Re: Restricting size of incoming mail
Hi, At 10:43 AM 4/15/97 -0400, Paul Wade wrote: This is because they are using a remote email client (netscape, etc.)? If so that is a big part of the problem. Right, we use a POP-3 client like Eudora or Netscape Mail to retrieve mail. Is this being caused because of spam and junk email? No, the problem is that users do not follow the policies of not sending e-mail attachments to users that retrieve mail via a dial-in connection. A temporary remedy is to use a shell and pine to delete the basura from the mailbox and then retreive the remaining messages with the client program. I have been thinking about using procmail. Quotas will possibly prevent important mail from being viewed if it arrives after mail with large attachments. Uhhmmm... you are absoluteli right, I had't thought about this. This leads me to think that a want to limit the size of incoming mail (in a user per user basis), not the size of the mail box. Some users can only send and receive files as email attachments. They are mulas and will not learn ftp. I want them to use FTP whenever they want to send a huge document. FTP should be more efficient. At least it doesn't increase the size of the file by applying a codification algorithm (like UUCode or MIME.) Thanks for the help. E.- -- Eloy A. Paris Information Technology Department Rockwell Automation de Venezuela Telephone: +58-2-9432311 Fax: +58-2-9431645 Cel.: +58-16-234700 Where does this path lead? said Alice Depends on where you want to go. Said the cat (Alice in Wonderland, by Lewis Carroll.)
Re: Oracle on Linux - Install Scripts
I have not know how to do what you ask but you might find useful an article that appeared in the Linux Means Bussiness section of The Linux Journal, issue November 1996. The article is called Running Progress on Linux. They also use the iBCS package. Good luck. E.- Subject: Oracle on Linux - Install scripts Has anyone been successfull installing Oracle7.3 Workgroup server for UnixWare or Oracle Webserver 2.1 on Linux (RedHat, Debian or Caldera) and care to share your experience/install scripts? I know that it has been done on SCO UNIX and there is a document that outlines the process. (I'm looking to do it on UnixWare 2.1) I know that I need iBCS package but has anyone written a HOWTO or step by step procedure to do this? I looked at the Oracle7.3 install scripts and it looks like they will have to be rewritten entirely. Any pointers will be appreciated. Please email me directly or CC: when you post. Exact version is Oracle7 Workgroup Server 7.3.2.2.0 for UnixWare Roger [EMAIL PROTECTED] -- Eloy A. Paris Information Technology Department Rockwell Automation de Venezuela Telephone: +58-2-9432311 Fax: +58-2-9430323
Is Bo frozen???
Hi, is Bo frozen already? When Rex got frozen I received a message from Debian-Announce saying so. I also read in www.debian.org that Rex was frozen. I think I have heard that Bo is frozen but I have not checked the FTP sites yet. Bye, Eloy.- -- Eloy A. Paris Information Technology Department Rockwell Automation de Venezuela Telephone: +58-2-9432311 Fax: +58-2-9430323
Re: bad FS errors after (almost) every boot
Hi, when I installed Debian 1.1 about 6 months ago I had filesystem corruption problems during and after installation. My box had a NCR53c810 on-board SCSI adapter so I had to use one of the special boot disks. After fighting with the installation for a couple of days I realized that the boot disks for the NCR53c810 had a kernel compiled with some parameters that caused problems in my hardware setup. I solved the problem by creating a boot disk (using the bootdisk package) with a kernel customized for my hardware. Now, my system has been up and running for half a year with no problems at all. I also upgraded succesfully to 1.2. Why don't you try re-installing with a custom boot disk that has a custom kernel just for you hardware? Regards, E.- I am gradually getting desperate. Almost every time I turn on my linux box e2fsck finds massive errors. Often times somewhere under /usr/lib/terminfo. I have no idea what I could be doing wrong. I always shutdown my system with a proper shutdown or halt command. Here is a short description of my setup: I am using 2 x 2GB Quantum SCSI HDs. On the first there is a Win95, a Win3.11 partition and 2 FAT data partitions. At the very end of this HD there is the OS/2 Boot Manager. On the second HD, there is another Win 3.11 partition (primary partition) and the root, var and swap partitions for Linux. /home is a link to /var/home, /usr is on the root partition. Hm, did I forget anything important? Oh, yes, I've got the Debian 1.2.4 distribution installed. At first, I thought that Win95 is the reason for this behavior, but in the meantime I found that I get these errors even if I didn't boot any other OS in between. So there must be something else wrong. Two days ago it became even worse: I got fs errors while the system was running. All kinds of bad message appeared in my console. I wanted to save them so I could tell you the exact words here, but all I got was Cannot save, fatal fs errors or something similar. :-( Please help, I have no idea how to fix this! Thanks so much in advance, Andy. Andy Spiegl, PhD Student, Technical University, Muenchen, Germany E-Mail: [EMAIL PROTECTED] URL:http://www.appl-math.tu-muenchen.de/~spiegl PGP fingerprint: B8 48 24 7B DB 96 6F 1C D9 6D 8E 6C DB C2 E7 E9 o _ _ _ - __o __o /\_ _ \\o (_)\__/o (_) --- _`\,__`\,__(_) (_)/_\_| \ _|/' \/ -- (_)/ (_) (_)/ (_) (_)(_) (_)(_)' _\o_ ~~~ -- Eloy A. Paris Information Technology Department Rockwell Automation de Venezuela Telephone: +58-2-9432311 Fax: +58-2-9430323
Re: biff(1)
Hi, Has anyone successfully managed to get biff to work in debian? i know there's some interaction with comsat, but i never worked out where comsat got messages from (something to do with procmail? shudder). FYI, using latest unstable sendmail as MTA. I love biff. I do not like to press Enter in bash to get a new mail notification. I like biff because it is asynchronous. I also have sendmail as my MTA. By default, sendmail uses deliver as its local delivery program. deliver does not notify comsat so biff does not work if you use deliver. You can use procmail (that's what I am using) and it'll work pretty well. You have to make a minor change to the sendmail.cf file, or better to the sendmail.mc file. Here is mine: -- # # This file is used to configure sendmail for use with Debian systems. # divert(0) VERSIONID(`@(#)sendmail.mc 8.7 (Linux) 3/5/96') OSTYPE(debian)dnl FEATURE(masquerade_envelope)dnl FEATURE(always_add_domain)dnl FEATURE(use_cw_file)dnl FEATURE(use_ct_file)dnl FEATURE(nouucp)dnl FEATURE(nodns)dnl FEATURE(local_procmail, /usr/bin/procmail) define(`LOCAL_MAILER_ARGS', `procmail -Y -d $u') MAILER(local)dnl MAILER(smtp)dnl Cwzeus.ven.ra.rockwell.com MASQUERADE_AS(ven.ra.rockwell.com)dnl ## Custom configurations below (will be preserved) define(`confMAX_MESSAGE_SIZE', `100') define(`confTRY_NULL_MX_LIST', `True') define(`confCOPY_ERRORS_TO', `postmaster') define(`SMART_HOST', smtp:mailrelay.mke.ab.com) LOCAL_NET_CONFIG R$* @ $* .$m. $*$#smtp $@ $2.$m. $: $1 @ $2.$m. $3 -- Don't pay attention to the custom configuration stuff. It's just for my site. The key lines are: FEATURE(local_procmail, /usr/bin/procmail) define(`LOCAL_MAILER_ARGS', `procmail -Y -d $u') If I remember right, the first line is needed because the default definition has procmail in /usr/local/bin/ and Debian puts it in /usr/bin/. After changing your /etc/mail/sendmail.mc file run sendmailconfig and tell it to use the existing sendmail.mc file, do not generate a new one. After this, you should be able to use biff as in other *unixes. Let me know if you have problems. Good luck. E.- --- Eloy A. Paris Information Technology Department Rockwell Automation de Venezuela Telephone: +58-2-9432311 Fax: +58-2-9430323
Backup procedure
Hi, backups!!! Backups have always been a mistery to me, a pain I should say. However, I realize the importance of backups. I am looking for a backup procedure to implement here at my company. We have a couple of Novell servers, several Linux boxes (two of them are production servers very important to our Information Technology structure) and an H-P 9000 server running HP-UX. Each server has a dedictaed tape unit attached (DDS, aka DAT.) Right now we are making full backups from Monday to Friday so we can go back only five days. We are not taking tapes off-site either. I want to know about rotation methods (I've heard about Granfather-Father-Son, 10 tape, Hanoi Towers but the information is not clear), when to take tapes off-site, Debian backup utilities, etc. Any URL will be appreciated. I am search Yahoo right now but without luck so far. Thanks in advance, Eloy.- -- Eloy A. Paris Information Technology Department Rockwell Automation de Venezuela Telephone: +58-2-9432311 Fax: +58-2-9431645 Cel.: +58-16-234700 Where does this path lead? said Alice Depends on where you want to go. Said the cat (Alice in Wonderland, by Lewis Carroll.)
Finding files not present (challenge to your intelligence)
Hi, I was given a text file containing one file name (no full path name) per line. My task consists of searching the entire filesystem and generate a list of the files that are NOT present. At first I thought this was a task easyly solved with a couple of awk's, sed's, and find's. However, it's been four days now and I still haven't finished this stupid thing. Right now I am working with Perl but haven't finished. I think I am close but just wanted to know if someone here at debian-user has a simple solution, the shortest one. The problem with using find is that this command does not return an error code when it does not find a file. Thanks in advance. E.- -- Eloy A. Paris Information Technology Department Rockwell Automation de Venezuela Telephone: +58-2-9432311 Fax: +58-2-9430323
Finding files not present (challenge to your intelligence)
Hi, I was given a text file containing one file name (no full path name) per line. My task consists of searching the entire filesystem and generate a list of the files that are NOT present. At first I thought this was a task easyly solved with a couple of awk's, sed's, and find's. However, it's been four days now and I still haven't finished this stupid thing. Right now I am working with Perl but haven't finished. I think I am close but just wanted to know if someone here at debian-user has a simple solution, the shortest one. The problem with using find is that this command does not return an error code when it does not find a file. Thanks in advance. E.- -- Eloy A. Paris Information Technology Department Rockwell Automation de Venezuela Telephone: +58-2-9432311 Fax: +58-2-9430323
Re: Finding files not present (challenge to your intelligence)
Hi Martin, How about using find and generate a list of ALL files, then grep `indivual_lines_from_your_list` list_of_ALL_files and checking grep's return value? Well, what I am doing right now is generating a list of the files that are present (using find and redirecting its output to a file). Later, I search this file for files not in my original list, so I know which files are missing. I am using Perl to do this. Regards, E.- -- Eloy A. Paris Information Technology Department Rockwell Automation de Venezuela Telephone: +58-2-9432311 Fax: +58-2-9430323
Re: Finding files not present (challenge to your intelligence)
Hi,
thank you very much for your solution; it is very sharp. I was doing
things the hard way. I like the shell solution very much.
Regards,
Eloy.-
On Mar 26, Eloy A. Paris wrote
: Hi,
:
: I was given a text file containing one file name (no full path name)
: per line. My task consists of searching the entire filesystem and
: generate a list of the files that are NOT present.
1. The perl solution:
---
#!/usr/bin/perl
# * probably overkill, if your file list
# is rather short.
# * probably storing the existing files in a @Array
# and grepping is faster
# * probably a shell line with sort / uniq / comm would
# do the same job
# generate a list of existing files
open(IN, find . -type f -printf \%f\n\|) or die;
while(IN) { chomp; $Files{$_} = 1; }
# read in the search list and see, the file exists
open(IN, searchlist) or die;
while (IN) { chomp; $Files{$_} or print Missing: $_\n; }
2. The Shell solution
find -type f -printf %f\n | sort files.exist
sort searchlist files.search
comm -1 -3 files.exist files.search
Heiko Schlittermann
-
Heiko Schlittermann / Internet Unix-Support
Kamenzer Str. 52 D-01099 Dresden
Voice: +49-172-7909055 Mail: [EMAIL PROTECTED]
--
Eloy A. Paris
Information Technology Department
Rockwell Automation de Venezuela
Telephone: +58-2-9432311 Fax: +58-2-9430323
Re: POP3 Server
What is the lastest version of the POP3 Mail Server software. I seem to be getting reports from users who say they have 'Remove Mail from the Server' selected so that when a mail download completes to their mail reader it should remove it, but it does not. Almost every single problem report comes from users running Netscape Mail or Microsoft Mail. The problem seems to point to the mail reader, because when they switch to pegasus or eudora the problem goes away. BUT, just in case it is a POP3 Server probelm.. I'd like to upgrade think the lataset version of the qpopper package is 2.2. It works wells with Netscape Mail and should work well with Internet Explorer as well, and with Remove Mail from the Server selected. Regards, E.- -- Eloy A. Paris Information Technology Department Rockwell Automation de Venezuela Telephone: +58-2-9432311 Fax: +58-2-9430323
Thank you very much (was: Extremely rare Apache configuration)
Hello guys, I just want to express publicly my gratitude to all of you (direct recipients of this message) for having pointed me in the right direction regarding the question I posted today to debian-user: Can one proxy server be configured to go through another proxy server for certain addresses??? All of you recommended to use Squid, and let me tell you something about it: Squid is a miracle!!! I am really impressed with it and how configurable it is. Credits also go to Debian GNU/Linux and its developers for having a Squid package that works out of the box. Once again: thank you very much. Regards, Eloy.- -- Eloy A. Paris Information Technology Department Rockwell Automation de Venezuela Telephone: +58-2-9432311 Fax: +58-2-9431645 Cel.: +58-16-234700 Where does this path lead? said Alice Depends on where you want to go. Said the cat (Alice in Wonderland, by Lewis Carroll.) -- This message was delayed because the list mail delivery agent was down.
Extremely rare Apache configuration
Hello everyone, I have the following situation: my company has a HTTP proxy server to access Internet WWW sites. All browsers are required to use this proxy because we all are behind a firewall. I am overseas and connected to our corporate headquarters through a very slow satellite link. I can not give my users unlimited access to the Internet through the proxy because of the limited bandwidth, which is needed for more important tasks. There are certain hosts in the Internet my users need to access, but again, to do this they HAVE to go through the proxy. What about this: set up Apache locally as a proxy server for my local users and, at the same time, to have Apache contacting the corporate proxy server to access Internet hosts. I know it sounds confusing and do not know if I am understood... Another way of putting it would be: can one proxy server be configured to go through the another proxy server for certain addresses??? Any help will be very very appreciated. Thanks in advance, Eloy.- -- Eloy A. Paris Information Technology Department Rockwell Automation de Venezuela Telephone: +58-2-9432311 Fax: +58-2-9431645 Cel.: +58-16-234700 Where does this path lead? said Alice Depends on where you want to go. Said the cat (Alice in Wonderland, by Lewis Carroll.) -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
poppassd_1.2-4_i386.deb bug?
Hi, To get poppassd working under Debian 1.2, I had to change the P2 string in line 184 of poppassd.c from: enter new password: to: changing password for *\nenter new password: This fixes the problem of getting the poppassd program hung while waiting for passwd to finish execution. Are you the right person to report this or a need to got to someone else? I think a new version is needed to solve this since the package does not work out of the box. Finally, there are problems when the user types an invalid password (less than 6 chars. or without both numbers and letters). To solve this I execute the passwd command with the -o option, to force any password typed to be valid. I do not know how to fix poppassd to work without doing this. The problem seems to be that the passwd program is too picky and behaves very interactively (too much, I would say.) Regards, E.- -- Eloy A. Paris Information Technology Department Rockwell Automation de Venezuela Telephone: +58-2-9432311 Fax: +58-2-9431645 Cel.: +58-16-234700 Where does this path lead? said Alice Depends on where you want to go. Said the cat (Alice in Wonderland, by Lewis Carroll.) -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
Re: poppassd_1.2-4_i386.deb bug?
Please submit this as a bug to our bug tracking system. I just installed the package and I can't change the password with Eudora either. It times out waiting for the server to respond. OK, I'll do so. I did not do it before because even though I am die-hard Debian user, it's the first time I need to report a bug. In the meanwhile, if you want to get poppassd working, apply the patches I mention. Regards, Eloy.- -- Eloy A. Paris Information Technology Department Rockwell Automation de Venezuela Telephone: +58-2-9432311 Fax: +58-2-9430323 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
Debian review in Linux Journal
Hi, I'd like to know if someone can give me a link to a place where I can read the Debian review made by the Linux Journal staff in their November 1996 issue. From www.debian.org I thought the review was in September's issue so I order the back issue and sadly found out that in that issue there was just a distribution comparisson. Thanks in advance. E.- -- Eloy A. Paris Information Technology Department Rockwell Automation de Venezuela Telephone: +58-2-9432311 Fax: +58-2-9430323 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
Re: No ftp? Have I missed something?
Um, how am I supposed to connect to the ftp site if I don't have ftp? Heck, I don't have any inet capabilities except for ping! Maybe I missed something in the documentation, but it appears that have a pretty useless system installed here. Is there someplace I can look that will actually tell me how to set Debian up? I swear, this is really annoying. I have installed many complicated OS in my life, but this takes the cake :( Does everyone go through this crap, or did I get defective files or something? *sigh* What do you mean? If you can ping any host in the Internet then you have anything you want (FTP, Telnet, News, WWW, etc.) - unless you are behind a firewall or something. Any way, what's the specific problem? E.- -- Eloy A. Paris Information Technology Department Rockwell Automation de Venezuela Telephone: +58-2-9432311 Fax: +58-2-9430323 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
Re: Dayly maintenance time (was Re: Innd problem)
Hi Mike, Was the time of the maintenance script changed from 6:42 AM in 1.1 to 3 AM in 1.2? Yes, it has been changed to 03:08. However that is the setting in the default crontab. If you upgrade a working installation, the crontab will not be adjusted. You'll have to do that yourself: that's basic system administration. That's right, an upgrade of the cron package won't touch /etc/crontab unless one explicitly tells dpkg to go ahead and use the maintainer's version of this configuration file. In my case, when I upgrade from 1.1 to 1.2, I am going to background the dpkg process (Z option) and examine carefully the differences between the old and new configuration files of each package since I do not want to break things just for keeping the old files. Bye. E.- -- Eloy A. Paris Information Technology Department Rockwell Automation de Venezuela Telephone: +58-2-9432311 Fax: +58-2-9430323 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
Re: Dayly maintenance time (was Re: Innd problem)
If the 6:42 AM update time is a problem, just go into /etc/crontab and change the time to whatever you feel is best... I know, I know. I just wanted to know if the default time was changed from 6:42 AM to 3 AM because I think it is a more reasonable time for a dayly maintenance (at 6:42 AM I already have users working.) I know I can change /etc/crontab to suit my needs but maybe the not-so- knowledgeable user won't. E.- -- Eloy A. Paris Information Technology Department Rockwell Automation de Venezuela Telephone: +58-2-9432311 Fax: +58-2-9430323 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
Re: ping reply without OS (was: Re: Unidentified subject!)
Yes, but I've seen it quite often. The symptoms are you can't connect with anything, rlogin, telnet or such programs; however the machine gladly replies to ping. I think the situation arises when an OS _has_ been running on the machine and then crashes or hangs; in this state the ethernet card has been assigned an IP address, and then it just answers any ping requests that comes its way, without any intervention of the OS. Well, I am not an expert but I think that in these cases where the OS crashes and one can still get ping responses from the dead box it is because the part of the kernel that handles ping request is still alive. I guess it is because this is a very low level task, not as telnet, rlogin and such programs that are at a very high level. What you all think? E.- -- Eloy A. Paris Information Technology Department Rockwell Automation de Venezuela Telephone: +58-2-9432311 Fax: +58-2-9430323 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
Re: maintaining apache-style auth files
Does someone know of a good set of utilities for doing this? The format is simple, just like the 1st 2 fields in /etc/passwd. I am actually using it for a squid proxy authentication file. There is a program called htpasswd that is part of the Apache package that allows you to add users to the authentication file. E.- -- Eloy A. Paris Information Technology Department Rockwell Automation de Venezuela Telephone: +58-2-9432311 Fax: +58-2-9430323 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
Re: Unidentified subject!
ping response can be handled by network hardware without OS running. Uhhmmm... hadn't heard of this ever before. To reply to a ping request the network card has to have an assigned IP address... -- Eloy A. Paris Information Technology Department Rockwell Automation de Venezuela Telephone: +58-2-9432311 Fax: +58-2-9430323 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
Re: [1.2 installation] resolv.conf without bind; or should I just use bind too?
Should I just create resolv.conf my hand, or should I get the one from the bind package for whatever useful defaults (including explanatory comments) it contains? According to my /var/lib/dpkg/info/bind.list file (Debian 1.1), the bind package does not include /etc/resolv.conf. I think this file is build when configuring the base package from input given by the user. Any way, if you missed that part of the installation, you can build the file by hand: just create it with your favorite text editor and include a line like: nameserver xx.yy.ww.zz where xx.yy.ww.zz is the IP address of your ISP's name server. You can take a look at resolv.conf's man page (man resolv.conf) but for this you'll probably need the bind package installed. Finally, you do not need to run named (BIND) if you get connected to the Internet through your ISP and are not in a LAN. E.- -- Eloy A. Paris Information Technology Department Rockwell Automation de Venezuela Telephone: +58-2-9432311 Fax: +58-2-9430323 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
Re: Switching to Debian from old Slackware 2.1; orientation?
I installed qmail, an alternative to sendmail, on my system by hand, because there is no Debian package yet. So I created a script 'qmail' that starts qmail, if you give it the argument 'start' and stops it if you give the argument 'stop'. I placed the script in /etc/init.d. I want qmail to start when I enter runlevel 2, so I had to create a link in /etc/rc2.d to the script 'qmail': ln -s /etc/init.d/qmail /etc/rc2.d/S19qmail The 'S' means start qmail, the 19 makes it run between S18netbase and S20xinetd. I also deleted the link, '/etc/rc2.d/S20sendmail' because I didn't want sendmail to start. (But I still have the script in /etc/init.d, in case I ever want to use it again.) Finally, I created a link ln -s /etc/init.d/qmail /etc/rc6.d/K19qmail so that qmail is killed when I reboot (runlevel 6 is reboot). Isn't the recommended way of installing scripts in /etc/rc?.d to use the Debian command update-rc.d? E.- -- Eloy A. Paris Information Technology Department Rockwell Automation de Venezuela Telephone: +58-2-9432311 Fax: +58-2-9430323 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
Re: Problems list for 1.2 upgrade/install
This is the current list of known problems with the 1.2 installation. . . . Hey, thanks for maintaining this. I am waiting for my I-Connect Debian 1.2 CD ROM. When I put my hands on it this list will be very helpful when I upgrade from 1.1. By the way, I saw in one message from the list that someone is maintaining this list at http://ece.wpi.edu/~rulnick/GlinuX/debian-1.2-faq.html. i haven't taken a look at it, though. Regards, E.- -- Eloy A. Paris Information Technology Department Rockwell Automation de Venezuela Telephone: +58-2-9432311 Fax: +58-2-9430323 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
Re: New 1.2 install / Observations / X-window startx problem.
3. OK now, request help for this: My preference in using Linux is doing most of my editing work on a console and only when I need graphics go to X-windows. So I prefer to use startx and not xdm. The XFree86 3.2 that comes with slackware works fine. However, with debian, the screen flashes once, X-windows seem to come up for a second and then promptly exit. Obviously, I need to do some additional configuration. Something to do with the window managers perhaps? I would appreciate any pointers / suggestions on this. I remember when I first installed X in my Debian 1.1 box that X came up for a second or so and the exit. After I examined the log files in /var/log I realized that the problem was that my old XF86Config file that worked under a previous installation of X under Slackware was referencing my mouse as /dev/mouse, being /dev/mouse a symlink to /dev/psaux, the real device. After I changed the XF86Config file to reference the real file in /dev everything worked fine. Try looking at the differences between your Slackware setup and your new Debian setup. Take a look at the log files from the X server. The answer has to be there. Regards, E.- -- Eloy A. Paris Information Technology Department Rockwell Automation de Venezuela Telephone: +58-2-9432311 Fax: +58-2-9430323 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
Re: List of installation problems for 1.2
I have been maintaining the list, and will continue to add problem reports as they surface. So far (I'm still recovering from the holiday) I have seen no reports that indicate any new problems. Should I just post the list to debian-user about once a week? Once a week sounds good to me. Thanks! -- Eloy A. Paris Information Technology Department Rockwell Automation de Venezuela Telephone: +58-2-9432311 Fax: +58-2-9430323 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
Re: access log not writing
Changed out my access log for my apache server this morning (once a month). Moved the current one to .0 and put an empty file in and set all the chmod's and chown's to their right settings. But when I test it with lynx or my web browser, nothing is written to the access.log file. I have included below what the directory is set as. Is there something that I may have missed? Did you try restarting the httpd daemon? I think there is one Debian specific command to do this but can't remember. You can do a kill -HUP pid of appache-httpd to restart Apache. Good luck. E.- -- Eloy A. Paris Information Technology Department Rockwell Automation de Venezuela Telephone: +58-2-9432311 Fax: +58-2-9430323 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
Dayly maintenance time (was Re: Innd problem)
It means exactly what it says. Every day (by default at 3:00 AM) the maintenance script news.daily has to run. This probably hasn't happened for some reason or the other since a couple of days. If you fix that, the system will stop sending you mail about it.. I am waiting for my I-Connect Debian 1.2 CD ROM that was already shipped so probably I'll know the answer when I upgrade from 1.1... but in the meanwhile: Was the time of the maintenance script changed from 6:42 AM in 1.1 to 3 AM in 1.2? If this was the case I am glad because 3 AM is a more reasonable time for a dayly maintenance script that 6:42 AM. Regards, E.- -- Eloy A. Paris Information Technology Department Rockwell Automation de Venezuela Telephone: +58-2-9432311 Fax: +58-2-9430323 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
Re: Dual-Boot on a non-sys disk
Now I have someone that wants to install a dual-boot win'95/Linux system, but the only available hard drive space is on drive E: (a new gig drive). Can I just set it to a boot disk in the BIOS, repartition the new drive and pray that it works? Or do i need to install it to the C: drive? You can install Linux wherever you want. To boot it up you will use a boot manager (such as LiLo - Linux Loader) or you'll boot from a floppy disk which is configured to mount the root file system in any hard disk you have (for example, /dev/hdb2 - the second partition of your second IDE drive, could be your DOS E: drive). Good luck. E.- -- Eloy A. Paris Information Technology Department Rockwell Automation de Venezuela Telephone: +58-2-9432311 Fax: +58-2-9430323 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
Re: rm -r/-R/--recursive doesn't seem to work
I just built a distribution tree from a CD on my hard disk. I tried to
remove all the TRANS.TBL files from the tree by using:
rm -r TRANS.TBL
If I do this in a directory that has a TRANS.TBL it, and only it, gets
removed. If I do this in an upper directory without a TRANS.TBL file, I
get the error no such file or directory
Am I doing something wrong, or is this a bug in rm?
Dwarf:
try this:
find . -name TRANS.TBL -exec rm \{\} \;
(you need the backslashes to escape {} ; because these have special meaning
in the bash shell - see find(1) for more details on the -exec parameter
of the find command)
Them rm -R command does do what you want. I have used it to delete an entire
subdirectory structure as in: rm -Rf subdir. Here, everything under subdir
gets deleted, including the directory subdir itself.
Good luck.
E.-
P.S. I love the find command :-)
--
Eloy A. Paris
Information Technology Department
Rockwell Automation de Venezuela
Telephone: +58-2-9432311 Fax: +58-2-9430323
--
TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to
[EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
Re: rm -r/-R/--recursive doesn't seem to work
Them rm -R command does do what you want. I have used it to delete an entire ^^ subdirectory structure as in: rm -Rf subdir. Here, everything under subdir gets deleted, including the directory subdir itself. OOPPP!!! I really meant: the rm -Rf command DOES NOT do what you want Sorry for the mistake. Eloy.- -- Eloy A. Paris Information Technology Department Rockwell Automation de Venezuela Telephone: +58-2-9432311 Fax: +58-2-9430323 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
Re: It happened again.
It just happened again. Only this time I was only running Lynx. As soon as I killed lynx, it stopped. It only managed to get up to 5megs in my swap space. This is weird. It seems as though it only happens when I'm connected. I've never noticed or remember it happening when I was not. Did you find anything strange in the log files? I remember one time that my swap space usage was increasing pretty quickly. I found out it was because /var/log/daemon.log was being filled with output from named. I got too much junk that it filled my partition. Regards, E.- -- Eloy A. Paris Information Technology Department Rockwell Automation de Venezuela Telephone: +58-2-9432311 Fax: +58-2-9430323 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
Re: Poppasswd
Hi, The -o option of the passwd command (in Debian 1.1 - don't know about 1.2) disables all checks. Hacking poppassd to call passwd with the -o option should be trivial. E.- This is a very tricky problem. poppassd is a small app that calls passwd to change the password. Well, it passes some arguments to passwd and expects some responses from it. All very fine. Now, the newer debians use a passwd program that does a few checks on the password. If the new password is too similar to the old one, or if it is too short, or if it has too many repeated characters, etc. it will issue an error message and prompt you for a new (hopefully better) password. poppassd was not desinged to deal with this. It just waits for passwd to issue the prompt re-enter new password while password is saying too simple: try again. Try entering a very random, 8 chars password to see if it works. To solve your problem, you must try to find a passwd program that doesn't do these checks (at the expense of security) or hack poppassd to be smarter. I don't know if debian has a simpler passwd program. Hope this helps, Nelson [EMAIL PROTECTED] -- Eloy A. Paris Information Technology Department Rockwell Automation de Venezuela Telephone: +58-2-9432311 Fax: +58-2-9430323 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
Re: Network trouble (?) with kernel 2.0.27
I've been working with a fresh instalation of 1.2 including a reconfigured kernel 2.0.27. I'm getting lots of messages like the following in /var/adm/messages: Dec 19 22:37:19 blackbird last message repeated 569167 times Dec 19 22:38:19 blackbird last message repeated 577432 times Dec 19 22:39:19 blackbird last message repeated 569323 times Dec 19 22:40:19 blackbird last message repeated 575027 times Good, but what's the last message? E.- -- Eloy A. Paris Information Technology Department Rockwell Automation de Venezuela Telephone: +58-2-9432311 Fax: +58-2-9430323 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
Re: Apache Installation
While I have your attention, how does one restart the server once these changes are made? /etc/init.d/apache reload I did not know of this sophistication. I just do a kill -HUP apache's PID Regards, E.- -- Eloy A. Paris Information Technology Department Rockwell Automation de Venezuela Telephone: +58-2-9432311 Fax: +58-2-9430323 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
procmail as local mailer for sendmail
Hi, A couple of weeks ago I changed the default local mailer for sendmail (deliver) with procmail (this was a suggestion of [EMAIL PROTECTED]). The reason for this change was that deliver does not notify comsat of new mail so I ca no get asynchronous notification of incoming mail through the biff program. Well, everything works fine except that with the default setup some situations are not handled right: when I am using vi and am connected through modem to the system and get disconnected because of noise, the cron daemon sends me an e-mail leeting me know how to recover the file I was editing. Well, in the list of processes I see a sendmail -t and a procmail -Y -a -d eparis. The procmail process never finishes so sendmail (the parent process) does not finishes either. I guess the problem is this -a switch that according to procmail's man page, requires a parameter following it. I don't see this parameter when the cron daemon sends the message to me. I can solve the situation by getting rid of the -a switch in the local mailer args. This is the important part of my sendmail.mc file: FEATURE(local_procmail, /usr/bin/procmail) define(`LOCAL_MAILER_ARGS', `procmail -Y -d $u') MAILER(local)dnl If I get rid of the define statement above I get the problem described. This is because the default for the LOCAL_MAILER_ARGS define is `procmail -Y -a $h -d $u'. Does any one has a clue of what's happening with this -a $h switch and why the cron daemon has problems sending local mail? Thanks in advance. E.- -- Eloy A. Paris Information Technology Department Rockwell Automation de Venezuela Telephone: +58-2-9432311 Fax: +58-2-9431645 Where does this path lead? said Alice Depends on where you want to go. Said the cat (Alice in Wonderland, by Lewis Carroll.) -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
Re: Warm Boot Problems
At 10:25 AM 12/18/96 -0700, you wrote: I seem to be having trouble booting with Debian version 1.2. Hardware: 486 sx-40 (AMD) 20 M memory 820 M Seagate HD using DM when in Dos When I cold boot (reset button) The boot floppy will load and run When I WARM boot, (shutdown and restart) the loader hangs at the word 'David' when loading one of the cd-rom divers. When I try to boot using loadlin from dos, first, the speed shown is 13.11 BogoMips vs 26.11 from booting from the floppy and the program hangs up when loadng the keyboard driver. Are you using any network card? If not I have no clue of what's going on. If yes try putting /sbin/ifconfig eth0 down just before reboot -d -f in /etc/init.d/reboot. This will take care of shutting properly the Ethernet card that may be causing problems. Good luck. E.- -- Eloy A. Paris Information Technology Department Rockwell Automation de Venezuela Telephone: +58-2-9432311 Fax: +58-2-9431645 Where does this path lead? said Alice Depends on where you want to go. Said the cat (Alice in Wonderland, by Lewis Carroll.) -- This message was delayed because the list mail delivery agent was down.
Re: deliver and biff and comsat
I chose to just change the ling in sendmail that points to procmail, since I don't regenerate sendmail.cf very often. The symlink would work too, though. Hi, thanks again for your response. I chose the symlink method because I generate sendmail.cf very often. It works great. As you know, sendmail depends on deliver. I think it should depend on a local delivery agent (local mailer) and not just deliver, and the user should be given the choice of which local mailer to use. Thanks, Eloy.- -- Eloy A. Paris Information Technology Department Rockwell Automation de Venezuela Telephone: +58-2-9432311 Fax: +58-2-9430323 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
deliver and biff and comsat
Hi, I really miss the biff utility in my Debian system that is running sendmail. After a long search I found out that is the local mail delivery agent the responsible to notify comsat of the arrival of new mail. Then, comsat displays asynchronously a notification of new mail. Since I am using sendmail instead of smail, the MDA is deliver. Well, it happens that I do not see any options to make deliver talk to comsat so I am screwed up. I tried smail sometime ago and saw that it does not rely on deliver because it has this capability built-in. Besides, there is one configuration flag called notifycomsat that is used to turn on and off this behavior. All this introduction is to ask if someone knows a way of getting a sendmail based system to notify comsat of new mail arrival. Does any one know of a patch for deliver or another MDA that can work with sendmail? Thanks in advance. Eloy.- -- Eloy A. Paris Information Technology Rockwell Automation de Venezuela Telephone: +58-2-9432311 Fax: +58-2-9430323 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
Re: How do I upgrade from 1.1 to 1.2?
Hi, At 08:39 PM 12/5/96 -0600, you wrote: This is the KERNEL numbering system, not the Debian dumbering system. For the record, x.y.z where y is even is a stable kernel. y odd means a development kernel. For debian, every numbered release (with the exception of 1.0) is a stable release. The development releases (which are in general relatively stable, although things break from tinme to time) are only given code names. But I thought Debian 1.2 (rex) was unstable (at least until it is released.) Then, if only development releases are given code names, how come buzz (Debian 1.1) is stable? Thanks, Eloy.- -- Eloy A. Paris Information Technology Rockwell Automation de Venezuela Telephone: +58-2-9432311 Fax: +58-2-9430323 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
How do I upgrade from 1.1 to 1.2?
Hi, I am sorry if I ask a couple of dumb questions... 1) When 1.2 is released, will it be named 1.2 or 1.3? I thought odd revision numbers were for stable releases and even ones for development releases (I guess this was a side effect after a wrong version of Debian was put on CD, I think originally it was the opposite was: odd for development and even for stable) 2) Can I upgrade from 1.1 to 1.2 (or whatever it's called when it's released) without changing my current system configuration or breaking any packages or something? Thanks to everyone. Eloy.- -- Eloy A. Paris Information Technology Rockwell Automation de Venezuela Telephone: +58-2-9432311 Fax: +58-2-9430323 -- This message was delayed because the list mail delivery agent was down.
It could happen to you (was Re: network cards and POST failures)
Hello everyone, at last, the strange problems I was having with my Linux box at reboot time is solved. I followed Donald Becker's recommendation about putting /sbin/ifconfig eth0 down at the very end of the reboot script and this fixes everything. Now everything is crystal-clear to me. I agree with Bruce that the Debian distribution should take care of shutting down Ethernet devices before rebooting. As an example, I was getting this error in my desktop machine everytime I rebooted the Linux box: eth0: Bus master arbitration failure, status 88f2 Since my desktop machine is not a production machine (I am the only one that uses it) and that I first saw this error after a kernel upgrade, I did not waste much time finding the cause. Each time I got that error I just cycled power to the computer and everything worked fine after that. Well, the Ethernet card here is also a NE2100 compatible (it's a BocaLAN Card, using the lance.c driver) and it does DMA. After I put /sbin/ifconfig eth0 down at the bottom of /etc/init.d/reboot the problem went away. Beware!!! This problem could happen to you and you'll waste hours trying to find what's wrong. Regards, Eloy.- P.S. Somebody may want to post this to debian-devel@lists.debian.org because I have no posting priviledges there. At 12:46 PM 12/2/96 PST, Bruce Perens wrote: Here's what Donald Becker had to say about Eloy Paris' problem in which a network card caused RAM POST failures. This is perhaps something that we should take care of in the shutdown scripts after NFS partitions are unmounted and daemons are killed. We can use the output of netstat -i to figure out what interfaces to shut down. From: Donald Becker [EMAIL PROTECTED] On Sat, 30 Nov 1996, Bruce Perens wrote: Eloy Paris [EMAIL PROTECTED] has a DEC PC that fails the RAM POST only when the Boca PCnet32 card is present and a recent Linux kernel has just shut down. My uninformed speculation is that the card could be doing wild DMA, and I'm wondering if there's any special step that should be taken to shut down this card before rebooting. That's a common BIOS bug. The warm boot code should disable the bus-master capability on all cards. What is happening is that the card is continuing to receive and store packets into memory. The work-around, which we must use here as well, is to put /sbin/ifconfig eth0 down /sbin/ifconfig eth1 down /sbin/ifconfig eth2 down /sbin/ifconfig eth3 down in you shutdown script. (Usually /etc/rc.d/rc.6, but Redhat is different.) This will disable the cards. -- Eloy A. Paris Information Technology Rockwell Automation de Venezuela Telephone: +58-2-9432311 Fax: +58-2-9430323 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
Re: new sendmail version won't start
At 11:18 AM 12/2/96 +0100, you wrote: While using dselect to install some new software, the program downloaded a new version of sendmail: 8.7.6-2. This version won't start however, here is the trace: . . . Same happened to me. Thanks God I did not install it in a production machine. Also, the last gzip upgrade depends on a libc5 that is not part of Debian 1.1 (buzz) My guess is that it's part of Rex (unstable) I wonder what happened to the last buzz update... In both sendmail and gzip cases, I went back to the previous versions. Regards, E.- -- Eloy A. Paris Information Technology Rockwell Automation de Venezuela Telephone: +58-2-9432311 Fax: +58-2-9430323 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
Re: 2 lpd processes; extra processes - possibly related?
At 04:34 PM 11/30/96 -0800, you wrote: I am running a basically off the shelf debian installation that is about 2 weeks old, so I haven't messed with it much. The first oddity is that, upon close inspection of ps -aux, I note two copies of /usr/sbin/lpd running. Any ideas why? I poked through the rc* stuff, but not being real used to the layout (coming from slack 3.0), I don't notice anything particularly obvious. Probably it's because there is something in the print queue (try the command lpq to see what's in the queue) If you see something try lprm - as root to get rid of all jobs in the queue. The second thing that makes me curious is the fact that there seem to be a lot of processes going by that I am not able to follow. What might they be? Well, you are running X with a clock, a Windows Manager (fvwm), a couple of X terms, a PPP connection, etc. It's normal to have this number of processes, unless you're talking about something else. Good luck. Eloy.- -- Eloy A. Paris Information Technology Rockwell Automation de Venezuela Telephone: +58-2-9432311 Fax: +58-2-9430323 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
Stable unstability (was Re: modules 2.0.0-14)
At 12:17 PM 12/1/96 -0600, you wrote: modules 2.0.0-14 has destroyed my System. Every attempt to install a module has created a kernel Oops. I had to build a new kernel with all needed drivers linked in. Downgrading to modules 2.0.0-13 solved this Problem. . . . Same here, with the same configuration. Email me for details. Why does it seem so many drastic changes are being made to stable? For example, the PCMCIA package has been removed and apparently rolled up into the kernel-sources package. Why? I agree: why so many drastic changes are being made to stable? The new gzip (in Debian 1.1.16 I think) depends on a libc5 that is not part of stable so if users want to upgrade a package that is part of stable (this gzip), they'll have to unstable and grab the libc5 package. Also, the last sendmail seems to have problems. Regards, Eloy.- -- Eloy A. Paris Information Technology Rockwell Automation de Venezuela Telephone: +58-2-9432311 Fax: +58-2-9430323 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
Re: Gosh, I can't believe this!!! (was: who's guilty...)
At 09:24 PM 12/1/96 -0800, you wrote: I'll offer this guess: the troublesome card does DMA, and is not being shut down before the kernel reboots. It writes to its previously programmed buffer address during the memory test, which of course causes the memory test to fail. Warm boot works because it does a very cursory memory test (if any), allowing only a short time window for the card to interfere; cold boot fails because of a more thorough test. Perhaps the change between kernels 1.2 and 2.0 was that the driver now takes advantage of the card's DMA capabilities? Or something isn't getting shut down properly when switching to runlevel 0? Or the driver doesn't finish or cancel outstanding operations when it's closed? Uhhmmm... your explanation sounds very good to me but dmesg reports that DMA is not needed when it detects the card. Also, the DOS utility used to configure the IRQ and I/O address of the card does not provide the capability to configure a DMA channel. ??? Regards, E.- -- Eloy A. Paris Information Technology Rockwell Automation de Venezuela Telephone: +58-2-9432311 Fax: +58-2-9430323 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
Re: Goodbye, all! (Whatever became of X3.2)
At 06:49 PM 11/29/96 -0500, somebody wrote: However since you have said so, I officially hand in my resignation. I wish you good luck with Debian in the future. Come on guys!!! I am not a Debian developer (yet) but I am a Debian user and I think we need everybody's effort in order to make Debian the best Linux distribution ever available. I do not think we should fight just by a stupid discussion about wheter to put X Free 3.1 or 3.2 in the new Debian release. I thought (maybe I missed something) that everybody agreed on delaying the release until X Free 3.2 was fully tested, because of the security hole in 3.1. So what's the problem? I hope nobody leaves the Debian project for such a thing like this. Have a nice day. Eloy.- -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
Re: Who is guilty, the kernel or Debian?
Hi, At 02:05 PM 11/28/96 PST, you wrote: Can you revert to the 2.0.6 kernel and tell us if that one causes a problem? Nope, it didn't work. I re-installed a fresh kernel source tree (2.0.7), re-configured it for my basic hardware (IDE, NE2100 compatible card, serial driver, MS-DOS+ext2 fs support and nothing else), and re-booted and it did not work. This time a got: LILO Loading Linux Uncompressing Linux... crc error -- System Halted Now, if I press the reset button, or cycle power, Linux will come back as if nothing happened. It will be up for days... unless I execute a reboot command (reboot, shutdown -r). In my opinion, this error is another side effect of the real cause. Consider the three different effects I am seeing (one at a time, of course): 1) A sequence of 1-3-3 beeps: this means 1st 64KB RAM chip or data line failure (according to the computer's manual) 2) The BIOS displays something like Memory failure at : expected and found - Decreasing available memory, please run setup program. 3) CRC error after decompressing the kernel. All of the above situations could be caused by kernel interaction with BIOS, possibly related to memory management bits not cleared by reset on your system (as Bruce Perens said) I have another machine exactly the same as this one (same model, same memory) I can take my hard drive to that machine and I bet it will do the same. It has to be a software problem. Uhhmmm... should I try and go back to 2.0.0??? Any ideas??? I am lost. I had not seen anything like this before. Eloy.- -- Eloy A. Paris Information Technology Rockwell Automation de Venezuela Telephone: +58-2-9432311 Fax: +58-2-9430323 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
Who is guilty, the kernel or Debian?
Hello everyone, I am writing this message while I am frustrated after fighting over and over and over with a Debian 1.1 installation. This is my story: Problem --- - EVERYTHING WORKS FINE EXCEPT AFTER REBOOTING. If I start Linux after a cold start (reset switch or power cycle) Linux starts just fine; I can compile the kernel, I can do networking (IPX, TCP/IP, SLIP, PPP, DNS, etc.), I can do EVERYTHING... everything but rebooting: if I shutdown -r +1, Ctrl-Alt-Del or reboot one of two things will happen: 1) After Linux does all the shutdown stuff (killing processes and unmounting file systems) and just after the Rebooting... message is displayed on the screen I will get some beeps generated by the BIOS. The number of beeps is 1-3-3 and according to the user's manual this means 1st 64KB RAM chip or data failure 2) After the Rebooting... message is displayed, the computer seems to reboot but the BIOS checks RAM and finds something wrong: I get a message similar to Memory failure at : expected and found - Please run setup program. I have found that sometimes I also get the error CRC error after Lilo decompresses the kernel. Known Facts --- - Computer: Digital DECpc LPx 466d2 (genuine Intel 80486DX2 running at 66 MHz), 16 MBytes of RAM, and 128 KBytes of L2 cache. This computer has been running Slackware 1.3 with kernel 1.2.13 for about one and a half year. It's been a SMTP, WWW, POP3, and DNS server (dedicated) in a network of 50+ users with no problems at all. - Hard drive: first tried a 1.2 GBytes IDE hard drive, and after I started to have problems (see above for a description of the problem), I switched back to the original Digital hard drive (224 MBytes) - didn't help, I got the same problems. - Network card: BOCALANcard-VL (local bus a NE2100 compatible). I have used always the lance.c driver. - It seems the server is 100% stable (if I do not reboot) - Debian 1.1. I have applied all upgrades in buzz-updates (up to Debian 1.1.14) - Kernel: started with 2.0.6 and when I switched to 2.0.26 problems started. I tried kernel from 2.0.20 and up and always got same problems. What I have done 1) I have installed Debian 1.1 several times. First times using one of the standard boot disks and the last time I built one custom boot disk (using the boot-disks and kernel-package packages) - did not work. 2) I have installed Debian from two different sources: FTP (through dpkg-ftp) and from SCSI CD-ROM. Did not work. 3) Recompiled the kernel from 2.0.20 and up and nothing good happened: if there is a problem with the kernel, it is already present in 2.0.20. My next move would be to try with the older kernel I have (2.0.7) 4) The first time I thought the problem was due to the fact that there is no BIOS option to support LBA. Because of this I changed my 1.2 Gig. disk with the one that came originally with the computer (a 224 Meg disk) I reinstall Debian on it and the problem showed up again. 5) (I am not too sure I did this one) I disabled the secondary cache. Theories - I do not think the problem is hardware: again, this server ran Slackware 1.3 with kernel 1.2.13 for one and a half years. I DID NOT change anything for Debian 1.1 and kernel 2.0.x. - Something broken in the kernel. - There is something wrong with the reboot program. - I am using the new IDE driver. May be there is something wrong with it? Conclusion -- I am lost. I have installed Linux (Slackware and Debian) many times in computers ranging from desktops to notebooks. I have used many kinds of hardware (PCMCIA, SCSI, IDE, multiport serial boards) with Linux (Debian and Slackware) and never had a problem. I am posting this message here because I am a Debian user. However, I do not know if the problem is Debian or the kernel. This machine will be a very important part of our IT structure, it will get rid of a FreeBSD machine, it will be the gateway between two networks connected through a leased line and a dedicated PPP link. I can run the machine the way it is now but then I won't be able to reboot remotely. I can not go 4 blocks away every time I need to reboot the server. I NEED STABILITY!!! (even at reboot time) I want to know who is guilty: the kernel? Debian? Somebody else? I am looking for help, support, new ideas, new clues, anything... PLEASE HELP!!! Thank you very much Debian user. Regards, Eloy.- -- Eloy A. Paris Information Technology Rockwell Automation de Venezuela Telephone: +58-2-9432311 Fax: +58-2-9430323 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
Re: Who is guilty, the kernel or Debian?
Hello Bruce, thanks for your response. At 02:05 PM 11/28/96 PST, you wrote: Does the system come up without complaint after a power-clear? When I turn on the computer Linux starts with no problems at all (no complaints) If I press the reset switch the same thing happens. Can you revert to the 2.0.6 kernel and tell us if that one causes a problem? Absolutely. I said in my message that was going to be my next move. I will do that right now. It sounds like a kernel interaction with BIOS, possibly related to memory management bits not cleared by reset on your system. I was thinking about something like this but it should have happened after some change in a 2.0.x kernel because with 1.2.13 I am pretty sure it never happened. If it is not the kernel, it could be the reboot() function (the one in unistd.h that is been called from the halt.c, part of the SysVInit package.) I forgot to mention in my message that after the 1-3-3 beeps, the system hangs (nothing is wrong with Linux afterwards because file systems were unmounted and the system was brought down orderly) I need to cycle power or press the reset button to restart. I have noticed in computers that work fine with 2.0.x that the reboot caused by a shutdown -r ... is warm, not cold as it seems to be in the case of the computer I am having problems with. The only other things I can think of are the hardware tweaks we do: the software watchdog timer, and the serial interrupt reprogramming. Those are unlikely. If you remove the watchdog and hwtools packages, that would eliminate those programs as candidates. My kernel is not compiled with watchdog timer support nor real time clock support nor something new or unconventional, just the basic stuff. I have not installed watchdog nor hwtools either. Well, let me go back to 2.0.7 and we'll talk later. Thanks, Eloy.- -- Eloy A. Paris Information Technology Rockwell Automation de Venezuela Telephone: +58-2-9432311 Fax: +58-2-9430323 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
