From: Digestifier To: Linux-Admin@senator-bedfellow.mit.edu Reply-To: Linux-Admin@senator-bedfellow.mit.edu Date: Tue, 30 Aug 94 19:13:26 EDT Subject: Linux-Admin Digest #2 Linux-Admin Digest #2, Volume #2 Tue, 30 Aug 94 19:13:26 EDT Contents: Re: process analysing tool now available (Thomas Breitfeld) Re: How to FTP recursive directories? (Yufan Hu) Linux hangs up and no trace why (Jozsef Kadlecsik) HELP! Intermittent connect & ARPing probs (Greg Corteville) Re: How do I get my 525MB SCSI streamer to backup 300MB? (Kai Dupke) Re: How to kill the unkillable ? (H. Peter Anvin) Re: Slackware 2.0 Problem with recompileing the kernel (Greg Corteville) Re: How to mount CD__ (Mitsumi) (Klaus Ebner) Re: How to kill the unkillable ? (James C. Bowen) Re: named keeps crashing (Barry Yip Kam-wa) Re: color_xterm doesn't read .bashrc (Jozsef Kadlecsik) SOCK_PACKET: Why not reading outgoing packets ? (Morten Jammer) Security hole in ntalkd Re: Need suggestions on Linux security (Colin Fletcher) **** HOW TO FIX YOUR CDROM!! **** (Serge Solski u) ---------------------------------------------------------------------------- From: breitf@infotech.tu-chemnitz.de (Thomas Breitfeld) Subject: Re: process analysing tool now available Date: 30 Aug 94 09:06:30 GMT stephenb@scribendum.win-uk.net (Stephen Benson) writes: >Hi. I downloaded xps and it built with no problems, but when I run it, >there's nothing in the windows (sometimes I see the icons) and I get these >error messages: >[extract error messages] >:: >:: >:: [...] >-- >+ stephen benson + + + + + linux 1.0.9 + + xfree86 2.1.1 + >+ stephenb@scribendum.win-uk.net + + + + + + + + + + + I guess you have an old psdatabase or the Symbol table doesn't correspond to the current kernel. Try to run Xps -U first to create a new psdatabase. If this doesn't work make sure the unstripped kernel image is the same as the kernel image used for booting. The unstripped kernel image will be normaly found in /usr/src/linux/tools/zSystem. The easiest way to create a new one is to recompile the kernel. Hope this will work Greetings Thomas -- _____________________________________________________________________ Thomas Breitfeld | Phone: (049) 371 531 3241 | Email: breitf@infotech.tu-chemnitz.de | Fax: (049) 371 531 3242 | University of Technology Chemnitz ----------------------------- ------------------------------ From: yufan@iscm.ulst.ac.uk (Yufan Hu) Subject: Re: How to FTP recursive directories? Date: 30 Aug 1994 11:17:45 GMT Jeff Arnholt (arnholt@mayo.edu) wrote: : Question: how can I use FTP (or some other solution) : to copy all of the contents of a directory AND everything : below that directory? IE, I want everything in : /pub/Linux/distributions/Slackware/slakware and below. : Currently, I can only grab one disk at a time. : Is there a giant tar'd file somewhere of the disk : sets? That would also solve my problem. Some FTPD has extra flavour to enable you to tar the directory first and then send the whole directory as a tar'ed file. To do this you need just put an .tar at the end of the directory name. You can also copy files compressed by appending .Z, or .gz, etc. to the file names. I think src.doc.ic.ac.uk support this. Hope this is of help. Yufan. ------------------------------ From: kadlec@cern.school.kfki.hu (Jozsef Kadlecsik) Subject: Linux hangs up and no trace why Date: 30 Aug 1994 15:16:09 GMT Hi, We have a 486 DX2/66 machine with the following devices and settings (Slackware 2.0): Adaptec AHA1542CF I/O=330 IRQ=12 DMA=5 SCSI ID=7 Quantum LP120S SCSI ID=0 Quantum LPS540S SCSI ID=1 Sony CD-ROM CDU-561 SCSI ID=6 SMC 8216 Ultra I/O=280 IRQ=15 RAM=C800 Trident TVGA8900D Everything works: we can read-write any partition, we can mount the CD-ROM or play an audio CD and the network is running. But sometimes the system hangs up, especially when NOBODY uses the system. Then we can type in commands and hit Enter, but nothing executed. We can switch between the virtual consoles, that's all. We changed the motherboard, the Adaptec card, the disk, the SCSI cable but nothing helped. There is only one "clue", but it doesn't help us: we get this "failed" message at every boot: Aug 30 12:23:51 cern kernel: Calibrating delay loop.. ok - 33.22 BogoMips Aug 30 12:23:51 cern kernel: failed Has anybody ever encountered such a problem? Thanks ANY help or hint, Jozsef Kadlecsik ------------------------------ From: gcortevi@nyx10.cs.du.edu (Greg Corteville) Crossposted-To: comp.os.linux.help Subject: HELP! Intermittent connect & ARPing probs Date: 30 Aug 1994 12:07:16 -0600 My Linux box is going seriously weird. The machine is refusing all inbound network connections for periods of up to five hours or more. Sometimes it works fine. Other times it will boot a user off for no apparent reason. The system is constantly responding to pings but refusing connections to all ports. The syslog files are filled with lines such as: Aug 29 22:34:47 trident kernel: <6>ARP: ARPing my own IP address 35.8.156.105 ! The system remains running locally. I am pretty sure it is running kernel version 1.1.0. The system runs on a 386SX/16 with 4 megs of RAM and about 12 megs swap. The distribution was originally Slackware 1.1.1 (and still is pretty much the same). The system also runs an NE2000 network card. These problems just started happening a few days ago. We made no changes to the system since that time or before that time. It seemed as though it just happened. Can somebody tell me how to rectify this problem? If not, please tell me what causes the system to "ARP" its IP address. Also, if you believe this problem lies outside of the system (a network problem) let me hear some of your reasoning [although I doubt this since it responds to pings when it refuses connections]. Any help would be greatly appreciated as my net access is far more limited without this machine. -- Greg Corteville | "We are the knights who say Nee!" gcortevi@nyx10.cs.du.edu | - Monty Python's QFTHG gcortevi@trident.lbs.msu.edu | ---[PGP Key available via finger]------------------------------------ ------------------------------ Crossposted-To: comp.os.linux.misc From: dupke@koma.han.de (Kai Dupke) Subject: Re: How do I get my 525MB SCSI streamer to backup 300MB? Date: Sun, 28 Aug 1994 15:38:14 GMT Tiger Zhao (tiger@netcom.com) wrote: : essenber@dutiws.twi.tudelft.nl (Eelco H. Essenberg) writes: : >Hi all! : >Apologies if this is a FAQ, but I'm having trouble making backups with : >my new Conner (Archive) Viper 2525 tape streamer. I'm going on the : >following assumptions: : > * the Archive Viper 2525 can backup 525MB on 1 tape, right? : > * a DC6525 tape can hold 525MB (or thereabouts), right? That's right. You could store up to 525MB on a DC6525. As i uses a QIC-Streamer whith 6250 or 6525 too, i tested this some days (month) ago. But: If you think, that you back up 200 (or so) Megs to tape, because you use 'df' before backing up, this is not right. Within my system i'm back up 23Megs more for the /proc/kore-file, wich isn't shown in the df-command. You could test ist, if you use the dd-command (and you got/have the time to try it). Gruss kai, dupke@koma.han.de ------------------------------ From: hpa@ahab.eecs.nwu.edu (H. Peter Anvin) Subject: Re: How to kill the unkillable ? Reply-To: hpa@nwu.edu (H. Peter Anvin) Date: Tue, 30 Aug 1994 18:34:59 GMT Followup to: By author: jbowen@thufir.cs.umn.edu (James C. Bowen) In newsgroup: comp.os.linux.admin > > >_every_ signal has been tried. but it just refuses to die ! can anyone shed > >any light on how to get rid of processes in this state ! > > I'm no expert, but I believe this is an inherent property of unix -- a > process in a wait state waiting on an I/O request can not be killed by > any means. > I have wondered about if that is something that should be fixed in Linux... I have had thoughts about a SIGNUKE signal that would have the same effects as a SIGKILL (untrappable, terminate execution) but which would also immediately wake up the process so it could receive it. I rejected another idea, having a generic wakeup_and_send_signal syscall, since if the process trapped the signal it might get confused about the non-standard behaviour. /hpa -- INTERNET: hpa@nwu.edu --- Allah'u'abha --- IBM MAIL: I0050052 at IBMMAIL HAM RADIO: N9ITP or SM4TKN FIDONET: 1:115/511 or 1:115/512 STORMNET: 181:294/1 or 181:294/101 Most inappropriatly named startup command, winner: Microsoft Windows ------------------------------ From: gcortevi@nyx10.cs.du.edu (Greg Corteville) Subject: Re: Slackware 2.0 Problem with recompileing the kernel Date: 30 Aug 1994 12:44:44 -0600 kiko@chopin.udel.edu (Brian Curti Harvell) writes: >Hi I just recently optained Slackware 2.0 and I am now trying to reconfigure >the kernel to my likeings but I keep getting errors. They are something like >gcc: Internal compiler error: program cc1 got fatal error 6 I too have the exact same problem! Please post any fixes or replies that could shed some light on this for all of us. (I'm still stuck on kernel 1.1.0) -- Greg Corteville | "We are the knights who say Nee!" gcortevi@nyx10.cs.du.edu | - Monty Python's QFTHG gcortevi@trident.lbs.msu.edu | ---[PGP Key available via finger]------------------------------------ ------------------------------ Date: Mon, 29 Aug 1994 08:20:25 +0200 From: Klaus_Ebner@zaphod.fido.de (Klaus Ebner) Subject: Re: How to mount CD__ (Mitsumi) Hoi, just in addition > later to mount do: > mount -r -t iso9660 /dev/mcd /cdrom > the type : -t iso9660 is the standar on CD_ROM format... > -r only read (otherwise linux refuse to mount) On my System I need to use the /dev/mcd0 device, but I don't need to use the option -r. It mounts just fine... -- Gruss Klaus Fido : 2:2453/30.68 Internet : klaus@zaphod.rhein.de ------------------------------ From: jbowen@thufir.cs.umn.edu (James C. Bowen) Subject: Re: How to kill the unkillable ? Date: Tue, 30 Aug 1994 15:07:19 GMT In <33si7r$kjk@styx.uwa.edu.au> cjcason@yarrow.wt.uwa.edu.au (Christopher Cason) writes: >Occasionally, I have a program that I absolutely _cannot_ kill, short of >a shutdown. These seem to occur at times when the program is attempting >to access a part of a disk that has been corrupted (I have had filesystem >problems from time to time.) >_every_ signal has been tried. but it just refuses to die ! can anyone shed >any light on how to get rid of processes in this state ! I'm no expert, but I believe this is an inherent property of unix -- a process in a wait state waiting on an I/O request can not be killed by any means. A bad disk sounds like a much more serious concern to me:-) -- *** Jim Bowen ** We have met the enemy, *** *** CS Systems Staff ** and *** *** jbowen@cs.umn.edu ** they is us. - Pogo *** ------------------------------ From: barry@ns.japan-online.or.jp (Barry Yip Kam-wa) Subject: Re: named keeps crashing Date: 30 Aug 1994 17:32:26 +0900 In article , Jason Fiset wrote: >I'm having a problem with named. Whenever, someone runs the command: > >host -la > >named always crashes! I've tried almost everything I can think of to stop >this from happening but it doesn't seem help. I'm running Slackware 2.0.0 >on a 386-40 with 8M of ram. I also upgraded to a new named binary that a >friend of mine gave me. Thanks. > This works fine on my linux box running BIND-4.8.3-BETA9. The linux box is running 1.1.38. I remembered that the older bind did some times died mysteriously after running for a while though not sure whether this is related to what you experience. This one works fine for about a month so far. Barry -- barry@japan-online.or.jp ------------------------------ From: kadlec@cern.school.kfki.hu (Jozsef Kadlecsik) Subject: Re: color_xterm doesn't read .bashrc Date: 30 Aug 1994 11:24:37 GMT Thanks to all, who answered my question. The setting of *LoginShell: true in .../app-defaults/XTerm solved the problem. Best regards, Jozsef Kadlecsik ------------------------------ Date: 30 Aug 1994 16:25:00 +0100 From: morten@gurke.allcon.com (Morten Jammer) Crossposted-To: comp.os.linux.development,comp.os.linux.help,comp.os.linux.misc Subject: SOCK_PACKET: Why not reading outgoing packets ? Hi ! Im developing a networkaccountingprogram and my question is: Why can the socket typ SOCK_PACKET only read outgoing packets when the interface is in promiscious mode ? - Morten : Morten Jammer : Ich sei, gewährt mir die Bitte : Norderstrasse 97: in eurem Netzwerk : 24939 Flensburg : der Dritte ## CrossPoint v3.02 ## ------------------------------ Subject: Security hole in ntalkd From: ptmcinti@indyunix.iupui.edu () Date: 30 Aug 94 09:36:49 -0500 Greetings: Last night, someone rebooted my machine using the ntalkd program. Here's a copy of my system logs prior to the reboot: /var/adm/messages Aug 29 18:48:54 MindWay pppd[1925]: local IP address 134.68.249.62 Aug 29 18:48:54 MindWay pppd[1925]: remote IP address 134.68.249.3 Aug 29 19:18:48 MindWay in.ntalkd[1986]: connect from 142.51.20.1 Aug 29 19:18:48 MindWay in.ntalkd[1987]: connect from 142.51.20.1 Aug 29 19:20:51 MindWay in.ntalkd[1990]: connect from ramsey.cs.laurentian.ca Aug 29 19:22:54 MindWay in.ntalkd[1992]: connect from ramsey.cs.laurentian.ca Aug 29 19:24:57 MindWay in.ntalkd[1994]: connect from ramsey.cs.laurentian.ca Aug 29 19:26:59 MindWay in.ntalkd[1997]: connect from ramsey.cs.laurentian.ca Aug 29 19:29:02 MindWay in.ntalkd[1999]: connect from ramsey.cs.laurentian.ca Aug 29 19:30:40 MindWay in.ntalkd[2002]: connect from starbase.NeoSoft.COM Aug 29 19:31:05 MindWay in.ntalkd[2003]: connect from ramsey.cs.laurentian.ca Aug 29 19:33:07 MindWay in.ntalkd[2008]: connect from ramsey.cs.laurentian.ca Aug 29 19:35:10 MindWay in.ntalkd[2013]: connect from starbase.NeoSoft.COM Aug 29 23:00:25 MindWay syslogd: restart /var/adm/syslog Aug 29 19:35:10 MindWay talkd[2013]: Bad address, family 0 Aug 29 19:35:10 MindWay talkd[2013]: recv: Connection refused Aug 29 19:35:10 MindWay talkd[2013]: Bad address, family 0 Aug 29 19:35:10 MindWay talkd[2013]: recv: Connection refused Aug 29 19:35:10 MindWay talkd[2013]: Bad address, family 0 Aug 29 19:35:10 MindWay talkd[2013]: recv: Connection refused Aug 29 19:35:10 MindWay talkd[2013]: Bad address, family 0 Aug 29 19:35:10 MindWay talkd[2013]: recv: Connection refused .. I really don't understand how they could have done this. Someone mentioned that since ntalkd speaks UDP, they could have used that protocol to (somehow!) cause a reboot ... I'd appreciate any insight you might have. At this point, i've disabled the talk daemon completely out of fear of another reboot. Thanks. ===================================================== Paul T. McIntire ptmcinti@indyunix.iupui.edu Indiana UniXersity Purdue UniXersity at Indianapolis Office of Integrated Technologies Web: http://indyunix.iupui.edu/~ptmcinti ===================================================== ------------------------------ From: cfletche@PROBLEM_WITH_INEWS_GATEWAY_FILE (Colin Fletcher) Subject: Re: Need suggestions on Linux security Date: 30 Aug 1994 22:11:27 GMT In previous life Thilo Wunderlich wrote: > smithgr@cs.colorado.edu (Gregory P. Smith) writes: > [it's too easy to change boot password...] First thing is first. To change the ami bios setup this means that you have to have access to dos. I used to run a linux box w/ 80+ users and after we installed linux and put a password on the ami-bios. There was no possible way to boot from a, short of taking the cover off the computer ... After you have installed linux and put a password on the ami-bios, how can anyone boot dos without the password and taking the cover off? my $.02 Colin Fletcher ------------------------------ Crossposted-To: comp.os.linux.help From: sols7520@mach1.wlu.ca (Serge Solski u) Subject: **** HOW TO FIX YOUR CDROM!! **** Date: Mon, 29 Aug 1994 14:54:18 GMT I've finally found out why most of us have CDROM's that work on installation, but don't work when running Linux regularily. The problem is that the wrong driver is being used. You have to use the "standard" driver rather than the "enhanced" driver. This may not work for all of you, but it worked for me. 1) If you look in /root/linux/drivers/block (this may be in another directory for you), you'll see "cdu31a.c", "cdu31a.c-enhanced", and "cdu31a.c-standard." I suspect that those of you not using the CDU31A drivers will see something similiar for whatever CDROM you're using. If you type "ls-l cdu31a.c", you'll see that "cdu31a.c" is linked to "cdu31a.c-enhanced." If it's linked to "cdu31a.c-standard" then forget about this fix -- you have another problem. Move "cdu31a.c" to cdu31a.c-backup (just in case), and then copy "cdu31a.c-standard" to "cdu31a.c" Instead of copying, you could just change the link, but I forget the exact syntax of the command. Basically, all you have to do is make "cdu31a.c-standard" the CDROM driver, which is called "cdu31a.c". 2) Now you have to recompile the kernel. Do whatever you need to do on your system to recompile the kernel, and don't forget to tell Lilo what you did. I just did a "make clean", then "make dep", then "make zImage" and copied it to "/" by hand (backup up the old kernel, of course.) Then I used "liloconfig" to reinstall Linux, and then rebooted. I skimped on this description, so do whatever need to be done on YOUR system to recompile the kernel. I'm sure I broke several rules on recompiling kernels, but I don't care (nya, nya.) 3) Upon rebooting, I found that I still had the error appearing on the startup screens. I did a manual mount "mount -t iso9660 -r /dev/cdu31a /cdrom" and it worked fine. Yahoo! I then took the line about the CDROM out of /etc/fstab, and put a line in /etc/rc.d/rc.local to mount the CDROM, "mount -t iso9660 -r /dev/cdu31a /cdrom". This issues the mount command when you boot-up. You'll have to "umount" and "mount" the CDROM whenever you change CD's -- I believe this is normal. 4) Party! Rejoice! This annoying problem is gone! This proves that if you really want to get anything accomplished, you have to wait until 3:00 am until you're truly inspired. Let me know how this turns out for you! -Mark -- "Key chuckles. 'If Skinny Puppy, in terms of the movie _Alien_, is a chest-burster, then Doubting Thomas is more of a face-hugger,' he informs, as if that were an explanation." -Keyboard, Jan '92 ------------------------------ ** FOR YOUR REFERENCE ** The service address, to which questions about the list itself and requests to be added to or deleted from it should be directed, is: Internet: Linux-Admin-Request@NEWS-DIGESTS.MIT.EDU You can send mail to the entire list (and comp.os.linux.admin) via: Internet: Linux-Admin@NEWS-DIGESTS.MIT.EDU Linux may be obtained via one of these FTP sites: nic.funet.fi pub/OS/Linux tsx-11.mit.edu pub/linux sunsite.unc.edu pub/Linux End of Linux-Admin Digest ******************************