs, I just burnt a data DVD, put it in my faster DVD-ROM drive to check the md5sums of the written data. Everything was fine, so I took it out and started to burn a new data DVD and I get a k3b error "cannot umount cdrom drive" or something similar, didn't save output. So I check what's mounted and df, and both DVD drives were showing that they were mounted with 3.1GBs of data even though supermount should've umounted them. So I guess something is up with fstab and mounting my discs. Is there some fix in the updates for anything like this? Just updated and installed everything but haven't rebooted if udev or something I have no clue about needs to be run from boot. Also just left my external plugged in as I've left it connected and on through my last few mandrake installs and it always was fine, also as I can assign a mount point during install and not have to do it later, at least that always worked in the past.

Using mandrake 10.1 official, trying to mount my 120Gb seagate external HD, which is this entry with lspcidrake "usb-storage : Seagate|Seagate Mass Storage [Mass Storage|SCSI|Bulk (Zip)]". This drive was detected and a mount point was set during installation/partitioning, I did a custom partitioning to make sure it wasn't erased as it's nearly full of files. After I installed the OS and booted I looked under /mnt/external, and checked mount, and it's not mounted. I've tried to use "mount -t vfat /dev/sda1 (and /dev/sda) /mnt/external" and I get the error "mount: permission denied". This same drive was detected, and mounted automatically under 9.2 official, I reinstalled 9.2 a few hours ago and checked again and it detects and mounts it perfectly, then I tried again with 10.1 and I get the "permission denied" error. Not sure what the problem is other than the 2.6 kernel and the udev system maybe, any help would be great. Thanks.

Have you used the same dial-up account on another OS/computer before with no problems? Do you have another network connection that works on your Mandrake installation? Just some basic questions, but I think your problem is with your DNS resolution. I've had the same thing happen a few times when setting up a dial-up connection on MDK installs, sometimes the DNS just doesn't seem to work properly. Could you post, and or check your /etc/resolv.conf file? It should look something like this, (without my made up IP addys)- [default@localhost default]$ cat /etc/resolv.conf search advant local # ppp temp entry nameserver 234.167.114.10 # ppp temp entry nameserver 159.3.222.10 # ppp temp entry Also your /etc/ppp/resolv.conf file should have your DNS servers listed. It should look the same as your /etc/resolv/conf file except without the "search..." and "# ppp temp..." parts.

That's weird I've tried the same scan before on a fresh install on "msec level 3" and my ports were reported as "closed". I guess it probably has to do with the fact that I installed iptables, (just so I could set up the firewall later), but didn't specify any rules yet so everything was on "ACCEPT" gmac - BTW alot of the scans that you are getting aren't malicious. I drop lots of packets from sites I'm browsing or using various programs. I drop stuff from this site and places like justlinux.com or slashdot all the time. My firewall rulesets are tight and various extraneous packets just get dropped for one reason or another. The same thing happens when I'm using windows. Also if you have a dynamic IP addy you sometimes will get traffic your way that was meant for the person who previously had that addy. I've seen things like packets meant for Kazaa, Yahoo Messenger, and Gnutella flooding in and dropped when I dial in and get someones old IP. Of course I imagine that there's just as many scans that are malicious smacking up against your firewall. If you're still using Windows I'd recommend checking out Oupost Firewall instead of ZoneAlarm as it's very customizable and you can set up each and every application according to what ports, hosts, and directions are allowed for all internet traffic.

Personally I like http://www.pcflank.com/ as a web based port scanner. The advanced port scanner can be used for normal or SYN scans, and you can specify each and every port that it scans. There's also exploit tests and a few other scans that are pretty good for checking out some security settings. Shields up! is more windows based and only scans a few ports like 137, 21, 80 and some of the more basic services like telnet. With PC Flank you can check out ports that UNIX services run on like 6000 or anything else. First thing I do when checking out a new firewall or iptables ruleset is pull an nmap and then head to pcflank.com to see how things are going. It's a pretty good preliminary step.

Yikes... sorry if I offended you. I didn't mean to imply that if you had mp3s that you got them illicitly. I was just trying to help and whatnot and give you my own audiophile opinion on lossy versus lossless compression and wavs. Sorry also for thinking that you were going to encode to mp3, I didn't read your post too carefully and I had heard grip could encode to mp3 after ripping and I didn't know it could encode to ogg. Although I guess I should've realized that "hey this is linux, ogg is everywhere". I think if you already have used cdparanoia on your discs like I said, (?), you could make a mix CD in any of your burning programs without bothering with any type of encoding as your wav files will make a higher quality CD than would the ogg files. So if you are backing up all of your discs to ogg I'd use the wav files to burn the discs before I encode them, just IMHO. But if you've already got oggs of all your other discs that you're going to be burning the mix CD with then I wouldn't see the point in using the wavs as you've got most the stuff in one format and you might as well use the ssame format across the board. Personally I like FLAC, but that's just me.

If you want to just make a mix CD and not an mp3 disc then I'd recommend just using "cdparanoia -B" from the command line on the discs you want to rip. You'll get .wav files instead of mp3's and save yourself some time that you would've spent encoding to mp3. Then you can burn the .wav file to a CD right away, and you won't lose any sound quality in the process. Cdparanoia by default will make a near-perfect copy of a disc. Whereas most ripppers can toss out chunks of audio data like it's trash. If you're going to hook the kid up with tunes you might as well make them the best quality possible. Personally I don't like the sound of mp3s that are encoded below 320 kbps. IMO lossless encoding is far superior to the mp3 format. Of course if you made the CDs from mp3s in the first place then that isn't really a factor.

I've had the same thing happen on two boxes after I had futzed around quite a bit with the MBR. After I zeroed out the HD with Autoclave things were back to normal, but I got the error on my main machine and as I've got a few partitions and OSs already installed on it I just set up LILO on a floppy and leave it in all the time, as I'd rather not have to zero out everything and reinstall 3 operating systems.

When I urpmi'd man and the other packages I didn't get any error messages at all, everything installed fine... but the actual files would not show up. I tried urpmi and rpmdrake on all of the files that wouldn't install. But I figured out, (I think), what I did to make this happen. I installed in text mode, deselected all of the package groups, and hit "individual package selection". Then you get three options... "With X", "With basic documentation, (recommended)", and "Truly basic install, (especially no urpmi!)". I didn't select any of the options as I wanted only the packages I selected individually, which included X, urpmi, and the man pages. If you select any of the options I've noticed that you get some more stuff than just what they say you'll get. So anyway urpmi and everything else worked just fine. But I think since I didn't select "with basic doc*", something screwy happened and I couldn't get the man pages and /usr/share/docs for anything! I guess when they give you that option that you had better take it if you want documentation... of course I didn't think that would be the case when I installed. Seems to be a bit of overkill really if that is the case. I guess it makes it impossible to install the man pages and whatnot, sort of strange. I think I might try re-creating the bug/feature on another box just to see if I'm right. Anyway thanks for the help and a big "Doh!" on my part.

Already tried using rpmdrake to get rid of man and manpages and then I installed them again. It didn't work. I did try out your suggestion of running a "rpm --rebuilddb" and trying it again, but they are not there... it installs and makes the directory, but there is nothing there. I also tried installing the packages "rute" and the "howtos" as I know they install into /usr/share/doc. They install fine but there are no files at all! I've checked, doublechecked, uninstalled, reinstalled and it will not make files in usr/share/man or in usr/share/doc. I'm about ready to try yet another install and see if the same type of thing happens again, but I really want to know WTF is going on.

So I've installed 9.1 on a box with a winmodem after I found a driver that would work. A conexant something with the HSF chipset. The rpm I installed was hsflinmodem-5.03.27lnxtbeta03042700k2.4.21_0.13mdk-1mdk.i586.rpm. Everything seemed to be going fine, the driver worked like a charm and all was good. Then after I setup urpmi and got a few updates I started getting some strange errors, "libcurl.so.* is corrupt", "wget is corrupt", the same sort of messages for almost every command I tried, cp, mv, everything I tried except reboot. I can't list the exact error message as I didn't save the output. So I decided to reinstall as the installation was only an hour or two old and I just wanted to get things back to normal. So I installed 9.1 and the driver again, and everything worked for a day or so. Then after I logged out of IceWM I got more strange errors when entering commands, "command perl not found", "wget not found", etc. So I reboot the box and log back in, now everything can be found fine, and I've had no more problems along those lines. But I decided to take a look at some man page, and I get the error "man wget not found". So I look at /usr/share/man and all the folders are there, but the pages are all gone! Then I look at /usr/share/doc and once again all the directories are there, but absolutely no files are there at all. Other than those two things everything else in /usr seems to be fine. Very strange. I've got 9.1 on several other boxes and have never had any of these problems before. I have no clue what the problem could be other than the winmodem driver, although I have no idea why that would make my man pages dissapear. I mean I know it's supposed to "corrupt the kernel", ;), but this seems ridiculous.

Doh! That's what I get for only looking at rpmfind.net and the main ftp site. I figured something didn't quite seem right about not having the 9.1 rpms around.

I like guarddog too, however this new version has presented a few problems. Like you said it normally starts the firewall whenever a network interface is brought up, which for me is a dial-up connection. Now it doesn't start at all, which I imagine is just some problem with chkconfig or something in init.d that just for some reason didn't work this time around. Or maybe something with the script as it normally, (in old versions), starts when I make a ppp0 connection according to syslog. I've just started manually running the script that guarddog already made for me whenever I bootup, as I'm too lazy to get around to adding it to init.d or chkconfig-ing it or whatever I'll have to do. I think I'm going to look into shorewall or just using a script for iptables now that I have a general idea of futzing with text config files for my firewall, and I do not like the idea of getting the entire kde desktop just for one program. edited for clarity

Just a note, the 9.1 rpms cannot be found on rpmfind. It seems that they only have rpms up to 9.0 as of now. Lets hope they fix that. You should be able to find all the rpms on the mirror I linked to, or any of the "core file"plf mirrors that you can get to on the easy urpmi site - http://plf.zarb.org/~nanardon/urpmiweb.php Anyway if you snag any of those rpms you can just add them as a local source in urpmi sources manager off off "config > packages" . Specify the path and hit update once or twice to make sure it takes, (it always bugs out on me once or twice when I do so) and the dependencies, if there are any, will be taken care of for you when you install via rpmdrake or urpmi. Of course if you already have most of kde on your computer I imagine a normal "rpm -ivh" will work just fine for kppp. Wvdial will need a few more packages like I said earlier, I'm not sure what discs those would be on so you might look at the mirrors.

Just make sure if you call the new executable file "links" that when you are moving your old "links" command you keep it out of your path as well as keeping the path specified in the new executable valid. That is do not use /usr/bin/links as the path as that will still be in your $PATH. I'm sure you get the idea.

Don't forgt w3m, it's pretty smooth as well. Also check out surfraw from the 3rd install cd. Google and slashdot straight from the console with excellent search options you can slap on the CL. Having the most excellent google command is almost as cool as the search bar in mozilla. Also if you want to specify a homepage you can do so by making a simple script, (well not a script as there's just one command, but whatever you call it), that just contains the following - ------------------------------------------------ # Links with a homepage /usr/bin/links yourhomepage.com ------------------------------------------------ Call it something like home, or you could call it links and move the old links executable out of your path. Chmod 0755 the file you just made and move it to your /bin, or /usr/local/bin and you have links with a homepage. This works for any other executables and whatnot that you want to futz with the options and have it be easy to call from the CL.

I recently upgraded to using guarddog 2.1.4 on a new install of MDK 9.1 after having used 2.0.0 on all of my old installs. Besides the fact that the dependencies now make me install the entire KDE desktop instead of just a few kdelibs, now it seems that the firewall script, (rc.firewall), does not load when I boot up my system. I have to manually go into guarddog and hit the "OK" button to run the script. I know there has to be an easy way to fix this and whatnot, I've just always used guarddog since I've started with linux and do not know much about iptables besides hitting "iptables -L" to see what guarddog did each time I mess with it. Some help with how to fix running guarddog would be great, as well as any suggestions on something better to use as I'm not too fond of the new dependencies.

The easiest way for you to get a dialer would be to simply get a few rpms off of one of the mirror sites that host the core files. You can get the kppp dialer here - ftp://public.ftp.planetmirror.com/pub/man...-31mdk.i586.rpm Or from rpmfind or any other mirror. You can also get wvdial the same way, although you will need a few dependencies, like wvstream-something and one or two others. An rpm -ivh on the file will tell you what you need.

Thanks! That looks like it'll solve my problems.

After taking a look at the man page for msec I notice that my current config is in /var/lib/msec/security.conf CHECK_SHADOW=yes CHECK_SUID_MD5=yes CHECK_UNOWNED=no CHECK_SECURITY=yes CHECK_PASSWD=yes SYSLOG_WARN=yes CHECK_SUID_ROOT=yes CHECK_PERMS=yes MAIL_EMPTY_CONTENT=no CHECK_WRITABLE=yes CHKROOTKIT_CHECK=yes CHECK_PROMISC=no CHECK_SGID=yes RPM_CHECK=yes TTY_WARN=no MAIL_WARN=yes CHECK_OPEN_PORT=yes So can I just edit this file and set my options that way after an "msec -something"? Or do I have to edit a file in /etc/security/msec? And does a simple invocation of msec (1, 2, 3) setup my security level according to the *.conf file? Or what? I've tried reading through the msec man page a few times and doing a search on mandrake secure or what have you with no luck.

Maybe that's part of the problem. I don't have a perm.local. Here's an ls of /etc/security/msec/ level.local security.conf* server.4 server.5 level.local and security.conf are blank, the server.4 and 5 files are just lists of commands and whatnot. I'll try some of the security checks before long and see if those settings "stick".

Thanks for the response. Actually I already knew about msec, I was looking for a bit more refined level of security besides the defaults provided by level 1, 2, 3, etc. As even level 4 allows remote root login with a password. But anyway I bumped up security to "high" in DrakSec, and set my options, (console log on v12 being among my most valued), and it seems to work. Although if "high" is msec level 3 then it might just be on that and my specific options might not have taken since console log, (the only noticeable option I checked), is enabled by default on level 3. So I went rooting around for a config file and /etc/security/msec/security.conf and all the other files in there don't tell me anything about my settings at all. In fact they're empty. So I guess I'll try enabling some specific types of security checks that aren't part of msec level 3 and see if my mucking around with DrakSec has accomplished anything.

I recently installed 9.1 after using 9.0 for a long time and I've had a few problems configuring my system's security settings the same way. I go into the control center, then DrakSec to set up specific settings, like "allow_remote_root_login" under "Network Options, System Options, Periodic Checks". I set all of my options and everything seems to go normally except I've noticed that "enable_console_log" won't work no matter how many times I've tried to set it through DrakSec. Also all of my options are reset/dissapear everytime I start up DrakSec making me wonder if they are working, or not working like the console log on vt12. I've updated all of the packages via the official update mirrors in urpmi and I still have the same problem. Thanks for your help.