Re: Installing testing: can't get cable Internet after reboot
Edward C. Jones wrote: How can I solve my problem? Do I need to install dnsmasq on T? If so, can I download it to K then copy it somewhere in T? Take a incremental backup of T, noting which files which have changed since yesterday's daily backup, then restore T from yesterday's backup. Now your system should be functional again. Then go back to the list of changed files and play detective, working backwards in time until you find which change caused the failure. If you don't understand the causal mechanism of the failure, post again to this list. Finally, update all the changed files which are unrelated to the failure, and you're good to go. :-) -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
samba won't dance
First thanks for the advice on using Apt given earlier. I managed to get Samba and dependencies installed this way after connecting to my LAN gateway. Now having trouble with Samba - it lists the computer on my XP workstation but when I click on it get a msg that the network resource is inaccessible. UNCLELEO:~# cat /etc/samba/smb.conf [global] workgroup = FACE2INTERFACE encrypt passwords = yes smb passwd file = /etc/smbpasswd log file = /var/log/samba.log lock directory = /var/lock/samba share modes = yes [look] path = / read only = yes public = yes [tmp] path = /tmp read only = no valid users = dad marty ;[web] ; path = /mnt/web ; read only = no ; valid users = dad marty UNCLELEO:~# cat /etc/smbpasswd dad:1001:42110F925EAFD4CDAAD3B435B51404EE:D5EC2F34658F6BAA862ADE8B799FC5B7:[U ]:LCT-43A761D8: marty:1000:8A0043FB9C6CC5FBAAD3B435B51404EE:353B645F8DBC9332E460916669527083:[U ]:LCT-43A761E5: UNCLELEO:~# tail /var/log/samba.log open_oplock_ipc: Failed to get local UDP socket for address 17f. Error was Cannot assign requested address [2005/12/20 08:00:35, 0] lib/util_sock.c:open_socket_in(708) [2005/12/20 08:00:35, 0] smbd/oplock.c:init_oplocks(1357) open_oplock_ipc: Failed to get local UDP socket for address 17f. Error was Cannot assign requested address [2005/12/20 08:01:16, 0] lib/util_sock.c:open_socket_in(708) [2005/12/20 08:01:16, 0] smbd/oplock.c:init_oplocks(1357) open_oplock_ipc: Failed to get local UDP socket for address 17f. Error was Cannot assign requested address [2005/12/20 08:01:16, 0] lib/util_sock.c:open_socket_in(708) [2005/12/20 08:01:16, 0] smbd/oplock.c:init_oplocks(1357) open_oplock_ipc: Failed to get local UDP socket for address 17f. Error was Cannot assign requested address UNCLELEO:~# Based on the log, tailed above, have I configured ok but started the daemon incorrectly? I log onto my XP box as dad and the conf file is almost identical to working ones on other 'nix boxes on my LAN. tia, Marty Marty Landman, Face 2 Interface Inc. 845-679-9387 Webmaster's Bulletin Board: http://bbs.face2interface.com/ Web Installed Formmail: http://face2interface.com/formINSTal -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Need help with backing up a Windows 98/2000 computer using rsync on a Debian Server running ssh
TAC Forums wrote: Hi I have a Debian server running ssh which performs the job of a Backup Server. All the GNU/Linux workstations / servers on the network have a bash script that run in the cron to rsync the data folders to the main backup server at night. However, I am having difficulting finding an rsync binary for Windows that can talk to a Debian ssh server to sync the files. I don't know if it will help, but a modified rsyncd for Cygwin is available at the following link: http://sourceforge.net/project/showfiles.php?group_id=34854package_id=88133 The rsync tools available on the net appear to talk with a Whitebox/Redhat/Fedora ssh server but I'm having trouble with getting it to work on a Debian ssh server. Any ideas on what I should do to get this to work? Maybe I am using the wrong rsync tool? Is there another rsync tool written for talking to the ssh server of Debian? Your description seems to point to the Debian ssh server. Have you compared the servers closely and examined any differences? Regards -- TAC Support Team -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
samba won't dance
Got it - turns out I had my smb.conf on the wrong path. Marty Landman, Face 2 Interface Inc. 845-679-9387 Webmaster's Bulletin Board: http://bbs.face2interface.com/ Web Installed Formmail: http://face2interface.com/formINSTal -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Install kernel-source-2.6.8 update
Marcel Stoop wrote: On Fri, 2005-12-16 at 23:37 -0500, Marty wrote: So in other words, if you like to inform other (beginning) users, or just answer their questions about kernels on the list. Please make sure they use it the way they should, in debian that is. The method I proposed works just fine in debian. I suppose somebody here wants to start a flamewar about it. Count me out before it even starts. :-) -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: install samba from binary
At 02:40 PM 12/12/2005, Juergen Fiedler wrote: On Mon, Dec 12, 2005 at 02:33:43PM -0500, Marty Landman wrote: At 01:48 PM 12/12/2005, Andrei Popescu wrote: [...] UNCLELEO:~# ping -c 1 google.com ping: unknown host google.com What do you get if you do (for example) 'ping -c 1 64.233.187.99'? No internet connectivity although my LAN connectivity is working. Is there a simple way for me to fix that? Marty Marty Landman, Face 2 Interface Inc. 845-679-9387 Webmaster's Bulletin Board: http://bbs.face2interface.com/ Web Installed Formmail: http://face2interface.com/formINSTal -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Install kernel-source-2.6.8 update
Marco wrote: Hi all, I have installed Debian Sarge and I have an custom kernel. Yesterday, I downloaded from security.debian.org the kernel-source-2.6.8 (DSA 922) update with a apt-get update and after apt-get upgrade. This is the output of command: dpkg -l | grep kernel-source ii kernel-source- 2.6.8-16sarge1 Linux kernel source for version 2.6.8 with D What do I have to do for install this update? Should I recompile the kernel? Thanks Marco ** New kernel in 5 easy steps! 1) untar /usr/src/kernel-source-2.6.8.tar.bz2 2) Add your .config (e.g. from /proc/config.gz) 3) make oldconfig;make bzImage 4) replace old bzImage in /boot 5) run lilo -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Install kernel-source-2.6.8 update
Roberto Sanchez wrote: Marty wrote: New kernel in 5 easy steps! 1) untar /usr/src/kernel-source-2.6.8.tar.bz2 Check. 2) Add your .config (e.g. from /proc/config.gz) Check. 3) make oldconfig;make bzImage Why not use make-kpkg from kernel-package? If you have just one kernel, or don't use modules, I don't see much advantage, and I tend to err on the side of simplicity. 4) replace old bzImage in /boot See last. cp /usr/src/linux/arch/i386/boot/bzImage /boot Yep, it's a lot of typing. I see what you mean. 5) run lilo Switch to grub :-) You mean the bootloader that doubles as backup kernel? That seems like a cool idea, if I can just figure out what it's good for. :-) -Roberto -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: What would I do without partimage?
William Ballard wrote: I literally would be unable to use Microsoft Windows if I couldn't stay mostly booted in Debian and manage that godawfulness with partimage. Every time I boot into it I restore a clean partimage of XP, let it puke all over itself, then restore the cleanness. It's the only thing that makes patching Windows remotely tolerable. Eventually partimage will stop working on new versions of NTFS, and it seems to not be maintained anymore. It was removed from Sarge. Are there other tools that work like Ghost but in Linux? Partimage is great. Maybe this is just a case where in Unix you don't need weird tools to do a routine task, or what *should* be routine in any sanely designed OS.* I've always just used the cp command, e.g. cp /dev/hdadisk_image or cp /dev/hda1partition_image to back up a disk or partition, respectively. (dd also works but I just don't trust it.) I've never tried to gzip or bzip the resulting image file, but if that works then I don't see much advantage using partimage or Ghost. They may be of marginal value if they are smart enough to automatically detect and adjust to differing drive geometries. I don't think cp or dd can handle that by itself. *What prompts this remark is an industrial application I heard about, where a large number of identically configured systems used ghost (or similar tool) to reinstall a pristine copy of XP each time the systems were booted, i.e. at least once per day! (True story.) -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: kernel has null dereference during boot
[EMAIL PROTECTED] wrote:
While booting my newly installed woody system (Why not sarge?
It's a long story which will be told another time) the kernel
crashes, recovers, and fails to make one of my hard disk
accessible.
Here's the relevant part of the dmesg output:
...
...
Uniform Multi-Platform E-IDE driver Revision: 6.31
ide: Assuming 33MHz system bus speed for PIO modes; override with idebus=xx
PIIX: IDE controller on PCI bus 00 dev 38
PIIX: chipset revision 2
PIIX: not 100% native mode: will probe irqs later
PIIX: neither IDE port enabled (BIOS)
PIIX: IDE controller on PCI bus 00 dev 39
PIIX: chipset revision 2
PIIX: not 100% native mode: will probe irqs later
ide0: BM-DMA at 0x3000-0x3007, BIOS settings: hda:pio, hdb:pio
ide1: BM-DMA at 0x3008-0x300f, BIOS settings: hdc:pio, hdd:pio
hda: FUJITSU MPB3064ATU, ATA DISK drive
hdb: Mxo 89H , ATA DISK drive
hdd: HL-DT-ST CD-ROM GCR-8520B, ATAPI CD/DVD-ROM drive
ide0 at 0x1f0-0x1f7,0x3f6 on irq 14
ide1 at 0x170-0x177,0x376 on irq 15
hda: 12672450 sectors (6488 MB), CHS=13410/15/63
hdb: 264241407 sectors (135292 MB) w/2111KiB Cache, CHS=16383/127/127
hdb: INVALID GEOMETRY: 127 PHYSICAL HEADS?
Partition check:
/dev/ide/host0/bus0/target0/lun0: [PTBL] [788/255/63] p1 p2 p3 p4 p5 p6 p7 p8
p9 p10
Unable to handle kernel NULL pointer dereference at virtual address 0028
printing eip:
c382f3c1
*pde =
Oops:
CPU:0
EIP:0010:[c382f3c1]Not tainted
EFLAGS: 00010212
eax: ebx: 03f0 ecx: 0300 edx: 0045
esi: c384b144 edi: 0040 ebp: 0040 esp: c2c49ee8
ds: 0018 es: 0018 ss: 0018
Process modprobe (pid: 28, stackpage=c2c49000)
Stack: c384b144 0001 c384afe0 0001 c384b210 03002980 c382f458 0340
0002 c3852980 19e0 c3831fc5 c38524d9 c38529c0 c3851000
0001 0001 c01156fd c2c48000 400199d8 bfffc4dc bfffc49c 1800
Call Trace: [c384b144] [c384afe0] [c384b210] [c382f458] [c3852980]
[c3831fc5] [c38524d9] [c38529c0] [c01156fd] [c3851060] [c0106d93]
Code: 83 78 28 00 74 09 56 8b 40 28 ff d0 83 c4 04 80 a6 9e 00 00
3ext3: No journal on filesystem on ide0(3,2)
Adding Swap: 128516k swap-space (priority -1)
Real Time Clock Driver v1.10e
...
...
By the way, I'm also wondering about the weird characters in the line
that identifies my second hard disk, an 80G Maxtor, but I don't think
they do any harm in themselves.
hdb: Mxo 89H , ATA DISK drive
Things you can try:
-if there is a BIOS CHS setting for that drive, or it's set to AUTO turn it off (no
drive present). Linux will still find the drive. Conversely if the BIOS is already
set to no drive present, set it to the correct CHS setting for that drive.
-Try a different kernel
-Test the Maxtor on a different motherboard.
-If you are ambitious, recompile the kernel and printk the drive ID string to
see
how or why it's being corrupted, then submit a bug report or patch to the IDE guy
(especially if the bug is still in the latest kernel).
-
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: kernel has null dereference during boot
[EMAIL PROTECTED] wrote: OOPS. Forgot to identify the kernel. This problem occurs with kernels 2.4.18-1-386 2.4.18-1-586tsc With kernel 2.2.20 I get the complaint about the invalid number of physical heads, but not the crash. It used to work properly before I did the forced reinstall (The machine was rooted) When it worked, I believe I was using kernel 2.4.16-586. I would be suspicious that something tried to change the bootloader (e.g. LILO or GRUB) or the partition table (MBR) on the drive, in a possible (though unlikely) effort to defeat the cleanup attempt. First I would save copies of the partition table and bootloader for future forensic purposes. Then I would try the fdisk p (print partition table) command to see if it indicates anything wrong with the partition table. I would also run LILO on the drive with the -v -v -v -t options to see if it finds anything wrong (or run the equivalent grub diagnostic). If there are any remaining doubts I would follow up with an appropriate disk editor to examine the drive more closely. After locating and identifying any problem(s), I would rewrite the MBR and/or bootloader. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: kernel has null dereference during boot
[EMAIL PROTECTED] wrote: On Tue, Dec 13, 2005 at 02:39:11PM -0500, Marty wrote: First I would save copies of the partition table and bootloader for future forensic purposes. Too late! I'll try and remember that next time. By the way, how *do* you save copies of the partition table and bootloader? dd is commonly used, but then you have to know the sector offset to the bootloader. But being lazy I would probably just grab the first few megs and some backup superblocks by running this for a few seconds: cp /dev/hdbhddump where hddump is a partial image of the drive. If it's a small drive I'd probably just dump the whole disk. snip I installed using SBM to boot off of the woody CD. The installer formatted a new swap partition, and a new boot/root partition, and installed into that one. The installer's lilo was used to rewrite the MBR. That should pretty well finish off the MBR, no And the bootloader too. After thinking more about your problem I suspect that your new kernel has the wrong processor selected, maybe not the one you think it has. This will cause a kernel oops early in a boot cycle. As an experiment, you could try the lowest common denomenator processor option (386 or whatever). fdisk's p comment (the one in the installer did not notice anything wrong. Thanks for telling me about lilo -v -v -v. I'll try that next. Don't forget the -t unless you are really ready to write. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
install samba from binary
Hi, I've downloaded Samba as a .deb file. Is there an easy way for me to install from that? Marty Landman, Face 2 Interface Inc. 845-679-9387 Webmaster's Bulletin Board: http://bbs.face2interface.com/ Web Installed Formmail: http://face2interface.com/formINSTal -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: install samba from binary
At 11:23 AM 12/12/2005, Hodgins Family wrote: Hi, I've downloaded Samba as a .deb file. Is there an easy way for me to install from that? dpkg -i package should work. Check the dpkg man pages for exact syntax. I'm probably doing this all wrong, have put the distro in /root/debs UNCLELEO:~# ls -Al debs total 2184 -rw-r--r--1 root root 20 Dec 10 18:25 Packages.gz -rw-r--r--1 root root 2225400 Dec 10 17:59 samba_3.0.20b-1woody1_i386.deb UNCLELEO:~# dpkg -i samba_3.0.20b dpkg: error processing samba_3.0.20b (--install): cannot access archive: No such file or directory Errors were encountered while processing: samba_3.0.20b UNCLELEO:~# Can you help straighten me out here? P.S. How come you didn't bring it in through apt-get or synaptic or aptitude (just curious!) I've got a hosed up woody install on a marginal pc - 166Mhz/48MB ram and broken cdrom although it worked for the mini-iso install. OTOH if I can get samba working that'll be okay enough to keep this on my network. Thanks in advance. Marty Marty Landman, Face 2 Interface Inc. 845-679-9387 Webmaster's Bulletin Board: http://bbs.face2interface.com/ Web Installed Formmail: http://face2interface.com/formINSTal -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: install samba from binary
At 11:48 AM 12/12/2005, Martin Lefebvre wrote: dpkg -i samba_3.0.20b-1woody1_i386.deb UNCLELEO:~# dpkg -i samba_3.0.20b-1woody1_i386.deb dpkg: error processing samba_3.0.20b-1woody1_i386.deb (--install): cannot access archive: No such file or directory Errors were encountered while processing: samba_3.0.20b-1woody1_i386.deb UNCLELEO:~# . I guess the problem is that I put the deb on my /root/debs directory. Where should it go? Marty Marty Landman, Face 2 Interface Inc. 845-679-9387 Webmaster's Bulletin Board: http://bbs.face2interface.com/ Web Installed Formmail: http://face2interface.com/formINSTal -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: install samba from binary
At 01:09 PM 12/12/2005, Hodgins Family wrote: I was wondering why you plunked the file into /root/deb. Why not just move it to /home and retry the dpkg command (as root obviously!) UNCLELEO:~# dpkg -i /root/debs/samba_3.0.20b-1woody1_i386.deb Selecting previously deselected package samba. (Reading database ... 12262 files and directories currently installed.) Unpacking samba (from .../samba_3.0.20b-1woody1_i386.deb) ... dpkg: dependency problems prevent configuration of samba: samba depends on samba-common (= 3.0.20b-1woody1); however: Package samba-common is not installed. samba depends on libacl1; however: Package libacl1 is not installed. samba depends on libattr1; however: Package libattr1 is not installed. samba depends on libcupsys2 (= 1.1.13-1); however: Package libcupsys2 is not installed. samba depends on libkrb53; however: Package libkrb53 is not installed. dpkg: error processing samba (--install): dependency problems - leaving unconfigured Errors were encountered while processing: samba UNCLELEO:~# Now what? Marty Marty Landman, Face 2 Interface Inc. 845-679-9387 Webmaster's Bulletin Board: http://bbs.face2interface.com/ Web Installed Formmail: http://face2interface.com/formINSTal -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: install samba from binary
At 01:27 PM 12/12/2005, Johannes Wiedersich wrote: If your CD works for installation, but fails afterwards, there might be something wrong with your config. Maybe you should try to get some sarge CDs and install sarge. I've tried installing RH 9 since I already have the CD's for that as well as the floppies to boot and start installation from. But my cd wouldn't work when booting from the RH floppies either so assumed it was the drive itself. Marty Marty Landman, Face 2 Interface Inc. 845-679-9387 Webmaster's Bulletin Board: http://bbs.face2interface.com/ Web Installed Formmail: http://face2interface.com/formINSTal -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: install samba from binary
At 01:48 PM 12/12/2005, Andrei Popescu wrote: As it just said, you need all those packages listed there in order to install samba. Don't remeber from your original post...do you have internet? No. Not sure why though. :( My nic works and I can ping out, ssh in from my workstation and out to other network boxes. UNCLELEO:~# ifconfig -a eth0 Link encap:Ethernet HWaddr 00:A0:CC:32:CC:42 inet addr:192.168.0.222 Bcast:192.168.0.255 Mask:255.255.255.0 UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 RX packets:33311 errors:0 dropped:0 overruns:0 frame:0 TX packets:3784 errors:0 dropped:0 overruns:0 carrier:0 collisions:71 txqueuelen:100 RX bytes:5382739 (5.1 MiB) TX bytes:357660 (349.2 KiB) Interrupt:12 Base address:0x1200 loLink encap:Local Loopback inet addr:127.0.0.1 Mask:255.0.0.0 UP LOOPBACK RUNNING MTU:3924 Metric:1 RX packets:8 errors:0 dropped:0 overruns:0 frame:0 TX packets:8 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:0 RX bytes:560 (560.0 b) TX bytes:560 (560.0 b) UNCLELEO:~# ping -c 1 google.com ping: unknown host google.com UNCLELEO:~# route add default 192.168.0.1 SIOCADDRT: No such device UNCLELEO:~# Just do: # aptitude install samba and it will get and install all packages automagicaly. UNCLELEO:~# aptitude install samba bash: aptitude: command not found UNCLELEO:~# Marty Marty Landman, Face 2 Interface Inc. 845-679-9387 Webmaster's Bulletin Board: http://bbs.face2interface.com/ Web Installed Formmail: http://face2interface.com/formINSTal -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: install samba from binary
At 02:40 PM 12/12/2005, Juergen Fiedler wrote: What do you get if you do (for example) 'ping -c 1 64.233.187.99'? Network unreachable. I'm going to work on that first, then work on Johannes suggestion try to use Aptitude. Marty Marty Landman, Face 2 Interface Inc. 845-679-9387 Webmaster's Bulletin Board: http://bbs.face2interface.com/ Web Installed Formmail: http://face2interface.com/formINSTal -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: gcc internal error
At 03:08 PM 12/10/2005, David E. Fox wrote: Is the error repeatable - same source file, same error? Yes, I tried a few times. does it happen with other compiles? Don't know yet, wanted to do Samba first. I'd first try memtest86 - run it through and see if there are any memory errors found. Heh, did I mention this was installed using the mini-iso? UNCLELEO:~# memtest86 bash: memtest86: command not found UNCLELEO:~# find / -name memtest86 UNCLELEO:~# Did above as root. Or could this indicate a hosed install? I have been noticing problems with the cdrom for that box. I'm on a LAN so could reinstall maybe. Marty Marty Landman, Face 2 Interface Inc. 845-679-9387 Webmaster's Bulletin Board: http://bbs.face2interface.com/ Web Installed Formmail: http://face2interface.com/formINSTal -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
gcc internal error
I've recently installed Woody on an old PC - 166Mhz w/ 2.5 GB hd. Am trying to get Samba installed from the source; after doing ./configure, apparently ending ok my make stops with an internal gcc error. Compiling rpc_parse/parse_net.c with -fPIC Compiling rpc_parse/parse_reg.c with -fPIC Compiling rpc_parse/parse_rpc.c with -fPIC gcc: Internal compiler error: program cc1 got fatal signal 11 make: *** [rpc_parse/parse_rpc.po] Error 1 UNCLELEO:/tmp/samba-3.0.10/source# Any idea how deep a mess I'm in here? Haven't really gotten anything working on this box yet, so if I ought to start from scratch that might be the easiest way to go. Marty Landman, Face 2 Interface Inc. 845-679-9387 Webmaster's Bulletin Board: http://bbs.face2interface.com/ Web Installed Formmail: http://face2interface.com/formINSTal -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: remove a page about us on your site please
At 11:58 AM 12/8/2005, Steve Block wrote: On Thu, Dec 08, 2005 at 05:53:02PM +0100, Johannes Wiedersich wrote: Dr. Nicolas Bussard wrote: Hello, Could you please remove this page from your site: http://lists.debian.org/debian-user/2004/09/msg03184.html Thanks Regards stepnewz Just stop sending out spam and justice will stop prosecuting you! Thanks for not going to send any spam to anyone anymore! Johannes Not this again. Anything posted here is public, and should stay public, regardless. If you keep that in mind when you're sending mail you'll be ok. As well ask all transcripts of a public speech be destroyed. Not to mention http://www.archive.org/. :) Marty Marty Landman, Face 2 Interface Inc. 845-679-9387 Webmaster's Bulletin Board: http://bbs.face2interface.com/ Web Installed Formmail: http://face2interface.com/formINSTal -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: jerks
Glenn English wrote: I'm trying to make a Digital Audio Workstation (DAW) out of a SUN W2100z (dual AMD64). Every few seconds, at seemingly random times, everything freezes for ~50ms. Even the mouse. Sometimes. Reliable at the first card moved in Aisle Riot Solitaire and when Jack is running. Often when I'm typing email (chars don't get from the keyboard to the screen). I know that's not much of a description, but I don't know where to start looking for this. I'm assuming interrupts are being disabled entirely momentarily. My background says it's time for a bus monitor, but I don't have one handy. This happens with a 2.6 kernel from the current updated Sarge or a 2.4 from DeMuDi 1.2.1 distro. Anybody have any idea how to begin tracing this down? Or maybe a better explanation of what's going on? Is DMA enabled on your hard drive(s)? Is your kernel is compiled with the Preemptible Kernel (PREEMPT) option enabled? -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: installing a nic
At 03:10 PM 12/5/2005, Ryan Nowakowski wrote: On Sun, Dec 04, 2005 at 09:07:44PM -0500, Marty Landman wrote: I've got Woody installed from the mini-iso and owing to problems with the old nic which never got recognized have swapped in a Linksys LNE100TX which will take the tulip driver iirc. How do I go about installing though? An ifconfig -a shows eth0 after I ran a modprobe tulip but it's not up and running. Pinging only works for localhost and netstat shows nothing in particular. This is for Sarge but most of it should still be valid for Woody: http://www.debian.org/doc/manuals/reference/ch-gateway.en.html Thanks Ryan this is very helpful and I'm getting somewhere. Marty Marty Landman, Face 2 Interface Inc. 845-679-9387 Webmaster's Bulletin Board: http://bbs.face2interface.com/ Web Installed Formmail: http://face2interface.com/formINSTal -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
installing a nic
I've got Woody installed from the mini-iso and owing to problems with the old nic which never got recognized have swapped in a Linksys LNE100TX which will take the tulip driver iirc. How do I go about installing though? An ifconfig -a shows eth0 after I ran a modprobe tulip but it's not up and running. Pinging only works for localhost and netstat shows nothing in particular. Marty Marty Landman, Face 2 Interface Inc. 845-679-9387 Webmaster's Bulletin Board: http://bbs.face2interface.com/ Web Installed Formmail: http://face2interface.com/formINSTal -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Filesharing on small LAN
[EMAIL PROTECTED] wrote: Linux NFS is pretty unreliable in my experience, randomly failing in spite of network connectivity every couple months or so -- but I haven't used it seriously in a while (since 2.4.10 or so maybe), and it may have improved. (I've waited for some improvement since 2.0.30-something and gave up even checking in the 2.4 series, but who knows -- maybe a kernel hacker somewhere started caring about NFS). I use NFS almost on a daily basis. The only problem I've had is the occasional inability to umount an NFS volume, which required a reboot until I learned to use the umount command repeatedly with the -f option (I presume once for each open file), but this may be a umount or kernel bug, since there seem to be other chronic issues connected to dynamic block devices in Linux. In addition, this NFS problem seems to have disappeared in the last year or two (with Linux 2.6), so things may have improved as you surmise. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Stock Sarge segfaults, and later crashes!
I think this is the first time I have gotten segmentation faults using Debian stable, on stable hardware. I got them while, or after, reading several 1.44M floppy disks, mounted using mount or fdmount. Before the segfaults, I may have lost track of what I was doing and tried to mount a floppy more than once, possibly several times. The failure may also be related to previously using wine to unzip a DOS ZIP .exe file. I continued to lightly use the system for other unrelated tasks and it appeared to be stable, but later it crashed while shutting down, locking up with a stack trace dump, IIRC. (Another first for me, with Debian stable.) After fsck cleanup, the files I unzipped using wine, and several wine conffiles were gone, as were many other files throughout the root file system. The lost files were either open when I ran wine, or were subsequently opened. The kernel is a custom-compiled 2.6.8 stock debian kernel, on an all-stock, up-to-date sarge system. Due to the severity of the failure I am reluctant to try to reproduce it, but I hope this information helps someone else who runs into this problem. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Stock Sarge segfaults, and later crashes!
Roberto C. Sanchez wrote: It sounds like you may have a bad memory module. Go grab a Knoppix CD or DVD and boot it with the memtest or memtest86 command. You can check the cheat codes with F2 or F3 to see which it is exactly. Then let the box run for at least 12 hours on all the tests. Depending on the amount of memory and the speed of the machine, you may need to let it run significantly longer to do all the tests. I suspected that too, so I ran my own stress tests overnight without discovering a hardware problem. My tests have always found memory problems in the past, although sometimes they require a longer testing period. In addition, this system is my main personal computer, which has been stable for a long time. At this point it seems unlikely to be a hardware problem. I will run some more tests and report back if I find a hardware problem. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Stock Sarge segfaults, and later crashes!
Following up, after some research I found this c.o.l.m. thread may describe the same problem: http://groups.google.com/group/comp.os.linux.misc/browse_frm/thread/826d6ad5a33fc7a7/460a72cb19108f31?tvc=1q=2005+debian+segfault+OR+segfaults+OR+%22segmentation+fault%22+wine+OR+fdmount+OR+mounthl=en#460a72cb19108f3 or http://tinyurl.com/9xfb2 Note that, unlike the first poster in that thread, in my case I did get a kernel crash. Also note the punch line, which appears in the last post in the thread: if you want it the m$ way, just mount it with the sync option. if you want to use win, then use it. If this describes what I experienced, then it looks like it could be operator error because it's possible that when I lost track of what I was doing, I accidentally removed a mounted floppy without syncing. If so, it's surprising that a minor operator error should be so destructive and cause a kernel crash, so I would still probably consider it a bug, though not Debian-specific. The main problem may be that new users, possible coming over from the M$ world like first poster in this thread, would find it very easy to trigger this bug and cause major data corruption. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Fileserver performance
Cliff Flood wrote: Hi all, I just got a new Dell[0] workstation yesterday and have Debian Testing running on it. What I intend using this machine for is a file server (an iTunes share using a DAAPd, store time-shifted TV etc.), to replace an ancient PC running OpenBSD 3.5. I'm moving the large discs from this machine to the new machine. The discs are both Western Digital, 250GB w/1MB cache and 200GB with 8MB respectively. They perform well in the machine being replaced. I've hit a snag. When I mount a share from the Debian box on my PowerBook and play back some media, an avi for example, I get audio drop outs. I had a similar problem recently, related to a motherboard BIOS setting. If your NIC IRQ is set to edge-triggered, try changing it to level-triggered. I've seen other evidence that some Linux drivers may have a problem with edge-triggered IRQs. If you find that this is your problem, please report it here and I may research it further and consider filing a bug report. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Deleting XP and installing Debian
Scott Rebman wrote: Hi, I have recently acquired an older (Pentium II 350MHz) computer from my work and I was planning on getting rid of Win XP home that is currently on it and installing Debian. We use Debian at school and I was looking to play around with it at home to develop a better understanding of how it works. The problem is that I can't get it to boot up from the Debian install disk. I have binary 1, 2, and 3 burned and they seem to work fine on my main computer, but refuse to boot on the old one. If anybody has any ideas as to how I can get this done, please let me know. I do not need anything from the old computer so formatting it is totally ok. Thanks! Scott This is typical of old hardware. Try one of the installation floppy images on the installation CD-ROM, along with linux or DOS programs for making the install floppy. There are several kernels to choose from, depending on your hardware requirements. Read the instructions. After booting with the floppy, select CD-ROM as the installation medium. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Remote X login
marc wrote: Now it all works, I've got to say that this is a killer feature - although it would be nice to have multiple connects active simultaneously, even better with each as a virtual desktop in, say, KDE. You can use multiple virtual X displays for this purpose. Use startx in a virtual terminal to start the additional X displays, e.g. startx -- 1: (to start display 1, corresponding to F8) ctl-alt-F(7-12) are used to switch between the X displays, analogous to using ctl-alt-F(1-7) to switch between virtual terminals. More displays are available, but not keymapped by default. This feature is also used to allow different users to be logged into simultaneous local X displays. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
mysterious change in X authorization behavior
On one of my Sarge systems, if X access control is enabled, i.e. xhost yields the following output: access control enabled, only authorized clients can connect then if I su to root, X apps now start normally, as if I had previously executed xhost +. All my sarge systems are up to date, but only one exhibits this new behavior. The only difference I see related to X behavior, is that the changed system uses gdm instead of kdm. The others still output the following error message after su'ing to root: Xlib: connection to :0.0 refused by server I'm not aware of any recent configuration changes that could account for the change. Since this is a change in security-related behavior I am more anxious than usual to get to the bottom of it. Thanks for any clues about what may have changed, and how. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: The client-server terminology when referring to X (was: Remote X login)
Maxim Vexler wrote: The X server (which is running on the local laptop) obeys the demands of the application and draws what was requested from him, that is why the laptop in question is the SERVER. Not draws, but *displays* the client's output. Draw has a much narrower meaning in X or graphics display terminology. To understand X client/server design, it's helpful to review the history of UNIX and X graphical workstations. He served the application and drew what was requested from him. The significant resource consumption in on the _client_ and NOT on the server as one would expect. Not necessarily, and historically the resources were generally vastly different between the client and server, which is why they were partitioned that way. In the early days of X, CPU, storage and display resources were often too expensive to combine in a single box. It's not unlike giving an expensive printer its own server, nowadays. Then, X graphical displays were expensive and used mostly by CAD specialists, while everyone else used dumb terminals. What we now call thin clients would have been thin (X) servers then, with all the expense concentrated in the dedicated display hardware, with very little storage or general purpose CPU power. That is because all the rendering, data structures, CPU calculation cycles, I/O and so on is done on the client. The server (in X terminology) only receives raw drawing data, using the mentioned XDMCP transmission, and those act's in a vary mature manner He told me to, so I did it :) Even today, often the (server) GPU does more work than the (client) CPU, so your example is not necessarily valid, and definitely wasn't valid in the early days when graphics took relatively more processing power. Is my analogy correct ? /// Skipping the user interaction phase, in hope that someone could fill me in... If the laptop in the example is the server, and yet the user is obviously working on that laptop... How then does the application knows what to ask the server to drew next ? Who sends the input from the user back to the application so that it in turn could instruct the X server what to draw next ? /// I don't know the details, so I expect someone more knowledgeable to correct me or fill in the details, but IIRC the X clients and X servers (there can be more that one of each) communicate through the same X protocols whether they are local or remote. If client and server are both local then the physical network interface is bypassed and X protocol packets are routed locally. (The X clients themselves don't know or care where or how their output is being displayed, and likewise the X servers don't know or care where the clients are located.) If the X client user is running a remote X client, then the X protocol connection between client and server is handled seperately from the login session connection of the X client, although in the case of an ssh login, X can be tunnelled with ssh. (This has the advantage of being far more secure, especially over insecure networks.) Would love to read some clarification on this subject, as it is known to confuse a lot of (newbie) users like me. Try the X.org web site. -- Cheers, Maxim Vexler (hq4ever). Do u GNU ? -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
marillat false alarm?
Since the Oct 29 update of realplayer: # chkrootkit -q /usr/lib/realplay-10.0.6/share/default/.realplayerrc Besides having no idea what chkrootkit is complaining about, what really bothers me is having no way to validate marillat packages, since I'm running stable. (That's another issue which I've tried to address without success.) Thanks for any help. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: marillat false alarm?
Roberto C. Sanchez wrote: I'm not sure about the chkrootkit complaint. However, if you don't like Marillat's packages, or that you can't verify them, then quit using them. They are unofficial anyway, so I don't see the problem. I don't know of any official counterparts with equivalent functionality. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Update on mysterious change in X authorization behavior
Marty wrote: On one of my Sarge systems, if X access control is enabled, i.e. xhost yields the following output: access control enabled, only authorized clients can connect then if I su to root, X apps now start normally, as if I had previously executed xhost +. All my sarge systems are up to date, but only one exhibits this new behavior. The only difference I see related to X behavior, is that the changed system uses gdm instead of kdm. The others still output the following error message after su'ing to root: Xlib: connection to :0.0 refused by server I'm not aware of any recent configuration changes that could account for the change. Since this is a change in security-related behavior I am more anxious than usual to get to the bottom of it. Thanks for any clues about what may have changed, and how. Acting on a hunch, I replaced gdm with kdm and the problem seems to have disappeared, so I guess that's my workaround and as far as I will likely go with it. What's still puzzling is that the current Sarge version of gdm is date Mar 31, 2005, but this change in behavior is most likely very recent. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Network Stress Testing
Nelson Castillo wrote: Hi, I want to debug several ethernet links between PCs in the same subnet. Which tools should I use? What should I read about TCP/UDP stress testing? nttcp is good for simple tests, especially for testing raw throughput. You will want to watch your error counts with ifconfig or equivalent tool. For system-level stress testing I use NFS and simple test scripts, e.g. copying then comparing a DVD over NFS, possibly with background stress tests running. Autonegotiation testing is often overlooked, and is a common source of compatibility problems. This takes a lot of patience and time, and preferably a large variety of NICs, hubs and switches. If you compile your own kernels, you can often turn on various useful debugging options in your ethernet drivers. For server stress testing netperf was once an industry standard, but it may be overkill, and I don't recall whether the Debian version supports real clients or is just a simulation. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Soundcard: via82xx and alsa
micobros wrote: Hello all, I've been trying to install this soundcard for a couple days now. Its an Onboard VIA82xx chip. I tryed it on different kernels (2.6.0, 2.6.10 and 2.6.14) with this configuration: snip When i try to launch alsamixer: # alsamixer alsamixer: function snd_ctl_open failed for default: No such file or directory Try gnome-alsamixer. I have a similar sound chip and have never gotten any of the alsa-utils to work. I substitute sox play for aplay. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
OT: Mirroring via disk image file
I plan to mirror then replace a hard drive, using an identical model drive. Instead of a device-device copy e.g. cp /dev/hda /dev/hdb I hope to do cp old drive image followed by cp image new drive where image is a disk image file of the original drive. Will the physical sector arrangment be preserved using the intermediate image file, or is that only possible with the device-device copy? I'm asking here because don't have the identical drives available to test my theory beforehand. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: OT: Mirroring via disk image file
Roberto C. Sanchez wrote: If the drives are *exactly* identical, you can move your old drive to be the slave on the secondary ide channel, which makes it hdd. Then place the new drive on hda. Boot Knoppix, or another suitable live CD distro, and then do this: dd if=/dev/hdd of=/dev/hda That's equivalent my cp example and it doesn't solve the problem. I want (and need) to use an intermediate image file. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Apt Needs Counseling
Marty wrote: Freddie Witherden wrote: Here is the result of using that command: [EMAIL PROTECTED]:~$ sudo dpkg --force-all -P webmin-core (Reading database ... 75153 files and directories currently installed.) Removing webmin-core ... /etc/webmin/webmin.acl: No such file or directory dpkg: error processing webmin-core (--purge): subprocess pre-removal script returned error exit status 2 /etc/webmin/webmin.acl: No such file or directory dpkg: error while cleaning up: subprocess post-installation script returned error exit status 2 Errors were encountered while processing: webmin-core That just means that you somehow deleted or corrupted the post-installation script and unless that runs dpkg doesn't know if the package is really purged. Sorry, I just noticed that I was mistaken about the problem here. It looks like the removal scripts are complaining about a missing file, /etc/webmin/webmin.acl. Try: touch /etc/webmin/webmin.acl Then try to purge it again. Does anyone know a way of forcing apt/dpkg to forget that a package even exists? If you don't care about whether it's really purged, you can just replace the post-installation script with a do-nothing executable script, i.e. one that just the runs runs true or the equivalent. Otherwise you could replace the script and rerun the purge command. Since I don't run webmin I don't have it handy, but you can just extract it from the .deb using dpkg -x filename, directory -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Ready to join the club..
Greg wrote: I'm a noob to Debian but I'm ready to install Debian to my current machine. (PIII, 512MB Ram, 2 HDs; 60 MB - main and 80 MB secondary). The first HD contains WinME (don't laugh) and the second will contain Debian in one partition and Windows files (mp3s, JPEGs) in the other. I've already partitioned the second HD and burned the installation image files onto CDs. My question is this, I want to use a boot loader that will load either WinME or Debian. Grub seems like the default boot loader per the installation docs I've read. During installation, will Grub be smart enough to see WinMe on the other drive and will itput the boot loader file on the main drive, the one that holds WinME? I don't know the answer to question question, but it raises a red flag, byt suggesting that you are planning to install Debian on a system with WinME present. That's an unnecessary and risky thing to do, especially for a noob. I strongly recommend against it, especially if your system is not 100% backed up. Instead you could just remove or disconnect the WinME drive, to avoid the risk of total and irretrievable data loss on the WinME drive. This is not opinion or excess alarm on my part-- the Debian installation software will give you the same warning. Furthermore since you are using separate hard drives, you can probably select the boot drive using the BIOS boot device selection menu (e.g. F8-Boot Menu) available on newer motherboards. There is probably no need in your case to use a boot loader for OS selection. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Ready to join the club..
Antony Gelberg wrote: He is new to Debian but sounds like he has a decent understanding of PCs. If he is stupid enough to delete his Windows partition during the install, it will certainly be the kind of mistake he learns from. :) d-i is very friendly and I think he should proceed with both drives connected. Easy enough to say when somebody else's data is at risk, and the friendliness of all the strangers in the world may not be enough to recover his lost data. Moreover the reference to stupidity doesn't seem to cover the possibility of bugs or poor design in the installation routine, which could cause or contribute to catastrophic data loss. Furthermore since you are using separate hard drives, you can probably select the boot drive using the BIOS boot device selection menu (e.g. F8-Boot Menu) available on newer motherboards. There is probably no need in your case to use a boot loader for OS selection. Talk about making an inelegant solution (dual-booting) even more so. Elegance is in the eye of the beholder. To me, such redundancy does not seem elegant in the least, just a source of additional complexity and modes of failure. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Bourn Shell Script While Loop Problems
Martin McCormick wrote:
Roberto C. Sanchez writes:
Why not just use wc and sed?
A good idea and also a good idea from Rick Pasotto [EMAIL PROTECTED]
to use head -n -6. There is, however, a problem with that last
suggestion in that the script is actually being run on a FreeBSD
system and FreeBSD's version of head does not support the -n flag.
Both Debian and FreeBSD, however, had the same problem with the loop.
The function gets called with a bunch of values passed as
arguments instead of getting one value per run.
When I use sh -x scriptname, the output doesn't look like a
normal loop.
Thanks for any more ideas.
I don't know how to fix the awk problem, but instead of:
zlength=`wc -l $zonename|awk '{print $1}'`
you could try:
zlength='wc -l $zonename|(read len fname; echo $len;)'
Martin McCormick WB5AGZ Stillwater, OK
OSU Information Technology Division Network Operations Group
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Bourn Shell Script While Loop Problems
Marty wrote: zlength='wc -l $zonename|(read len fname; echo $len;)' Sorry, those single quotes should be backticks. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: OT: Damaged harddisk and/or disk controller - ps
Kjetil Kjernsmo wrote: On torsdag 20 oktober 2005, 01:27, Marty wrote: It could be a spin-up problem due to a worn out motor. Right. It is actually something like that that's my primary suspect. This would probably be reported by smartctl from the package smartmontools. Ah, thanks for the pointer! Got the daemon installed now, and did a run, I'm seeing these errors: Error 149 occurred at disk power-on lifetime: 1024 hours (42 days + 16 hours) When the command that caused the error occurred, the device was in an unknown state. Error 148 occurred at disk power-on lifetime: 1024 hours (42 days + 16 hours) When the command that caused the error occurred, the device was in an unknown state. Error 147 occurred at disk power-on lifetime: 1017 hours (42 days + 9 hours) When the command that caused the error occurred, the device was in an unknown state. Error 146 occurred at disk power-on lifetime: 1017 hours (42 days + 9 hours) When the command that caused the error occurred, the device was in an unknown state. Error 145 occurred at disk power-on lifetime: 1017 hours (42 days + 9 hours) When the command that caused the error occurred, the device was in an unknown state. Does that mean anything to you? That looks familiar. In my case it also warned of imminent drive failure, smartctl -H indicated that the disk was failing, and the BIOS SMART function warned about the disk. Swapping out the power supply seems to have cleared up all problems and the disk now passes all tests. Of course there's a chance your drive motor could really be failing. It's also possible that your power supply can't handle the powerup surge, and this can also mimic the motor spin-up problem. I've recently had this problem. Try swapping the power supply, preferable with a more powerful one. OK, it could be, but I put my hand in there last night after Alvin's post, and again this morning after power-up, and the PSU is cool, so it is nothing to indicate it is over-heated, at least. Furthermore, it is a 340W PSU, which should, according to spec, give more than enough power for my system, unless it is something wrong with it, of course... Surge performance has nothing to do with steady state power or heat dissapation. It's the ability to handle transient loads, and depends mostly on the quality of the design and components in the power supply. It may also be related to component aging or other degradation of the power supply. Is there any way I can measure the actual consumption? You probably need a storage scope and current probe. The best way is an inductive probe but that will probably be hard to find, and/or expensive. The quick and dirty way is to use a low value (~1Ohm) high power resistor calibrated against a measured 12V DC source, in series with the drive as it spins up. The storage scope measures the surge voltage across the resistor. This approach obviously has the drawback of increasing the voltage sag caused by the surge, so the smaller the resistor value the better. If you lack the test equipment, you might try to somehow use the motherboard 12V voltage sensor, if it's fast enough to catch the transient. That takes some imagination and effort. It would be interesting from several perspectives. I don't have a more powerful PSU available, and the cost of that means it gets even harder to decide what is the right thing to do... Since I need a larger disk anyway... If it's the power supply, then you may need to retire it anyway. Cheers, Kjetil -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: apt-0.6 off line usage
Brian Nelson wrote: Marty [EMAIL PROTECTED] writes: But again the main point is that the dist-upgrade option is not a supposed to be a routine procedure. Whatever gave you that idea? Switching to stable. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: apt-0.6 off line usage
Andrew M.A. Cater wrote: On Tue, Oct 18, 2005 at 09:38:58PM -0700, Brian Nelson wrote: Marty [EMAIL PROTECTED] writes: golfer wrote: The only way I seem to be able to get packages installed is to go back on line and do the 'apt-get dist-upgrade'. For one or two packages, this may be ok, but it's not something I want to waste time doing routinely. The dist-upgrade option is not intended for routine use. Unless you are mistaking it for the upgrade option, it may be a bug or a deprecated feature. An off-line dist-upgrade seems like a problematic feature to support. Huh? Consider the (admittedly rare) case - someone with CDs or DVDs and no readily available net access. [This happens to me at work more often than I'd like :( ] I agree that that's a valid definition of offline installation, but that's not what I was thinking when I used the word offline in my response to the original poster. He had described a process which I'd never considered before -- downloading the packages using apt-get's -d (download) option. I was expressing my doubts about the feasibility or safety of upgrading a distribution this way. But again the main point is that the dist-upgrade option is not a supposed to be a routine procedure. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: apt-get update problems
Jan C. Nordholz wrote: Hi! 1) several errors of followig form Release: The following signatures couldn't be verified because the public key is not available: NO_PUBKEY i used suggestions from http://lists.debian.org/deity/2005/08/msg00178.html to remedy the problem. what could cause this? apt tries to verify[1] the GnuPG signature of the Release files of each package server it connects to. By default, apt only knows of the Debian Archive Keys, which are used to sign the Release files of ftp.debian.org (and thus its mirrors, too). If you have external servers listed in your sources.list, their Release files will be signed by some other key - which apt doesn't know. 2) i can see the following Get:3 http://security.debian.org stable/updates Release.gpg [197B] Hit http://security.debian.org stable/updates Release Ign http://security.debian.org stable/updates/main Packages Hit http://security.debian.org stable/updates/main Packages i didn't mentioned it previously, so it's possible that the error is there for quite some time. btw i use amd64/testing so it may be caused by conflict between testing and stable. You need to use the security update site for testing. The recent announcement here included the following lines: We also invite you to add the following lines to your /etc/apt/sources.list file, and run apt-get update apt-get upgrade to make the security updates available. deb http://secure-testing.debian.net/debian-secure-testing etch/security-updates main contrib non-free deb-src http://secure-testing.debian.net/debian-secure-testing etch/security-updates main contrib non-free It's possible that your GPG verification problem is related to your attempt to use packages from stable, where it's not supported. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Apt Needs Counseling
Freddie Witherden wrote: Here is the result of using that command: [EMAIL PROTECTED]:~$ sudo dpkg --force-all -P webmin-core (Reading database ... 75153 files and directories currently installed.) Removing webmin-core ... /etc/webmin/webmin.acl: No such file or directory dpkg: error processing webmin-core (--purge): subprocess pre-removal script returned error exit status 2 /etc/webmin/webmin.acl: No such file or directory dpkg: error while cleaning up: subprocess post-installation script returned error exit status 2 Errors were encountered while processing: webmin-core That just means that you somehow deleted or corrupted the post-installation script and unless that runs dpkg doesn't know if the package is really purged. Does anyone know a way of forcing apt/dpkg to forget that a package even exists? If you don't care about whether it's really purged, you can just replace the post-installation script with a do-nothing executable script, i.e. one that just the runs runs true or the equivalent. Otherwise you could replace the script and rerun the purge command. Since I don't run webmin I don't have it handy, but you can just extract it from the .deb using dpkg -x filename, directory -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: OT: Damaged harddisk and/or disk controller - ps
Kjetil Kjernsmo wrote: And I have done nasty things to the system, just to see. For example doing aide --update, updatedb, an intensive write process and a CPU-intensive computation process simultaneously. It should put maximum stress on the system, both power-wise, disk use and CPU... Never seen any errors, and certainly no other problems... It is just that it isn't detected at startup... Detecting primary master None It could be a spin-up problem due to a worn out motor. This would probably be reported by smartctl from the package smartmontools. It's also possible that your power supply can't handle the powerup surge, and this can also mimic the motor spin-up problem. I've recently had this problem. Try swapping the power supply, preferable with a more powerful one. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Apt Needs Counseling
Freddie Witherden wrote: Hi, none of those worked. It seems to be because it can not remove (--purge) a package which does not have ant files and when I try to reinstall it apt tries to configure the packages which need it. I need a way of totally nuking those packages from apt's list so that it forgets that they even exist. This should always work: dpkg --force-all -P pkg-name Use caution, backup eveything and read the docs first. Try man dpkg and dpkg --force-help -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: apt-0.6 off line usage
golfer wrote: The only way I seem to be able to get packages installed is to go back on line and do the 'apt-get dist-upgrade'. For one or two packages, this may be ok, but it's not something I want to waste time doing routinely. The dist-upgrade option is not intended for routine use. Unless you are mistaking it for the upgrade option, it may be a bug or a deprecated feature. An off-line dist-upgrade seems like a problematic feature to support. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: exploring a M$ network
Joe Mc Cool wrote: The college I work for is an M$-only shop :-( I don't know if it is XP, or 2000 or whatever. I can plug my beloved sarge notebook into the network and use firefox etc no problem. But, how do I avail of the M$ disk spaces, printers etc ? Is samba (under the sarge) the way to go and if so what do I need to know about the M$ network in order to proceed ? To access a windows share you need smbclient. To print to a windows printer you need CUPS (or similar printing system) as well, and an entry like the following in /etc/cups.conf: DeviceURI smb://yourname:[EMAIL PROTECTED]/windowspc/windowsprinter Unblock smb and NETBIOS ports on your firewall. BTW if you use CUPS, I recommend the CUPS *server*, not just the client. I never was able to get the client alone to work and it seems to be a commonly reported problem. Thanks Joe -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: How to use old CPUs (Not Debian Specific)
Marc Shapiro wrote: OK. Like I'm guessing is the case with many of the users on this list, I have a multitude of PCs around the house. Four to be exact. I used to have three of them up and running, but that was before we moved. Now, since I have a DSL connection, I no longer have my desktop connected to an old machine with ISA slots just to use the hardware modem. We also have not yet found room to set up my daughter's computer. So only one PC is in use and the others are not even out of their moving boxes. Like I said, I don't have enough space in the new apartment to set up multiple computers, but I dislike having computing power going to waste. Can anyone suggest a way to network/connect all four to possibly distribute the load among them? I am considering building a custom case to hold all the MBs and only have a single monitor, mouse and keyboard connected. The case would, obviously, have to be fairly large, but it could then act as a table, as well, so the space would still be more efficient than four seperate cases which serve no othere purpose. I don't see much advantage over just stacking up the boxes in a corner or closet. In addition, the slow boxes aren't going to add appreciable power in a clustering arrangement, so that idea is out. This is the equipment that I have: My current Desktop unit - AMD Athlon XP 2400+ - 40 GB HD (less than half used) - DVD ROM / CDRW - 3.5 FD - Built in Audio - USB / Firewire Previous Net gateway- PII 350MH - 200 MB HD - 20 GB HD - CD-ROM - 3.5 FD - USB This one would make a reasonable spare desktop, especially if you avoid Gnome and KDE, and use a lean window manager instead. Daughter's PC - 486 SX 25 - 80 MB HD Other inherited PC - 486 SX 25 - 120 MB HD? (may actually be 80 + compression) The 486s are currently running Win3.1, but that can be remedied. In my opinion these are way too underpowered for most current X graphical apps. Even non-graphical debian utilities like apt will be too slow. Consider using one for a firewall(-router) and/or X-10 controller. Off the shelf DSL routers are poor firewalls and tend to have scary security holes. Leave Win3.1 on the other for legacy apps, games or Windows-only peripherals. I also have lots of cables of various sorts, extra mice and keyboards (that I hope NOT to use) a few small HDs (170 MB to 256 MB) a 10-BaseT card or two. 10Mb cards are good connecting to a DSL modem in a firewall box. Somewhere, there is also a 2GB HD that has a possibly non-funtional Win98SE install. It is probably in the PII box. 2GB would make good live mirror backup for a modest Debian installation. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: How to use old CPUs (Not Debian Specific)
Graham Smith wrote: The problem is there is a world of difference between doing up and old car as a hobby and trying to use a 486 as a desktop machine. You're ignoring the uses in between those two extremes. For example, why use a modern machine, which uses 3 or 4 times the power, just for a firewall or backup server? -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: How to use old CPUs (Not Debian Specific)
Hendrik Boom wrote: Actually, I've tried using old klunkers to do backups, and discovered that they can't take large hard disks. One of mine won't go beyond about 128 gig, tha other gets stuck somewhere between 2.5G and 80 G. I don't know if it will work for all old machines, but when I ran into that problem I turned off the backup disk in the BIOS, letting linux find it after booting from a smaller disk. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: SSH attack
Alvin Oga wrote: On Tue, 11 Oct 2005, Marty wrote: Thanks, you just reminded me of two more items for my ssh hardening plan: -deny root login -turn off sshd access after a specified number of failed login attempts, or any attempts outside the specific IP address range. those should be done BEFORE you go live .. ?? I mean before exposing an sshd port to the internet. - no machine i would be baby sitting would be turned on if those 2 minimum requirements is not met If your machines are all exposed to the internet or to an insecure LAN, then I don't see how you can safely use ssh at all. I would never attempt such a thing, so you are much braver than I. What I would do instead is limit ssh logins to a single heavily scrutinized, stripped and locked down, dedicated (internet) ssh server, which would be manually activated (maybe remotely) for each ssh use, and turn off all other times. - in the old days, i'd be running the latest/greatest ssh ... vs those that come with any distro ( it seems lot more stable now... not as many exploits ) as far as i'm concerned ... free audits is a good thing on non-critical machines ... let um play with those .. i get um by the thousands ... and i'm not gonna want any email just because one bozo decides to run a generic port scan or dictionary attacks Whatever, I would never dare expose any normal machine to the internet, especially one I have any responsibility for. I've never considered (until now) exposing any open port to the internet. I think that's what ISP servers are for. - that'd generate hundreds of thousands of false alarms - too many attempts will also raise a flag ( more than the number of your fingers ) Since I would enable the internet ssh server on a temporary and per use basis, to many email alerts would probably not be an issue. - critical machines are watched very carefully :-) c ya alvin -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: SSH attack
Dick Davies wrote: On 11/10/05, Marty [EMAIL PROTECTED] wrote: If your machines are all exposed to the internet or to an insecure LAN, then I don't see how you can safely use ssh at all. I would never attempt such a thing, so you are much braver than I. What I would do instead is limit ssh logins to a single heavily scrutinized, stripped and locked down, dedicated (internet) ssh server, which would be manually activated (maybe remotely) for each ssh use, and turn off all other times. 'maybe remotely' - aren't you just pushing back the problem? Yes it replaces one security headache with another, but having remote out-of-band access may be useful for other reasons, and therefore worth the risk. I first got the idea from ISPs which allow remote control of customer servers for reboots or maintenance. For example, I might use a modem on a system with no LAN connection, controlling an X-10 network. Then hopefully the worst damage an intruder could do is reboot or power off the servers. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: No Sound
Scarletdown wrote: On Tue, 2005-10-11 at 22:42 -0400, [KS] wrote: Scarletdown wrote: What do I need to modprobe to get this working? I tried modprobe es688, modprobe es1688, and modprobe ymf262, but those all failed with FATAL: Module not found. Try modprobe snd-es1688 and read the following page: http://www.alsa-project.org/alsa-doc/doc-php/template.php?company=ESS+Technologycard=.chip=ES688%2C+ES1688module=es1688 snip However, that site is useless to me, since I use oss insetead of alsa (which I have never been able to get to work). From /usr/src/linux-2.6.8/Documentation/sound/oss/README.OSS ESS ES1688 and ES688 'AudioDrive' based cards - Support for these two ESS chips is embedded in the SB driver. If you want to try another ALSA driver, from /usr/src/linux-2.6.8/Documentation/sound/alsa/ALSA-Configuration.txt: Module snd-opl3sa2 -- Module for Yamaha OPL3-SA2/SA3 soundcards. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: SSH attack
Alvin Oga wrote: On Tue, 4 Oct 2005, Henrique de Moraes Holschuh wrote: On Mon, 03 Oct 2005, Marty wrote: Correction -- it's in the hosts.deny man page. As others have already pointed out, sshd must be configured to start via inetd. Must it? It uses tcp-wrappers natively, it should not need inetd for ANYTHING concerning /etc/hosts.allow or /etc/hosts.deny. I should have said that inetd must be configured to start sshd, if you want it to mail information on refused login attemts. simple test ... ( use your positive or negative logic equivalents for these files ) /etc/hosts.deny ALL : ALL I'm not sure that will work with the manpage example I gave. At least you can get an equivalent effect by adding an entry for each server started by inetd. and try to make your ssh work .. when yu give yp .. you'll find that you will need to have /etc/hosts.allow sshd : 192.168.123.456 restart the inetd or sshd as needed Don't forget to add your ssh entry in /etc/inetd.conf. whether inetd is good or bad is a separate issue Now that you menion it, this is probably not a good way to detect failed ssh login attempts for two reasons: because those packets should probably be blocked at the firewall; and because it also won't report failed attempts on the permitted IP address space. Before suggesting that I skimmed over the ssh/sshd manpages to see if they supported an automatic email alert option. ssh did mention an option for supplying a users email address, but I didn't find out what its used for. At first I thought the best way to do this is in a network monitoring tools, but I don't know whether such tools would normally detect an application-level event like a failed login attempt. As for tools that just monitor logs, I'm not sure whether they can respond fast enough. c ya alvin -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: SSH attack
Alvin Oga wrote: On Mon, 10 Oct 2005, Marty wrote: simple test ... ( use your positive or negative logic equivalents for these files ) /etc/hosts.deny ALL : ALL I'm not sure that will work with the manpage example I gave. works for me ... no services coming in that is not supposed to I meant sending the email alert as described in the manpage. you'll find that you will need to have /etc/hosts.allow sshd : 192.168.123.456 restart the inetd or sshd as needed Don't forget to add your ssh entry in /etc/inetd.conf. everything is turned off, sshd is NOT listed in inetd.conf or xinetd.conf whichever one is being used .. and similarly for the [x]inetd daemon itself - sshd does its own magic based on the allow/deny entries I had forgotten about that. It really got me the first time I tried to run ssh, and it doesn't seem to be well documented anywhere. Still it seems better to start ssh from (x)inetd for security reasons. grep whatever you like from the gazillion log files for ssh this and ssh that I don't know what you're getting at here. The idea is to get a realtime email alert. c ya alvin -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: SSH attack
Alvin Oga wrote: On Tue, 11 Oct 2005, Marty wrote: grep whatever you like from the gazillion log files for ssh this and ssh that I don't know what you're getting at here. The idea is to get a realtime email alert. one can get any and all kinds of alerts till you're blue ( satisfied ) - just write it the way you like The problem is that method I proposed, using hosts.deny, doesn't cover all cases, no matter how much code I write. I already gave one reason and there are a few more I can think of. if you use somebody else widgets, you are restricted to what they did and what it does That's fine and even preferable, if it does everything I need. - if you are cracked or about to be cracked, you have less than 5 seconds to close the exploit before they can do the rm -rf / The whole point is to respond to the breakin attempts as quickly as possible, since most are not successful on the first attempt, and your example presupposed that the attacker immediately gets root access on the first attempt, which is not very likely. Thanks, you just reminded me of two more items for my ssh hardening plan: -deny root login -turn off sshd access after a specified number of failed login attempts, or any attempts outside the specific IP address range. c ya alvin -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Laptop clock is localtime, system time is off
Nate Bargmann wrote: I use a laptop that multiboots with XP and I just set up a second hard disk for it with Sid installed. As a result, the hardware clock is set to local time and I can't seem to change Sid's mind on this. Per the Debian GNU/Linux Administrator's Manual, the UTC variable in /etc/default/rcS controls this behavior. After changing it to no and restarting the system, my time is still off. Somewhere the system is getting told to subtract an offset from the hardware clock, and I wonder if I'm looking in the wrong place. There's a 2.6 APM kernel option called RTC stores time in GMT (APM_RTC_IS_GMT) you might want to check. Thanks! - Nate -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Debian Compatable UPS?
Henrique de Moraes Holschuh wrote: On Sat, 08 Oct 2005, Hugo Vanwoerkom wrote: all UPS's are Debian compatible as long they have an rs232 interface. Which statement I don't understand. Why is USB bad? What do I do with M$ code that is shipped with most? USB is not bad (but good luck trying to find USB power surge isulators, which are quite easy to find for RS232), but its support in Linux is not anywhere close to the RS232 support (smart protocols). I think the problem with any serial protocol is DC signalling, which is vulnerable to surges, a solved by an ethernet connection. I recently lost three serial ports in two computers due to a surge, one of which was connected to an APC UPS. (A SmartUPS 700) It defeats the purpose of a UPS somewhat, if it destroys your hardware. I've never heard of power surge insulators, and a web search turns up nothing. Are these specific to serial connections? Could you supply more information and maybe some web site links? -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Debian Compatable UPS?
Henrique de Moraes Holschuh wrote: On Sat, 08 Oct 2005, Marty wrote: I've never heard of power surge insulators, and a web search turns up nothing. Are these specific to serial connections? Could you supply more information and maybe some web site links? I have no idea how they are called commercially in english. I will do a search and see if I can find one in english. In the meanwhile, please try serial line surge supressor or maybe surge protector... I found this on the APC web site: http://www.apc.com/resource/include/techspec_index.cfm?base_sku=AP9825# You have to dig a little deeper for the price: http://tinyurl.com/a65bu And it only costs $209.99! What a bargain! Only a little more than what I paid for the UPS! -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Audio player error (amarok)
Jeremy Merritt wrote: I have recently been having an error with several audio players, including amarok and xmms. Amarok produces the following error message: Did you recently convert to udev? (See below) [GStreamer error] Could not open device '/dev/dsp' for writing. ** gstosselementc(752): gst_osselement_open_audio /root/bin system error: Resource temporarily unavailable xmms produces a similar error message about 'cannot write to /dev/dsp'. What is your xmms setting for options-preferences-output plugins? I suspect it's set to OSS. This probably means either you don't have OSS (or ALSA OSS emulation) or you don't have a /dev/dsp softlink to your dsp device. Either create the link, or if using udev, add a line similar to KERNEL=dsp2 SYMLINK=dsp to your udev.rules file, replacing dsp2 with your dsp device. It's likely that your issue with amarok is similar. Even if this solves your problem, consider switching to ALSA (if amarok supports it -- I'm sure that xmms does) because I think that OSS is going to be deprecated eventually. However, Noatun has _NO_ problem playing any audio files. This could mean it uses ALSA. Any suggestions? TIA __ Yahoo! Music Unlimited Access over 1 million songs. Try it free. http://music.yahoo.com/unlimited/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Upgraded from 512 to 1024 ram. Now, how to fine tune the system?
Ron Johnson wrote: On Fri, 07 Oct 2005 14:10:58 -0300 Bruno Buys [EMAIL PROTECTED] wrote: Ron Johnson wrote: On Fri, 7 Oct 2005 16:45:21 +0300 Bogdan Rotariu [EMAIL PROTECTED] wrote: Hello Bruno, Friday, October 7, 2005, 12:41:33 AM, you wrote: Just bought an extra 512mb ram module, to add to my existing 512mb. Free reports the total correctly, but i was wondering if is there any way that i fine tune my system for better use of this memory. That's the first time i run a 1gb pc :)! Any help? [snip] or update your kernel if u didn't alredy do that. Why? I didn´t understand, also. I was after any kernel parameter, command line option to include in my boot, or something like that, to tell my system how to use the memory. IF there is such a command. I remember having read some time ago about commands telling the kernel what to swap and what not, and that it had to do with mem upgrades. That´s it. Hope it cleared up the subject. Oh, ok, I think. The only kernel configuration option is to tell it what memory range you have. The choices are 1GB, 4GB 64GB. The sid 2.6.12 binary kernel has it set for 1GB. So, if I were to add more RAM to my current 1GB, I'd have to build a custom kernel. As for telling the kernel what to swap, AFAIK, there are no build or boot parameters to control that. Linux knows how much RAM swapspace you have, and does what it thinks is best. Due to kernel address map limitations, unless your kernel is configured for at least 4GB, only about 900MB will be recognized. You can tell if you have this problem by running cat /proc/meminfo. The first line of output should be: MemTotal: 1034116 kB Otherwise you need to recompile your kernel. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Upgraded from 512 to 1024 ram. Now, how to fine tune the system?
Bruno Buys wrote: Marty wrote: Due to kernel address map limitations, unless your kernel is configured for at least 4GB, only about 900MB will be recognized. You can tell if you have this problem by running cat /proc/meminfo. The first line of output should be: MemTotal: 1034116 kB Otherwise you need to recompile your kernel. Hi marty and others. Here's my meminfo. I guess I'm ok, then? [EMAIL PROTECTED]:~$ cat /proc/meminfo MemTotal: 1036524 kB That looks good. I expect a few MBs difference due to legacy hardware issues, chipset differences and BIOS settings. Finally you may want to think about testing the memory, especially if you suspect any instability, but that's a different thread. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: How to completely reinstall a package?
Steve Block wrote: On Fri, Oct 07, 2005 at 03:47:29PM -0500, Tim McDonough wrote: I'm running Debian (Sarge) and had used X-Windows very little on what's primarily a file server when I got a larger, nicer monitor for the system. It was not obvious to me how to reconfigure the system for the better capabilities of the monitor so I, perhaps unwisely, decided I could simply uninstall then reinstall X-Windows and KDE. I ran... apt-get --purge remove kde apt-get --purge remove x-window-system Once that was done I used apt to install these pieces again but there are apparently remnants of the originals around so it never takes me through the configuration screens and I cannot get the windowing programs to run. In fact I cannot even find the startx command on the machine. I'm certain this is some sort of classic newbie screw-up. Would someone explain what I've done wrong and how I might re-install things properly? Any help would be appreciated. Running dpkg-reconfigure package_name will take you through the configuration stuff again. As far as missing programs such as startx goes, I am unsure. From apt-file search startx: xbase-clients: usr/X11R6/bin/startx What probably happened is that apt-get automatically removed xbase-clients for dependency reasons, and reinstalling x-window-system won't automatically reinstall any reverse dependencies that got removed. This is a good lesson on keeping logs of of your apt-get session, or at least keeping a list of installed packages by running dpkg --get-selections *. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: How to completely reinstall a package?
Marty wrote: This is a good lesson on keeping logs of of your apt-get session, or at least keeping a list of installed packages by running dpkg --get-selections *. Sorry, make that dpkg --get-selections (no asterisk). -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
testing memory (was Re: Upgraded from 512 to 1024 ram. Now, how to fine tune the system?)
Bruno Buys wrote: Marty wrote: Finally you may want to think about testing the memory, especially if you suspect any instability, but that's a different thread. I did. That was the very first thing I did after hooking up the module. I was afraid what a defective module could do to my beloved debian system, so I run memtest86 for 30 min without problems, before booting debian with the new module. My prior experience with memtest was that it would find problems before 2 or 3 min running. What do you think? I've had modules that require hours or even days of heavy stress testing to elicit failure. Under normal curcumstances they might run for weeks or months without any obvious failure. I think this type of marginal defect is much more common than any that can be found in a few minutes of testing, because thr latter are much easier to screen by the retailer, middleman or manufacturer. I used the same set of tests whether I'm adding or changing memory, a processor or a motherboard. Even if one combination of components tests good, I retest them all if any one component is changed. To invalidate the kernel's file cache, I use custom shell scripts to continuously read and check a large number of files (e.g. looping debsums runs, checksum tests of my local debian archive, or copying and then checking the contents a DVD). In addition I run memtest and/or do looping kernel compiles while checking the results using a modified version of the old burnit script which was used years ago to test for defects in certain types of AMD processors. This test is a particularly effective memory test due to the use of self-modifying code by GCC. While monotoring a temperature probe inside the box, and temperature monitoring tools like mbmon and hdtemp, I reduce airflow and/or test in a small enclosed space, allowing the (modules') ambient temperature to rise to 40C or higher. Under these conditions I allow the tests to run from overnight to as long as a week in duration. In case of a crash I run the tests from a remote system via ssh sessions, so that the results are logged in the terminal sessions. These tests are probably overkill for a personal desktop system, but I run them because I dislike flaky hardware and because I always buy the cheapest hardware I can find with the expectation that a certain percentage of it will be returned or scrapped. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: boot stalls on ntpdate [was bootlog, control of modules]
Matt Price wrote: On 10/3/05, Edward J Shornock [EMAIL PROTECTED] wrote: Edward J. Shornock wrote: A moot point now... sysvinit (2.86.ds1-4) unstable; urgency=low this is great, thank you. I now have bootlogd running! (found htis out when I finally had an opportunity to reboot, bit of a stressful weekend). COmmenting out a bunch of modules helped my problems, thanks much for that help folks. I notice now that bootup is stalling on ntpdate. Here's what I see on bootlog... Tue Oct 4 21:13:32 2005: Running ntpdate to synchronize clockError : Temporary failure in name resolution Tue Oct 4 21:14:12 2005: . so it hangs for about a minute. I assume that the internet connection has not been brought up yet when ntpdate is run... any suyggestions what I should do? If it's a name resolution problem, then try using the IP address of a reliable time server, like your ISP's. If it's a time out issue then you can try lengthening the timeout period using the -t option. Try both if you're not sure. The init script for ntpdate is currently /etc/rcS.d/S51ntpdate, which i think is the default. That's just a link. The init script is /etc/init.d/ntpdate, but even that is not what you want. In debian, init scripts have files of the same name in /etc/default, reserved for custom configuration, so you can add the changes there (in /etc/default/ntpdate). thanks! matt -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: SSH attack
Jared Hall wrote: It looks like I am being rooted right now. How do I toss this guy off of my system. he has an IP address of 210.95.212.131 It's a kid! Whois returns Hanguk Kwangsan Technoledge High School. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: SSH attack
Alvin Oga wrote: - if it was a hole in ssh, ALL and i mean ALL other Debianites and possibly other Linuxites will be equally susceptable and some of of them will have noticed that they too were successfully attacked == == time for you ( marty ) change the way you use ssh and/or the way you == log into your PC and/or update your PC, or let it run and see if == you can stop them from loggin in == I don't knowingly expose any ports or services to the internet, including ssh. As you point out, there's no guarantee it can ever be done safely, due to the possibility (certainty?) of holes in ssh. If I did attempt it, however, I would at least take the following precautions: -always require private keys -use a port other than 22 -run sshd only at certain times of the day (e.g. 5 minutes per hour) -restrict ssh connections to a single remote IP address or subnet, in the firewall as well as in sshd -configure the ssh server to report any successful ssh login using email, and/or send a page or cell phone alert -do the same for mutliple failed connection attempts -in conjunction with the previous items, provide a way to remotely disable ssh or initiate an emergency shutdown in case of suspected atteck. This list come from my own limited knowledge of the subject. I'd appreciate any additions to this list from any security experts reading this. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: playing midi with timiditySOLVED
Joe Mc Cool wrote:
So, I started again.
aptitude remove timidity
and, to be sure to be sure, as root:
find . -name timidity* -exec rm {} \;
That won't necessarily work. You need to follow up with dpkg --purge timidity
To do it all in one operation, use the purge option of aptitude, or
else use apt-get remove --purge timidity
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: SSH attack
Landy Bible wrote: Marty wrote: -configure the ssh server to report any successful ssh login using email, and/or send a page or cell phone alert I can only guess at this point because I've not tried it. A crude example might be using a login script to detect whether the shell is starting in an ssh session, e.g: if [ `pidof ssh` != ] do /usr/bin/mail -s ssh login at `date` [EMAIL PROTECTED]tail /var/log/authlog ... Here I might call chat or a similar program to send the page or phone alert, using a modem in the system that's not used for my main internet connection. I would run something like this on the gateway/firewall and/or on an internal system which has ssh forwarded to it. I'm not sure which method would be more secure. -do the same for mutliple failed connection attempts Could some one point me at a way to do this? There are probably special tools for this, but an easy way is to use inetd: From http://www.freeos.com/articles/2896/: If you would like to know about the failed connection attempts to your machine then change the above entry to the following [in /etc/hosts.deny]. ALL:ALL:/bin/mail -s %s connection attempt from %c [EMAIL PROTECTED] The inetd man page gives an example for use with a specific service: /etc/hosts.deny: in.tftpd: ALL: (/usr/sbin/safe_finger -l @%h | \ /usr/bin/mail -s %d-%h root) An example using the SPAWN command is given here: http://www.debian.org/doc/manuals/securing-debian-howto/ch4.en.html#s-tcpwrappers ALL: ALL: SPAWN ( \ echo -e \n\ TCP Wrappers\: Connection refused\n\ By\: $(uname -n)\n\ Process\: %d (pid %p)\n\ User\: %u\n\ Host\: %c\n\ Date\: $(date)\n\ | /usr/bin/mail -s Connection to %d blocked root) -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: SSH attack
Marty wrote: The inetd man page gives an example for use with a specific service: /etc/hosts.deny: in.tftpd: ALL: (/usr/sbin/safe_finger -l @%h | \ /usr/bin/mail -s %d-%h root) Correction -- it's in the hosts.deny man page. As others have already pointed out, sshd must be configured to start via inetd. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: pointer for home networking
I forgot to add that for DNS and DHCP I recommend dnsmasq, which is very easy to set up. (For me just one line added to the config file.) -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: pointer for home networking
michael wrote: I've had a look about but can't find a basic guide to setting up a home network. There seems much discussion of 'deeper' stuff but I'm stymied for setting up my first home Debian/Linux network. I've a computer that did have Internet connection via ethernet to a modem router. It's now connected to Internet by a USB modem. I've a second computer with an ethernet card. So all I want to do is connect the latter to the former such that both can access the Internet... All pointers to (simple) (online) (Debian specific) references most welcome! Thanks, Michael This is not so much a how-to as a list of general themes that you might want to think about. There are plently of how-tos that get into the details, but don't hesitate to ask if you get stuck on any detail. First you will have to compile the appropriate drivers into your kernel. I recommend reading through the driver options while running make xconfig and paying special attention to the networking options, like ethernet bridging, filtering and forwarding. You will probably end up enabling nearly all the networking options. Then you will need the user space firewall software. I've had good results with the graphical apps called guidedog and guarddog, for firewall/bridging functions respectively. The are fancier ones but for home use I think those are fine, and also very easy to use. (Reading the docs for these will help give you a basic understanding of how your gateway will work.) After installing the firewall software, turn off all unused ports. In addition, carefully go through your init sequence and turn off unused servers and daemons. On my firewall I also remove the unused server packages and all other unused software, and access the box only through a serial port, for security. (This latter option has to be enabled in your kernel). If you are concerned about power, then you might also remove unused peripherals and extra memory, and underclock the processor. It takes very little processing power to function as a firewall. Consider using a fan-less heatsink for reliability reasons. Since the firewall serves a vital function, I make my more fail-safe by using a second drive, kept up to date using rsync, as a mirror of the root drive which I can always use if the root drive fails. A secondary purpose is for use in case I suspect that the machine has been compromised. Finally, you may want to install some system monitoring software like tiger, and some network monitoring tools like snort. When you see all the benefits of having a custom firewall, I predict you will never go back to an off-the-shelf router. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: NUM Lock , Home, End
John Hasler wrote: Marty writes: Some init scripts use configation files in /etc/default, and I guess the script's stop routine could store the NUMLOCK state upon system shutdown, if that state is accessable from the system. That would set it globally, not for each user. I agree, and I see that I missed your point. On the other hand it could solve the original problem in the majority of practical cases. Maybe the ultimate solution would be a mechanism for changing global defaults on the fly, either automatically or by user initiative. I'm thinking of something along the lines of udev/hotplug, which prove that even Unix can learn new tricks, although I'm not sure about the level of kernel involvement. Whether multi-user PCs represent a large enough user base to justify the effort, is another matter. Since Windows was brought up earlier in this thread, I'll just point out that the issues are much simpler because Windows doesn't allow more than one concurrent user logging in on the console, so it's not a fair comparison. Given a choice I would rather live with Debian's limitations. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: NUM Lock , Home, End
[EMAIL PROTECTED] wrote: you have numlock enabled for every console? and HOME return you in begining of command promt and END in the end? i have to press ctrl+a or ctrl+e for VC's use setleds for X use numlockx home and end work as expected for me. -matt zagrabelny uhhh i know how to set them but it is very stupid in my opinion ... you cant use numlock for anything else.. there are other arrows.. I never set them, and these keys work for me both in VTs and X apps, on my Sarge systems. Earlier versions of Debian and X had problems like this. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: NUM Lock , Home, End
Seth Goodman wrote: From: Marty [mailto:[EMAIL PROTECTED] Sent: Thursday, September 22, 2005 6:10 PM ... I never set them, and these keys work for me both in VTs and X apps, on my Sarge systems. Earlier versions of Debian and X had problems like this. Well, my BIOS sets NUMLOCK on, but Sarge and Etch both turn it off, at least with the gnome desktop. I have to manually turn NUMLOCK back on. I was referring to keymapping issues. Sorry if I misunderstood the issue here. I agree that the default state of NUMLOCK should be easily configurable, and possibly even stored between reboots. Even the bozos in Redmond have figured out that users want the computer to remember how _they_ like it without the user taking any action or editing any initialization scripts. If we want Linux to be the ubiquitous OS, and not just for the cognoscenti, we can't ignore details like this. Sounds like a good reason to file a wishlist bug report. -- Seth Goodman -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: NUM Lock , Home, End
Ron Johnson wrote: On Thu, 2005-09-22 at 18:29 -0500, Seth Goodman wrote: From: Marty [mailto:[EMAIL PROTECTED] Sent: Thursday, September 22, 2005 6:10 PM ... I never set them, and these keys work for me both in VTs and X apps, on my Sarge systems. Earlier versions of Debian and X had problems like this. Well, my BIOS sets NUMLOCK on, but Sarge and Etch both turn it off, at least with the gnome desktop. I have to manually turn NUMLOCK back on. Even the bozos in Redmond have figured out that users want the computer to remember how _they_ like it without the user taking any action or editing any initialization scripts. xnumlock If we want Linux to be the ubiquitous OS, and not just for the cognoscenti, we can't ignore details like this. File a bug. But against what? Each WM? I was thinking it should be handled during boot by an init script like keymap.sh (just a guess). Anything else that automatically changes the state of NUMLOCK (e.g. Gnome) could be considered a bug. First, however, you might want to check to see if one of the derivative Odistros like Ubuntu have already solved this problem. This is the just kind of integration issue they specialize in. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: NUM Lock , Home, End
John Hasler wrote: Marty writes: I was thinking it should be handled during boot by an init script like keymap.sh (just a guess). How would that set it correctly for each user? Some init scripts use configation files in /etc/default, and I guess the script's stop routine could store the NUMLOCK state upon system shutdown, if that state is accessable from the system. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: backup plan bare metal
Rodney Richison wrote: Would be interested in seeing what some of you use for a backup plan. Mainly for servers. Tar? Easy/quick way to restore bare metal? Rsync? Can it do bare metal? what about hard links? It's possible using a rescue floppy or knoppix. Again, easy/quick way to restoe entire debian server? I use a mirror drive. Recovery is as easy as selecting a new boot drive in the BIOS. You can replace or restore the original drive at your leisure. Unfortunatly, I've found mondo unreliable for bare metal. Though when it works, it's the cat's meow.. I don't see much advantage. You can just mount the replacement drive in a working system and restore it from the backup server. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Network (pcmcia card) not enabled after boot [newbie alert]
Peter Coppens wrote: If I remove pcmcia-cs (or at least make sure it is not initialized by removing the links from the /etc/rcx.d directories) I can't even manually start the network (with ifup -a) after booting I'm not familiar enough with pcmcia-cs nor hotplug to guess what's happening, but beware that for a variety of reasons just disabling a package is not the same as removing it, at least for the purposes of your test. For example the packaging system's diversions may affect both packages and the only way to eliminate any interaction may be to remove one or the other package. On the other since you're a newbie be cautious about removing hotplug, since it could make your system unbootable. Hopefully somebody more familiar with your issues can help at this point. If I first manually start pcmcia I can ifup the network. Any guidance would be appreciated. Thanks, Peter -Original Message- From: Marty [mailto:[EMAIL PROTECTED] snip If you have cardctl then I assume you are running pcmcia-cs. It's not clear to me why you would need both that and hotplug, because it seems that their functions overlap (although hotplug is listed in the recommended section of pcmcia-cs). Here is a link I found that supports this theory: http://lists.debian.org/debian-laptop/2003/07/msg00234.html -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: eth0 problem
Deephay Z wrote: Greetings all, since I need to install the wireless device driver for the Debian system, so I compiled and installed the 2.6.13.2 version kernel, but the eth0 (wired-ethernet card) now cannot be found using the newer kernel. Some kinda :eth0: no such device is displayed on the screen during booting procedure. I guess that's because I incautiously removed the necessary ethernet card driver when I was doing the make menuconfig part. How can I re-install those drivers for the eth0? (I am not sure which one is the precise driver, but I'm sure it is contained in the kernel source). Thx a lot! Regards, Deephay Z To identify the NIC device, try lspci (if it's a PCI device) or else grep for eth0 in your boot message logs. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Network (pcmcia card) not enabled after boot [newbie alert]
Peter Coppens wrote: All, I have installed the 'latest stable' Debian version on an (old) Dell laptop Cpi D300XT. The initial install was done with the laptop in a docking station and using the network adaptor that comes with that docking station. Everything went fine. Later I removed the laptop from the docking station and tried to enable a PCMCIA Xircom CBEM56G network/modem card. That is giving me a headache. The boot log contains ... Mon Sep 19 13:11:28 2005: Setting up networking...done. Mon Sep 19 13:11:29 2005: Starting hotplug subsystem: Mon Sep 19 13:11:29 2005:pci Mon Sep 19 13:11:30 2005: usb-uhci: already loaded Mon Sep 19 13:11:30 2005: insmod: /lib/modules/2.4.27-2-386/kernel/drivers/usb/host/uhci.o: init_module: No such device Mon Sep 19 13:11:30 2005: insmod: Hint: insmod errors can be caused by incorrect module parameters, including invalid IO or IRQ parameters. Mon Sep 19 13:11:30 2005: You may find more information in syslog or the output from dmesg Mon Sep 19 13:11:30 2005: insmod: /lib/modules/2.4.27-2-386/kernel/drivers/usb/host/uhci.o: insmod uhci failed Mon Sep 19 13:11:30 2005: uhci: can't be loaded Mon Sep 19 13:11:30 2005: missing kernel or user mode driver uhci Mon Sep 19 13:11:30 2005: yenta_socket: already loaded Mon Sep 19 13:11:30 2005: neofb: ignoring pci display module Mon Sep 19 13:11:30 2005:pci [success] Mon Sep 19 13:11:31 2005:usb Mon Sep 19 13:11:35 2005: usbcore: already loaded Mon Sep 19 13:11:35 2005: usbcore: already loaded Mon Sep 19 13:11:37 2005:usb [success] Mon Sep 19 13:11:37 2005:isapnp Mon Sep 19 13:11:37 2005:isapnp [success] Mon Sep 19 13:11:37 2005:ide Mon Sep 19 13:11:37 2005:ide [success] Mon Sep 19 13:11:37 2005:input Mon Sep 19 13:11:37 2005:input[failed] Mon Sep 19 13:11:37 2005:scsi Mon Sep 19 13:11:37 2005:scsi [success] Mon Sep 19 13:11:37 2005: done. Mon Sep 19 13:11:37 2005: Setting up IP spoofing protection: rp_filter. Mon Sep 19 13:11:37 2005: Configuring network interfaces...Internet Software Consortium DHCP Client 2.0pl5 Mon Sep 19 13:11:38 2005: Copyright 1995, 1996, 1997, 1998, 1999 The Internet Software Consortium. Mon Sep 19 13:11:38 2005: All rights reserved. Mon Sep 19 13:11:38 2005: Mon Sep 19 13:11:38 2005: Please contribute if you find this software useful. Mon Sep 19 13:11:38 2005: For info, please visit http://www.isc.org/dhcp-contrib.html Mon Sep 19 13:11:38 2005: Mon Sep 19 13:11:39 2005: eth0: ERROR while getting interface flags: No such device Mon Sep 19 13:11:40 2005: Bind socket to interface: No such device Mon Sep 19 13:11:40 2005: exiting. Mon Sep 19 13:11:40 2005: Failed to bring up eth0. Mon Sep 19 13:11:40 2005: done. ... After booting the card seems to be there (cardctl info shows it), but the network is not. If after booting I execute ifup -a it is enabled and the network works fine. I have tried to follow the procedure in http://www.debian.org/doc/manuals/reference/ch-gateway.en.html#fr67 (10.10.2 Triggering network configuration - hotplug) but that does not help me (same problem). Anyone any thoughts or pointers to documentation to help me out with this, Thanks, Peter If you have cardctl then I assume you are running pcmcia-cs. It's not clear to me why you would need both that and hotplug, because it seems that their functions overlap (although hotplug is listed in the recommended section of pcmcia-cs). Here is a link I found that supports this theory: http://lists.debian.org/debian-laptop/2003/07/msg00234.html -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Install USB Mouse, Lose Keyboard
Mike McCarty wrote: Recently, my girlfriend decided to try the jump to Linux. I helped her build a dual-boot system with Windows and Debian on it. She has run it fine for a couple of weeks, except that her mouse goes crazy sometimes. This occurs both with Debian and with Windows. So I suggested she has a hardware problem with the mouse. I gave her a USB optical mouse I have used before, and it works just fine with both Windows and with Debian. However, Debian no longer recognizes her keyboard. It sees the mouse fine, but not the keyboard. Windows still sees and uses both the mouse and the keyboard with no problem. You forgot to mention whether the keyboard is USB or not. (See below) You also didn't specify what it means to say that Debian no longer recognizes her keyboard. Are you running X when the keyboard is not recognized? X has an inherent design flaw that makes it prone to this problem (although I would expect it to lose the mouse as well). To sort it out you could check if the problem occurs on a virtual console, peferably with no X server running. The motherboard is about two years old with a 2+Ghz Celeron on it, 512 MB RAM. I have booted Knoppix 3.9 (Debian variant) and run memtest overnight with no problems indicated. Does anyone have any suggestions? Mike The obvious work-around is to try a PS/2 keyboard (or USB, if it's currently PS/2). For debugging, if the problem is with USB, I would look at the hotplug output in the log files. For example you could log via ssh and run tail -f /var/log/messages and plug and unplug the keyboard and see what hotplug and the USB driver outputs (or doesn't). You could also plug and unplug the mouse, while checking the keyboard and log output. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: minicom display problems
Miquel van Smoorenburg wrote: In article [EMAIL PROTECTED], Marty [EMAIL PROTECTED] wrote: Is there any way to get minicom to properly display the color graphics of an ncurses app like iptraf? I am using the multi GNOME terminal as my X terminal, but xterm seems to have the same problems. So far I've tried: -setting the minicom terminal type to ANSI -setting the remote shell TERM type to xterm (it starts up as VT100) -turning the minicom color on using the -c on option minicom -c on should start minicom in color. If ctrla-z shows the command summary in color, minicom supports color. Then you need the right terminal type on the remote system to tell its apps to use ansi-color. Try TERM=xterm-color or TERM=linux Thanks. Now I have color. The ncurses border line graphics are still not rendered correctly, however. So you have to change that manually. Use stty rows 40 columns 120 on the remote system, or run resize if that command is installed. (See man stty, man resize). These both work too. Thanks again. [I notice that resize on my Linux box doesn't set the tty properties but prints out COLUMNS= / LINES= .. how 1980's ..] I mercifully forgot everything I knew about terminals in the '80s, But now I am puzzled by two purely theoretical questions: 1) why don't LINE and COLUMN get listed when I run printenv? 2) How does resize retrieve the TTY properties via minicom, and why can't the remote shell automatically do likewise? -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: absurdly simple LAN problem
Anthony Campbell wrote: iface eth1 inet dhcp address 192.168.0.22 auto eth0 iface eth0 inet dhcp address 192.168.0.20 I could be missing something but why do you specify DHCP as well as a static address? I thought these options were mutually exclusive and I'm surprised you don't get some kind of error message. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
minicom display problems
Is there any way to get minicom to properly display the color graphics of an ncurses app like iptraf? I am using the multi GNOME terminal as my X terminal, but xterm seems to have the same problems. So far I've tried: -setting the minicom terminal type to ANSI -setting the remote shell TERM type to xterm (it starts up as VT100) -turning the minicom color on using the -c on option Another, possibly unrelated issue, is that iptraf and other programs like top or man which output to the entire terminal screen, only use a small portion of the available X terminal screen area (about 23 lines of text) regardless of how large I make the X terminal window. Man even puts pages of text on different areas of the X terminal and is nearly useless. My current workaround is to use a tiny X terminal window with these programs, but ncurses graphics is still a problem. In contrast, programs like ls which produce individual lines of output, seem to adjust and use available X terminal space properly. I suspect that a solution to either problem will solve both. I am not averse to RTFMing but am not sure which FM to read at this point. Thanks for any help. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: announcing the beginning of security support for testing
Could a list of md5sums be provided for this archive, like the file /debian/indices/md5sums.gz in the main (debian) archive? With the help of a simple script, this file allows me to check the package integrity in my mirror of the main debian archive. I am hoping that this method can be used for other archives as well, as an alternative to the currently recommended checking method. The problem with the secure-testing checking procedure (which is also used by security.debian.org and marillat archives) is that it requires apt 0.6.* Unfortunately, the version of apt in debian testing is only 0.5.28.6 and in any case it will be a long time before all of my systems run apt version 0.6 or higher. In addition, the recommended checking procedure only checks packages during installation, if I understand it correctly -- it cannot check the inegrity of an entire mirror archive. For my purposes, I need to check the integrity of all packages in my local archives, before I attempt to install them. Compounding this problem is the fact that rsync to the (primary) secure-testing archive is disallowed using the -c (checksumming) option, understandably so. rsync with checksumming has been my workaround with my local debian-security archive. *See http://www.debian.org/doc/manuals/securing-debian-howto/ch7#s-deb-pack-sign which is referenced by the Debian security FAQ. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: OT iptables question
Glenn English wrote: I'm updating a RH ipchains packet filter script from the dim past to iptables on Debian stable. I noticed that when I specified the network the host is on (by IP/mask), the iptables listing called it localnet. So I tried using localnet in the rule, and iptables seems to take it, and the chain seems to work. But I can't find any documentation about that keyword in man, in Rusty's HTML dox, or with google (lots of talk about it, but no dox). Is localnet a legit iptables network specification or an undocumented feature? What does it actually do (should I hang a CIDR mask on the end, or would that be redundant)? If the host responds to several IPs, does localnet cover then all? Or just eth0? How about eth0:1? It would be very handy because this script is to set filtering on all my DMZ and LAN hosts (by switching on their hostnames and IPs). I know I could just try it and see if it works, but this is to be the packet filter on the DMZ, and I'd like to do it as rigorously as I can. TIA... On my sarge system localnet seems to be defined in /etc/networks. Try man networks You might also try changing the network name there and see what happens. This raises another question for me, I don't understand why I cannot find the this file using dlocate or apt-file, or even using the package search tool on debian.org. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
OT: help with security question
Unfortunately there is a windows box on my network which is running Norton Firewall, with logs, documentation and a user interface that seem ambigious, simplistic and confusing, as if written in some kind of technical pigeon language. I was surprised when it reported an incoming ICMP packet by raising a dialog window asking if I wanted to make a firewall rule for the source of the packet, which I later thought looks like a router at some mid-level ISP (ae-1-51.bbr1.Chicago1.Level3.net [4.68.101.1]). Norton's recommendation was to enable incoming ICMP for that host, which I did. I then checked the firewall rules and found a specific rule just for that host. Now I'm trying to understand what it all means. I'm not very familiar with the details of IP and ICMP, much less windoze boxes and all their quirks. For some unknown reason Norton Firwall wants to distinguish between incoming/outgoing ICMP v bidirectional ICMP. There doesn't seem to be any way to log and inspect the actual packet headers or contents, so I don't know what kind of ICMP the firewall detected, nor even the definition of incoming ICMP. I don't think any application or service tried to ping the internet router, because when I manually try to ping another internet host from the windows host, Norton tries to make rule for *outgoing* ICMP, not incoming ICMP. This reinforces my slightly paranoid initial impression that some internet host is trying to ping or otherwise access the windows host behind my firewall, and a possible hardware firewall misconfiguration that allows it to happen. I double checked my hardware firewall, which is a Sarge box running guarddog and my ppp link to the local ISP. It's set to allow ICMP service from the internet to local hosts, but not in the other direction. I've always assumed that local hosts cannot automatically serve protocols, including ICMP, through the firewall without special configuration. Thanks for any help, suggestions or insights. The Debian firewall has all the tools I might need to track this down, e.g. psad, snort, bastille, etc. But I'm still learning how to use them, and I appreciate any help. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Kernel panics when booting off SCSI
Joel Barker wrote: I have been using two hard drives, an old IDE mounted at / and a brand new SCSI mounted on /home. A few days ago the IDE drive died. Fortunately, I had just copied all the data over to the SCSI drive (/dev/sda1). But when I try to boot off the SCSI drive, I get the following messags: Freeing unused kernel memory: 152k freed initrd-tools: 0.1.81.1 NET: Registered protocol family 1 Uniform Multi-Platform E-IDE driver Revision: 7.00alpha2 ide: Assuming 33MHz system bus speed for PIO modes; override with idebus=xx NFORCE2: IDE controller at PCI slot :00:09.0 NFORCE2: chipset revision 162 NFORCE2: not 100% native mode: will probe irqs later NFORCE2: :00:09.0 (rev a2) UDMA133 controller ide0: BM-DMA at 0xf000-0xf007, BIOS settings: hda:DMA, hdb:DMA ide1: BM-DMA at 0xf008-0xf00f, BIOS settings: hdc:DMA, hdd:DMA hdc: OPTORITECD-RW CW5201, ATAPI CD/DVD-ROM drive ide1 at 0x170-0x177,0x376 on irq 15 pivot_root: No such file or directory /sbin/init: 432: cannot open dev/console: No such file It's a udev quirl. You need the device file console in /dev. I usually copy the device manually after cloning a disk. Kernel panic - not syncing: Attempted to kill init! Here are the commands I'm using in GRUB to boot: root (hd0,0) --- this is the scsi disk kernel /vmlinuz root=/dev/sda1 ro initrd /initrd.img boot I tried building a new initrd (mkinitrd) with 'scsi_mod', 'sd_mod' and 'sr_mod' added to /etc/mkinitrd/modules, but no improvement. I then built an initrd with nearly all the modules from `lsmod` (except a few I was sure weren't needed, like soundcore) added to /etc/mkinitrd/modules. I got a few more messages when I tried to boot (USB Mass Storage support registered), but had the exact same error. I am using Debian 3.1. /sbin/init exists on both /dev/sda1 and the initrd. So does /dev/console. I DID specify the correct kernel version when building the initrd. If possible, I would like a solution short of recompiling the kernel. Thanks. ~joel -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Printer localhost:631 -- Don't Give Up Yet
Charlie wrote: My /etc/hostname file is:- ariestao my netaddress for this machine This is incorrect. It should just be your hostname, with no IP address. (Could this be the cause of your problem?) -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: resolved Re: sarge package installation segfaults
Marty wrote: Marty wrote: I reported Bug#301912, consisting of repeated perl warnings and segfaults during package installation. It turns out to likely be hardware data corruption caused by a new memory module or changed timing. I caught it by the venerable burnit kernel compile loop script, aided by distcc to keep cycle times reasonably low. Overnight memtest and debsum loops did not cause a failure, but dpkg was failing about every third run. Sorry about the false alarm. Marty If possible, try different memory modules. For some reason, dpkg seems particularly sensitive to memory errors. I once got repeated segfaults and perl warnings due to memory errors, when almost no other test (including memtest) was able to detect the failures. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
