Re: Machine/Domain Name
Chris, what you see below is what you need to change. I'm not sure what you mean when you say cwaiken.com is a redirector. If it just relays email for you that isn't good enough. When you attach to a SMTP server the HELO or EHLO is given with a single parameter: your fully-qualified domain name. Some servers are configured to try to check this domain name by resolving it to an IP and then checking to see if that IP matches the IP address of the peer on the connection. If such a check doesn't match it won't take mail. So, what you need to do is if you're on a dialup you need to update the exim config file as Terry has indicated and restart exim. Alternatively you could register with a dynamic dns provider, to have something like cwaiken.dyndns.org. In either case you can effect this by modifying your PPP ip-up script(s). Terry Boon wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On Fri, Jan 12, 2001 at 10:51:17AM -0500, Christopher W. Aiken wrote: I have exim setup with my ISP's smtp server for outgoing mail. Mail gets delivered to everyone I send to except to my office. How do I change the name of my machine to darkstar.cwaiken.com? cwaiken.com is my domain name at a re-director service and is resolvable and should work. You may want to look at primary_hostname in your exim.conf. The following is the automatic documentation of the variable: # Specify the name of the current host. This is used in the HELO # command for outgoing SMTP messages, and as the default for qualify_domain. # If it is not set, Exim calls uname() to find it. primary_hostname = [snip] - -- Terry Boon, Hertfordshire, UK [EMAIL PROTECTED] -- Jens B. Jorgensen [EMAIL PROTECTED] Shop online without a credit card http://www.rocketcash.com RocketCash, a NetZero subsidiary
Re: dhcp
Sebastiaan wrote: High, Run ps xa | grep dhcp. I you can see a line for dhcpd it is running, take a look at /var/log/syslog for dhcpd error messages. If there is no dhcpd running try editing /etc/init.d/dhcp, set run_dhcpd=1 and restart the dhcp daemon (/etc/init.d/dhcp restart). Debian sets run_dhcpd=0 to disable dhcpd from starting before a /etc/dhcpd.conf is properly set. I guess something went wrong during install. I did a apt-get install dhcp once and I remember that I got a message which tells me to edit /etc/init.d/dhcp. But I tried some other packages (bootp) and I got errors when uninstalling dhcp and it also would not reinstall. So I had to force some things, but now and empty /etc/init.d/dhcp is created and dhcpd.conf is not created at all. I tried to run /usr/sbin/dhcpd -d -f for debugging but that did not help. Is this an error in the dhcp package? I am using potato. Can anyone send me an original /etc/init.d/dhcp file? If you have the dhcpd package installed (which is what you'll need to run your dhcp server) then the example configuration file can be found as: /usr/doc/dhcpd/examples/dhcpd.conf.gz You mainly need to specify the range of IP addresses you'll dole out to clients. Since this will always vary from site to site the install package can't automatically set this and start up the server. -- Jens B. Jorgensen [EMAIL PROTECTED] ___ Why pay for something you could get for free? NetZero provides FREE Internet Access and Email http://www.netzero.net/download/index.html
Re: dhcp
Sebastiaan wrote: High, Run ps xa | grep dhcp. I you can see a line for dhcpd it is running, take a look at /var/log/syslog for dhcpd error messages. If there is no dhcpd running try editing /etc/init.d/dhcp, set run_dhcpd=1 and restart the dhcp daemon (/etc/init.d/dhcp restart). Debian sets run_dhcpd=0 to disable dhcpd from starting before a /etc/dhcpd.conf is properly set. I guess something went wrong during install. I did a apt-get install dhcp once and I remember that I got a message which tells me to edit /etc/init.d/dhcp. But I tried some other packages (bootp) and I got errors when uninstalling dhcp and it also would not reinstall. So I had to force some things, but now and empty /etc/init.d/dhcp is created and dhcpd.conf is not created at all. I tried to run /usr/sbin/dhcpd -d -f for debugging but that did not help. Is this an error in the dhcp package? I am using potato. Can anyone send me an original /etc/init.d/dhcp file? If you have the dhcpd package installed (which is what you'll need to run your dhcp server) then the example configuration file can be found as: /usr/doc/dhcpd/examples/dhcpd.conf.gz You mainly need to specify the range of IP addresses you'll dole out to clients. Since this will always vary from site to site the install package can't automatically set this and start up the server. -- Jens B. Jorgensen [EMAIL PROTECTED] _NetZero Free Internet Access and Email__ http://www.netzero.net/download/index.html
Re: BIOS local time to UT time
Just edit the file /etc/default/rcS and change the line with UTC to UTC=yes Lazar Fleysher wrote: Hi everybody, I used to have a machine running Windows and Linux, but now I decided to get rid off win. Currently, I have setup which relyes on BIOS clock set to local time. Since, I am converiting to linux only, I want to change that. I know how to set the BIOS clock to UT time :-), but which steps would I perform to reconfigure the operating system for that? Thank you Lazar -- Jens B. Jorgensen [EMAIL PROTECTED] ___ Why pay for something you could get for free? NetZero provides FREE Internet Access and Email http://www.netzero.net/download/index.html
Re: Subject: ATA-100, UDMA100, LILO
Yes, the better way to do this is to go into your BIOS and disable the motherboard' s IDE controller. That's the only way the Promise IDE controller can have hda. Vincent Gaines wrote: Subject: ATA-100, UDMA100, LILO I recently built a computer using the ASUS A7V motherboard, AMD 800, and an IBM Deskstar DTLA-307030 30GB UA100 hard drive. I wanted to install Debian GNU/Linux on my hard drive (it uses the Promise ATA-100 controller). 1) Upon reading the Debian Mail archives I surmised that the ATA-100 was not supported by the UDMA66 diskette install option from the ftp archiives. ( I also tried it!) 2) I installed using the UDMA66 diskettes and CDROM binaries from linux labs using my via udma66 controller ( I was later to find out to my dismay that the system uses hda, hdb, hdc, and hdd for the via controller and hde, hdf, hdg, and hdh for the promise ATA-100 controller). 3) I applied the patches from www.linux-ide.org to a linux-2.4.0-test6 kernel source from www.kernel.org and compiled my new kernel. 4.) When I booted on my custom kernel to my dismay I saw all devices operable however, my promise ATA-100 contoller was using hde. As far as I know there is no way to change this so. 5.) I hoped I could simply edit my lilo.conf file run liloconfig and voila!.No way! Since device hde did not exist at that time ( I was using hda, hdb) this coiuld not be. 6.) After many hours of strife ( and realizing that linux-2.4.0-test6 has the drivers in it for ATA-100) I decided to install a 2 GB maxtor on hda, put the custom kernel on it, and now Debian sees my hda, hdb (CDROM) and hde. 7.) I will now have to adjust my lilo.conf to boot windows 98 (hde1). There must be a better way to do this, someone please tell me . -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: POP - IMAP w/ Maildirs?
The Cyrus IMAP server (http://asg.web.cmu.edu/cyrus/) is designed to handle mail for users who don't have accounts on the machine. I'd take a look at it. I've been looking to do something like this and still haven't gotten it together. I'm going to suggest hacking UW IMAP to do this might be difficult. I didn't really consider it because I wasn't happy with its maildir support. Keith G. Murphy wrote: I have an application in mind where I download some folks' POP3 mail from their ISPs to our local server, then present their E-mail (they're Win95 users on our local network) using an IMAP server. So far I can do everything very simply using fetchmail + procmail -d + UW IMAP. However, I'd like to go to a solution, probably using maildirs, that does not involve messages in their home directories; in fact, maybe not requiring them to have accounts on our Linux server at all. I've been digging into the procmail documentation, and trying to figure out how to use maildirs in UW IMAP, but I've got the feeling someone out there has already done this. Any simple methods for this? TIA. -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: Internet Cafe
This shouldn't be too tough. I'd look at radius for authorizing and accounting. It's built for this kind of stuff. As for automatically logging people out this should be too tough. Just fire off a program when the xdm/gdm session starts. This little helper program could wait for a message from the box where your radius server runs. When the time is running out it can send a little message to the helper giving it a 5-minute warning. The helper would fire up a little message box (I believe gnome has a mini-app that'll do this--just display a text message) warning the user their time is running out. When the time actually runs out you can just kill the X server--or perhaps do something a little more graceful. ChrisHellberg wrote: I want to set up an internet cafe at a hostel and am investigating various ways of going about things. I think windows 2000 would be the best os to have the clients on, but the biggest problem is logging the pc's out when their credit expires. I'm a strong advocate of Debian GNU/Linux and open source so I would liie to use an opensource solution but not too sure if the clientele would be that comfortable with X. That being said, if I decide to go down the linux path, what would be the best way to log a workstation out when credit expires? I rekon it's a matter of creating a simple TCP/IP connection to a server and listening to logout requests :/ Cheers Chris -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: Which IMAP and POP3 servers ?
I feel I also have to throw in my opinion. First of all, I _highly_ recommend you use qmail and use maildirs. I also highly recommend using courier-imap. This will let your users gracefully deal with folders and the like. Brendan J Simon wrote: I need to install some IMAP and POP3 servers so users can read mail via Netscape/Outlook on other machines. I have exim as my mail server. I'm running Debian Potato on a PowerMac G4. The choice (according to apt-get -s install imap-server) of available imap servers is: imap 4.7c-1 courier-imap 0.31-1 The choice (according to apt-get -s install pop3-server) of available pop3 servers is: cucipop 1.31-13 qpopper 2.53-5 ipopd 4.7c-1 Does anyone have any preferences, suggestions or comments regarding these packages. I thought I would install imap and ipopd. Are these ok or are there better packages ? qpopper sounds interesting. Thanks for any advice, Brendan Simon. -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: ISDN newbie
If it has RS232 then you don't need ISDN4Linux as that package is for cards which don't look like a modem. If it's RS232 then it probably acts like a modem (accepts AT commands, etc.) and can be configured much like a modem. You need more info about the modem itself. Looking at its command reference should show you how to dial up your provider and get working. Mario Olimpio de Menezes wrote: Hi, I just received a isdn connection at home. Till now, I've a USR isa modem working great with Linux. The isdn standard in Brazil is the Europe Standard. The NT used has a RS232 adaptor and according to the telco, it has a internal isdn adaptor. Well, my doubts are: is it possible to use this RS232 to connect to Linux serial port and then to use ISDN4Linux? I read the ISDN4Linux faq but as far as I could understand, the faq deals only with the case where one has and isdn adaptor inside his PC. That is not my case, yet! Any pointer, clue, hints are welcome! Mario O.de MenezesMany are the plans in a man's heart, but IPEN-CNEN/SP is the Lord's purpose that prevails http://curiango.ipen.br/~mario Prov. 19.21 http://www.revistalinux.com.br -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: Drivers for DELL PERC 3/Si - RAID controller in PE2450.
Sorry, you'll have to install RedHat. Well, you could install debian but you'd need to hack up the install disks to use the RedHat 6.2 custom kernel provided by Dell. There's no other source for the driver. I was in the same boat and just decided to be lazy and install RedHat. [EMAIL PROTECTED] wrote: I want install Debian on Dell PowerEdge 2450 with DELL PERC 3/Si - RAID controller. This controller is based on Adaptec technology but I dont know wich. Does exists Debian driver for this controller? Is this contoller in Debian HCL :-)? Kamil Sotak MORAVIAPRESS, a.s. U pony 3061 690 02 Breclav Czech Republic Tel.:+420-627-305 173 Fax: +420-627-321 728 e-mail: [EMAIL PROTECTED] WWW: www.moraviapress.cz P.S.:Please use pure ASCII or ISO-8859-2 code -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: Mondo GNU like Norton Ghost
Have you checked out beoboot? I think there may be a free version and it can do what you want and even uses network multicast file transfer to avoid network saturation. Kent West wrote: Michael Banck wrote: On Mon, Aug 28, 2000 at 01:54:35PM +1000, Peter Firmstone wrote: For those who are interested there is a package called mondo which is GNU and does the same thing a Norton Ghost, namely allows you to take a system image and use it to install other computers or backup simply by booting from a cdr created earlier. I think this should become part of debian, anyone know how I can get it included in unstable? Has anyone used Mondo? Is it strictly a harddrive-to-harddrive solution, or can it be used over the LAN? I need something that allows me to duplicate a lab-full of PCs, like Ghost's multicast. Thanks! -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: Enabling Serial Port Console Login
No need to recompile the kernel. Your kernel should either have serial port support compiled in or in a module. On your server try 'cat /dev/ttyS0' If you get an error: cat: /dev/ttyS0: No such device Then your kernel doens't have support. Otherwise it does. Now you just need to run a getty on the port. This is the process that actually listens to the port for connections and allows you to log in. init starts the getty processes. Add the following line to /etc/inittab: S0:23:respawn:/sbin/getty 38400 ttyS0 To get init to reread /etc/inittab and start up a getty just run 'kill -HUP 1' or reboot. Now inittab will start a getty process on the port, setting it to 38400 baud. Of course you can change the baud to match whatever you're coming in at from your terminal server. You should now be table to log into your machine if it's got a crossover cable hooked up to COM1 (yes, you definitely do need a crossover cable unless it's a modem hooked up to COM1; computers and terminals are wired DTE on their serial ports so this is necessary). Peter Kim wrote: I need to be able to login to my Linux web server, via a communications server (Portmaster). The communications server will be connected to web server with a serial cable. I've connected my PC to my web server's serial port (COM1) and (COM2) with a cross cable and tried to login with Tera Term. Tera Term cannot connect. Nothing shows on the terminal and anything I type does not appear on the screen. I have heard that I have to recompile the kernel with some options set to make COM1 AND COM2 come alive. Questions: 0. Is it possible to have COM1, COM2, or BOTH come alive? 1. What options do I have to set when I compile my kernel? I am running make xconfig and I can't seem to find any option that turns this feature on on the GUI. Help! :-) Thanks, Peter -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: DNS problem
Anson Ng wrote: Hi, (I'm sorry for the long email, but since I've to tell situation) I've registered a domain name impko.com at Network Solutions. But since I haven't setup my DNS server and running properly, I parked it to Network solutions for the time being. I want to host the domain for my LAN, which consists of a Linux box (24.123.456.789) running www, mail, news, ftp for public access and the internal LAN (192.168.1.x). Last night, I configured a primary DNS server for impko.com, but I think I missed the reverse-IP lookup settings. Since I've only 1 external IP, i.e. 24.123.456.789, I named my Linux box serv1.impko.com and then www, mail, news, ftp, ns1 are all CNAME to it. And I got it working on my internal LAN, but I don't have a secondary. Here are my problems and questions: 1. I tried to unpark the domain from Network Solutions and it required 2 hostnames of the DNS server for the domain. So I provided them the official hostname (cr123456-a.ym1.on.wave.home.com) for the primary DNS, and for the secondary DNS, I typed in the primary DNS server hostname from my ISP since don't have another IP for the secondary. However it returned Host name invalid for both of my entries. It said the hostname must host the domain, but I already configured my own DNS. Each NS record must refer to a registered HOST. In order to register a HOST record in a domain you must own that domain. There can only be one HOST record per IP address. You need to register serv1.impko.com as a HOST with network solutions. You'll need a second NS server as well. It has to be a different machine than the first and it must also naturally be registered as a HOST. 2. Can I do it this way, my own DNS server for the primary entry and the primary DNS from ISP for the secondary? You can if they are actually going to provide DNS for your domain. 3. Since there is no DNS server on earth (except the one I host) can resolve ns1.impko.com, how can I use this DNS entry? This is what HOST records are for--so that the root servers which have the records for your domain can also give the IP addresses of the NS servers listed for it. 4. How can I restrict the resolve of the hostname, e.g. mydesktop.impko.com of my internal LAN for only internal use, i.e. not for public? But I still have to resolve the www, mail, etc. however they're in different network, 24.123.456.789 and 192.168.1.x Many sites large and small use a dual-dns configuration. You run two instances of named. You can do this because named allows you to specify that it bind to a specific IP address (or addresses). So, you run one instance which binds to the internal Ip address and one that binds to the external. In order to get this to happen in debian you'll want to either modify your existing /etc/init.d/bind or better yet copy this one to a new one called /etc/init.d/bind-internal and get it registered to start up using 'update-rc.d bind-internal' and modify this script to pass your other config script. 5. Can I setup both the primary and secondary DNS server on the same box using the same IP? No. There are organizations (or at least there used to be) of people who would secondary for each other. I don't remember any URLs for these places. Use google. 6. I use linuxconf to configure the DNS, when I add the domain to the DNS, there's a field named Main Server (the 2nd field), what should I typed in it if this is my primary DNS for the domain? I don't use linuxconf so I can't comment. 7. After I successfully host my domain, is there anything I've to inform my ISP to update? You first have to get them to agree to be a secondary DNS for you. If they agree you just need to give them your IP and enable (if you've disabled them) axfr transfers for their DNS server. I know there are too many questions, sorry. Hope you could help me on my questions, thank you. That's ok. The number and nature of your questions sound like you're in a little bit over your head. You should read man pages and RFCs. Don't expect people on mailing lists to tutor from being a newbie to being an expert in DNS administration. All this stuff isn't that hard but there are a bunch of things you need to understand to be able to set things up the way you need them to be. Just being able to run linuxconf won't be enough. In fact, you probably won't be able to use linuxconf at least not for the internal because linuxconf won't expect (I'm guessing) that you want to run two instances of named which you'll have to to have an internal version of your domain and an external. Best Regards, Good luck. Just look for and read the documentation. It's out there. Anson -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: FW: Dell PowerEdge 2450 RAID
Currently the only source for these drivers I know of is the binary ones supplied by dell. Dell has binary modules and also a kernel with support for this controller compiled into it. In order to get these you need to use the RedHat 6.2 disks which can be downloaded from the dell support site. In theory it should be not too difficult to extract the kernel out of these disks and put it onto a debian rescue disk and just use debian with the kernel they supply for redhat. I would be happy to send you the kernel and modules that work on this system. You'll have to build the debian install disks yourself. Alternatively, you can break down and run RH on the system. Kim O wrote: -Original Message- From: Kim O [mailto:[EMAIL PROTECTED] Sent: Monday, August 21, 2000 3:40 PM To: [EMAIL PROTECTED] Subject: Dell PowerEdge 2450 RAID Anyone out there know if Debian linux will support the onboard raid controllers that come in the DELL PowerEdge 2450 2U servers The RAID controllers on there are called : PowerEdge® Expandable RAID Controller 3 Single-channel (PERC 3/Si) and PowerEdge Expandable RAID Contoller 2 Dual-channel (PERC 2/DC) PERC 3/Si PERC 2/DC Single-channel Ultra3 (Ultra160) RAID controller designed to provide very low cost/high performance internal RAID to PowerEdge 2450 customers. Dual-channel Ultra-2/LVD RAID controller designed to provide low-cost/high-performance/ high availability add-in RAID solution for all PowerEdge customers -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: Transferring files between a windows machine and a Debian Linux machine
You best bet is to turn on file sharing on your windows box and then use the smbclient to access these shares in your linux box. Matt Gagné wrote: I need to transfer some program files and stuff from my windows machine (where I download the stuff) to the Linux machine (where I am learning to set it up) Can someone tell me the best way to do that? Thanks. -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: Linux client through MS Proxy Server 2.0
Doing NTLM authentication should be possible. I posit this since the samba team was able to figure out how to do it for their project. I once spent a little time looking at it but didn't get that far. If you're really determined you should probably drop some questions on one of the samba lists. CHEONG, Shu Yang [Patrick] wrote: Hi guys, Finally, I have found the answer to my problem: Acessing the Internet with a Potato box via MS Proxy Server 2.0. Here it is:- MS Proxy Server basically allows 3 types of authentication, namely : Anonymous, Basic and NTLM (also known as NT challenge/response). The first speaks for itself. The send basically require the client to send a user name and password for authentication before access is allowed...however the user name and password sent is in clear text. And finally, NTLM require the client to send a user name and password, which is encrypted, before access is allowed. The MS Proxy Server which I am trying to authenticate before the Potato box is allowed access to the Net has been set to use NTLM authentication...and here is the bombshellthe only browser (AFAIK) which can authenticate using the NTLM handshake is MS IE...what do you know..talk about cornering corporate users into using nothing but MS IE So it looks like the only place where I can tinker with Linux and other *nix is at home. Well, I am still hoping that someone knows of a way to authenticate the the MS Prxy Server using the NTLM handshake but not with MS IE. Patrick Cheong Information Systems Assurance Measat Broadcast Network Systems E-mail: [EMAIL PROTECTED] Visit us at: http://www.astro.com.my -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: java
Goeman Stefan wrote: Hello, I know, this is not really a question about Debian. You got that right. The question is actually very simple. I want to convert a string to a double. There is a method in the java.lang.Double named parseDouble When I insert a line like double d = Double.parseDouble(x); in my program, I get the error: Method parseDouble(java.lang.String) not found in class java.lang.Double This method is only in java.lang.Double since JDK 1.2. You must have an earlier version. There is also another method in the java.lang.Double class that should do the same, i.e. method valueOf Now, when I insert a line like double d = Double.valueOf(x); in the program, i get the error: Incompatible type for declaration. Can't convert java.lang.Double to double. Yup, because Double.valueOf() returns type java.lang.Double, not simply double. What you want is: double d = Double.valueOf(x).doubleValue(); in the program, (Casting does not solve this problem) Nope. There's no such thing as Double::operator double() such as there _might_ be if this were C++, which it isn't. Does anybody know what is going wrong?? Greetings, Stefan. -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: Threads and mysql
Error 11 is EGAIN (which means Try Again which generally means there was an attempt to allocate some resource which had run out at the time). There could be numerous reasons why mysql can't run more than 256 processes, e.g. because the threads share a single file descriptor table and they're all trying to open too many files. You might want to run it under strace to see what the last thing it did was before it exited with an error. George Chavdarov wrote: Hi allhave following problembox is debian/linux kernel 2.2.17pre6 with mysql 3.22.32and it seem to impossible to have more than 256 processes of mysqlerror iserror(11) this is specific OS errorany ideas George Chavdarov System Administrator - www.dir.bg -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: grep crashes machine
Frodo Baggins wrote: Christian Pernegger scripsit: Hallo! I just stumbled upon the following. If I do # cd / # grep -r * stuff it outputs files for maybe half a second then hangs the whole machine. The last matches were under /dev. I can't remember grep ever scanning device files, but maybe it's just me... are you shure? shouldn't be # grep -r stuff * ? Anyhow... It will have a problem when grep-ing never endig devices such as /dev/zero. Right, so something a little better might be find / -type f -exec grep stuff {} \; -print That way you only grep on regular files. You might even want to throw in a -xdev to keep from going through /proc. -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: J2EE SDK
Why not just convert them to debs yourself. The alien package will do this. Dominic Blythe wrote: just been checking out java.sun.com and they've got binaries of J2EE SDK for redhat 6. I presume these are .rpm, although i haven't checked it out. anybody know of any (plans to make a) .deb for these? Dominic Blythe Programmer __ BCP Ltd, BCP House, Charles St, Stockport, SK1 3JY t: 0161 335 3000 x131 f: 0161 335 3001 e: [EMAIL PROTECTED] -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: dhcp network
If your DSL router is in fact a DHCP server you can just install the pump package and run it on eth0. If your router isn't a DHCP server (or doesn't proxy DHCP) then I couldn't begin to guess how you'd get the IP address. [EMAIL PROTECTED] wrote: I am trying to hook up my new Linux box, running Storm2000 to my existing network. There already is a router (Linksys BEFSR41) used to connect my DSL modem to the network and act as a fire wall. Could you please tell me how to set the Debian Linux to get the IP address dynamically from my router. Thank YouSteve -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: 2gig file
Brian Schramm wrote: OK. I am dealing with a 2gig file size limit on Linux. I have been reading up on this all day to very little luck as to how to fix it. If I recomend someone to go to debian for an upgrade will this fix it? I am after any fix I can get and at this time changing dist. is just as good for me as anything else. Please someone help me see the lite on this. This is due to a limitation of the ext2 filesystem. I believe the reiserfs filesystem supports longer files than this but you'd need to patch your kernel source to add this. I also don't know if it's really ready for prime time yet. -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: MS Proxy with Linux
The MS Proxy Server supports the special, proprietary Microsoft Proxy Server service and also optionally supports the Socks4 protocol. The socks4 protocol must be enabled by the administrator. Once this is enabled you may use the debian socks4-client (or whatever it's called) with the run-socks script to run programs socksified. I have been successfully using this for over a year now. - Original Message - From: Timothy C. Phan [EMAIL PROTECTED] To: Debian User debian-user@lists.debian.org Cc: [EMAIL PROTECTED] Sent: Thursday, July 13, 2000 9:01 AM Subject: MS Proxy with Linux Hi, Have anyone been able to use MS Proxy Server with Linux client? Please explain the configuration of the MS proxy server as well as the Linux client box! TIA --- tcp -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null
Re: linux + wan (frame relay)
We used a Sangoma Wanpipe card in a linux box for about a year and were very happy with it. We only stopped using it because we moved our office and opted for DSL in the new space. This card can also be purchased with a built-in CSU/DSU which is what we got so other than your linux box you only need to purchase this card which we paid I think $800 US for. I would recommend this card to anyone. I would imagine that most needs would be filled with this card although there are almost certainly things a cisco can do that a linux box can't. Mario Olimpio de Menezes wrote: Hi, I would like to know if Linux (Debian/GNU) can work with wan protocols, especifically frame-relay? That's, if I buy a wan card, can I route with Linux? Should I expect some troubles? Some limitation? Can I safely substitute a Cisco router with a linux+wan card? Any kind of info is welcome; better if it's quick! :-)) Thanks, []s Mario O.de MenezesMany are the plans in a man's heart, but IPEN-CNEN/SP is the Lord's purpose that prevails http://curiango.ipen.br/~mario Prov. 19.21 http://www.revistalinux.com.br -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: SSL for smtp pop
Robert Varga wrote: On Wed, 28 Jun 2000, Jens B. Jorgensen wrote: Q.2/ Do Outlook Express 5.0 onwards Netscape 4.6 onwards offer SSL support for both pop smtp? If they do not, any other mail clients do? Netscape 4.6 does support secure pop-3 and smtp. I haven't tried the secure smtp before but I have used pop-3 over ssl and can testify that it works. How can you enable POP3-SSL in Netscape Messenger? I haven't been able to find a setting related to that, only related to TLS and SMTP-SSL... I have tried Communicator 4.7 on NT4.0. You know, I created a profile to try this myself and sure enough you're right, there's no option to enable secure POP-3. I could've sworn I used to use secure POP3 with netscape (now I just use IMAP) but I guess I must not have. Sorry for the disinformation. Eudora probably does support it if that's an option. -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: Java support? (apache + tomcat)
Oh, if you're running slink then that's the problem. Unfortunately the package doesn't exist in slink, probably because it wasn't packaged until potato was underway. The jdk package available in slink is called jdk1.1 so 'apt-get install jdk1.1' should work fine. BTW, http://www.debian.org/distrib/packages is very helpful for looking up packages. Also, I've been running potato for a while and it's pretty stable except that every once in a while a package is submitted which won't install properly. I deal with this by simply resisting the temptation to daily 'apt-get update; apt-get dist-upgrade' which it seems so many people do and then moan and groan about how this or that package doesn't work. If you're willing to put up with this occasional convenience it's worth going to potato for all those packages you'd really like to run. Gregory Guthrie wrote: At 03:32 PM 06/27/2000 -0500, Jens B. Jorgensen wrote: I know jserv is available, because I have it (1.1-3). I installed the jdk myself because I wanted 1.2.2. Are you running potato or slink? (I'm running potato). -- slink, and : alpha{root}.43: apt-get install jserv Reading Package Lists... Done Building Dependency Tree... Done E: Couldn't find package jserv and... E: Couldn't find package jdk E: Couldn't find package jsdk I want to run current java tools, and don't find any in the modules that dselect shows.. -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: SSL for smtp pop
Wilson Yau wrote: Maybe this is not quite a relevant question should be posted on this mailing list, but I'd be very grateful if anyone can give me quick correct answers to the following questions: Q.1/ I have a mail server running Debian/GNU Linux Potato 2.2 w/ kernel 2.2.15. Is there any open source utilities I can use to enable SSL support for both pop smtp? The qmail-src package provides the source for the excellent MTA qmail. See: http://www.qmail.org/top.html . SSL (read: TLS) support in qmail is available through a patch which is listed on the qmail page: http://www.esat.kuleuven.ac.be/~vermeule/qmail/tls.patch . I haven't tried to compile and use this myself. Actually if I may insert my opinion TLS is worthless for general use because you can only count on the mail being encrypt from you to your MTA. After that it will be relayed to the next SMTP server and that one probably doesn't support TLS so it'll be sent in the clear. SSL is lots easier for POP because with POP SSL is handled differently. With SMTP support is added through a special command (STARTTLS) which tells the server to start up TLS. This is necessary because SMTP always has to run on port 25 since all systems assume this. With POP, a different port is used. Standard pop-3 runs on port 110 and pop-3 over ssl uses port 995. So, if you want to run pop-3 over ssl you can use the stunnel package. The stunnel package works by just sitting in between the client and any old pop-3 server you care to run. Q.2/ Do Outlook Express 5.0 onwards Netscape 4.6 onwards offer SSL support for both pop smtp? If they do not, any other mail clients do? Netscape 4.6 does support secure pop-3 and smtp. I haven't tried the secure smtp before but I have used pop-3 over ssl and can testify that it works. Thank you for your quick help! Sure, if you want any tips on qmail or setting up secure pop-3 I'd be happy to lend my experience. -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: Help with COM port based terminal access
This is easily done by running mgetty on the port. Note that booting your computer keyboard-less can be a problem. Many systems will hang in the BIOS POST if they can't sense the keyboard is present. On some systems this can be turned off in the BIOS setup. This has nothing to do with linux. I believe you can buy little plugs which go in the keyboard connector which will make the system think there's a keyboard present. I also found that I had to modify the kernel source to get linux to boot without a keyboard but this was with a 2.0.X (where X == I don't remember what) in order to get it to work because there was some startup assembly code which was making a call to clear the keyboard buffer and it would cause the system to hang (or crash, can't remember which). Hopefully this is alleviated in more recent kernels. Serial console support should be set up in your kernel config (you're going to need to compile your own kernel) and lilo can be set up for serial console support as well. Good luck, Dinesh Nadarajah wrote: Hi: I am trying to setup several linux systems that I would like to administer over a network and/or via the serial port. The serial port based admin. is important in some cases when network is down etc. Is there any way of running a Kermit like program to connect to a linux box through COM ports and login to the system. This feature (Monitor less, keyboard less, mouse less) operation is available on IRIX, AIX and Solaris. There there a similar service I can run in linux. Thanks in advance. -Dinesh -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: Help with COM port based terminal access
Oh, ok. Now I'm with you. Sure, for this you don't want mgetty. Just use the regular getty that's already on the system. Here's what you do. First edit /etc/inittab to get init to run a getty on your chosen serial port (I'll assume for purposes of exposition you want to use the first serial port, /dev/ttyS0). I'll also assume you want to run the port at 38400bps. Add the following line to /etc/inittab: T0:23:respawn:/sbin/getty -L ttyS0 38400 vt100 Now to get init running this do 'kill -HUP 1' (init is always the first process in linux and so will always have the pid 1, the HUP signal tells init to reread /etc/inittab and act accordingly for any changes). Now the getty should be running. On your laptop you'll want to install a terminal program. I prefer seyon (which is an X app) or minicom (curses app). Hook up your null-modem (Rx and Tx are swapped) serial cable, run your terminal program, set the bps to 38400 and you should be in business. You should get a login prompt. Dinesh Nadarajah wrote: Sorry for the earler blank message. What I am trying to do is very simple actually. I would like to hookup my laptop to the seriel port of the linux machine and use a program like hyperterminal on the laptop to access the linux machine on COM1 or COM2. How will I do that. mgetty seems like a modem based access program that looks for dial tone etc. There is no info. on how to setup mgetty for this. Have sucessfully done this? Thanks a lot in advance for the advice. Regards, -Dinesh --- Jens B. Jorgensen [EMAIL PROTECTED] wrote: This is easily done by running mgetty on the port. Note that booting your computer keyboard-less can be a problem. Many systems will hang in the BIOS POST if they can't sense the keyboard is present. On some systems this can be turned off in the BIOS setup. This has nothing to do with linux. I believe you can buy little plugs which go in the keyboard connector which will make the system think there's a keyboard present. I also found that I had to modify the kernel source to get linux to boot without a keyboard but this was with a 2.0.X (where X == I don't remember what) in order to get it to work because there was some startup assembly code which was making a call to clear the keyboard buffer and it would cause the system to hang (or crash, can't remember which). Hopefully this is alleviated in more recent kernels. Serial console support should be set up in your kernel config (you're going to need to compile your own kernel) and lilo can be set up for serial console support as well. Good luck, Dinesh Nadarajah wrote: Hi: I am trying to setup several linux systems that I would like to administer over a network and/or via the serial port. The serial port based admin. is important in some cases when network is down etc. Is there any way of running a Kermit like program to connect to a linux box through COM ports and login to the system. This feature (Monitor less, keyboard less, mouse less) operation is available on IRIX, AIX and Solaris. There there a similar service I can run in linux. Thanks in advance. -Dinesh -- Jens B. Jorgensen [EMAIL PROTECTED] __ Do You Yahoo!? Get Yahoo! Mail - Free email you can access from anywhere! http://mail.yahoo.com/ -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: Java support? (apache + tomcat)
I know jserv is available, because I have it (1.1-3). I installed the jdk myself because I wanted 1.2.2. Are you running potato or slink? (I'm running potato). Gregory Guthrie wrote: I want to run current java tools, and don't find any in the modules that dselect shows.. I did get apache, but are: jserv, tomcat, jdk, jwsdk, ... available? Also, is there an easy way to ask apt-get what it knows is available? like dpks -S ? Thanks. Gregory Gregory Guthrie [EMAIL PROTECTED] (515)472-1125Fax: -1103 -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: REPOST: ypserv consumes giant amounts of memory???
^^-^^ -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: postfix or sendmail tempting
What you should do is not mess with scanning /var/spool/mail/X. Instead install procmail. With procmail you can set up a script which is called each time when the messages is received (and not until the whole thing is there!). [EMAIL PROTECTED] wrote: Hi all! I've got the job to make accepting mails to an e-mail address a bit unusual. What's unusual. I will explain. When a mail arrives to the address it shouldn't appear in the /var/spool/mail immediately. Because there is daemon which scans the mailbox all the time and as soon as it sees a mail arriving it takes it from the box and passes it to a program which works on it and then sends the output to a database. But when the mail starts to arrive the system starts to write it to the disk immediately. ..And the daemon takes of , let's say the first 15 bytes (From, mal client etc.) but not whole letter. That's the problem because the body contains the datas to the program. The only solution I see is to move it temporarly to a /temp box or directory and as soon as the whole mail arrived pass it to the accurate mailbox so the daemon could catch it to delivery(the whole mail) to the program. Oh, yes, and all the headers have to be untouched so the From and Reply to and Envelope _must_ be the same as they arrived. The ASCII art to this (as I imagine) : mail to [EMAIL PROTECTED] -- temporary box/dir(until closed) -- -- [EMAIL PROTECTED] /-- daemon -- program -- DB2/ How can I do it with postfix (preferable) or sendmail or any other sendmail compatible MTA RTFMs and HOWTOs are welcomed. Thanx, Ago -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: simple dhcp questions...
That is weird. The option routers line is indeed supposed to set the default gateway. This works fine for me. I will say though that if your laptop is actually running WinNT and you have two ethernet interfaces (one pcmcia, one in the docking station) this is pretty much standard behavior. You have to release the one you're not using using NT's ipconfig command. Jonathan Lupa wrote: [potato, dhcp.deb, linux 2.2] I am setting up a dhcp server so I can connect my laptop to my home system and always get an expected setup while getting a different setup whenever I plug it in at work. Sounds simple. :) I'm having 2 problems detailed below. I'm trying to get the following: ip: 192.168.2.4, netmask: 255.255.255.0, bc: 192.168.2.255, dns: 192.168.2.1, gw: 192.168.2.1, netbios: 192.168.2.1 To that effect I use the following... ==dhcpd.conf== option domain-name JJLNet; option domain-name-servers gw.JJLNet; default-lease-time 600; max-lease-time 7200; subnet 192.168.2.0 netmask 255.255.255.0 { range 192.168.2.10 192.168.2.100; option broadcast-address 192.168.2.255; option routers gw.JJLNet; } host noghri { hardware ethernet 00:50:8B:9D:AF:A7; fixed-address Noghri.JJLNet; option routers gw.JJLNet; option host-name Noghri; option domain-name JJLNet; option all-subnets-local 1; option broadcast-address 192.168.2.255; option netbios-name-servers gw.JJLNet; } = And get back this: == gw:/etc# init.d/dhcp restart Internet Software Consortium DHCP Server 2.0 Copyright 1995, 1996, 1997, 1998, 1999 The Internet Software Consortium. [...] /etc/dhcpd.conf line 24: expecting identifier. option all-subnets-local 1; ^ Configuration file errors encountered -- exiting exiting. == Which is odd because of this: ==man dhcp-options== option all-subnets-local flag; This option specifies whether or not the client may assume that all subnets of the IP network to which the client is connected use the same MTU as the subnet of that network to which the client is directly connected. A value of 1 indicates that all subnets share the same MTU. A value of 0 means that the client should assume that some subnets of the directly connected network may have smaller MTUs. OK, so I get rid of the offending line, and everything SEEMS ok until I actually connect. Everything seems to be cool except that my default IP gateway is not getting set. Isn't that supposed to be the routers line? (The ip gateway is still set to the gateway on the network at work)... Thanks! Jonathan -- [EMAIL PROTECTED] GPG public key available from http://lupavista.jamdata.net/gpg.asc -- Lament 1750: If I only had a radioactive decay source and a fast free-running oscillator... -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: https / SSL question
Andrew Kae wrote: Hello, I'm hoping someone has had experience w/ HTTPS / SSL. There seems to be very little documentation on HTTPS. I just have a few questions: 1) Are all socket connections in HTTPS (specifically version 1.1) done through SSL? Yes (technically, the protocol now specified is TLS, the successor to SSL). 2) In HTTP 1.1 (not 1.0) you can send multiple GETs through the same socket connection. Can you do the same in HTTPS (1.1)? Naturally. HTTP is transparent through the SSL connection (see http://www.isi.edu/in-notes/rfc2818.txt). Whether or not multiple requests can be made depends on the server and client supporting this through the 'Connection' header. 3) Does anyone know which versions of browsers HTTPS is compatible with such as IE or Netscape? Hmmm, hard to say. I think both have support https for quite a while. -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: samba auto-unloads
smbd and nmbd are only used for serving linux filesystems to other machines. That's not the problem. I have noticed that the newer versions of smbmount let the connection time out and get closed. Brian Stults wrote: Hello, I've been using the samba client to mount NT drives to my linux box for some time without any problems. Now, on my home linux box, I'm mounting drives from my linux box at work. It works very well at first. However, when I come back the next day, the mounted drive produces I/O errors. It turns out that the smb and nmb daemons have automatically unloaded from my work box. When I login to the remote machine and restart samba, the drives can be mounted again. Why does this happen? Is it related to the daemons versus inetd issue? I don't fully understand this. There are entries in my inetd.conf like this: #:OTHER: Other services #off# netbios-ssn stream tcp nowait root/usr/sbin/tcpd /usr/sbin/smbd #off# netbios-ns dgram udp waitroot/usr/sbin/tcpd /usr/sbin/nmbd -a But it appears they are commented out. I also have the standard startup script in /etc/init.d for samba. I sifted through the SMB-HOWTO, but couldn't find reference to this. Can someone help? Thanks. -- Brian J. Stults Doctoral Candidate Department of Sociology University at Albany - SUNY Phone: (518) 442-4652 Fax: (518) 442-4936 Web: www.albany.edu/~bs7452 -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: netdate missing in potato?
There's ntpdate which as the name suggests uses ntp and there's also rdate now which utilizes the simpler FRC868 protocol. Sebastian Canagaratna wrote: I don't seem to find netdate in potato. I am pretty sure it was there in slink. Is there any replacement for it? Sebastian Canagaratna Department of Chemistry Ohio Northern University Ada, OH 45810 -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: FW: IMAP + QMAIL
Did you make sure to compile courier-imap with PAM support? I don't remember from when I compiled it but you probably have to specify this. This is what I'm using. Your AUTHMODULES should contain authpam. It would seem that authshadow should've worked but it wouldn't if you're using md5 passwords (this is specified when you install the box). Once you've got authpam in there you'll need to add an /etc/pam.d/imap file if it doesn't exist. Mine is just: authrequiredpam_unix_auth.so shadow nullok accountrequiredpam_unix_acct.so sessionrequiredpam_unix_session.so Give that a try. If you want to test in a more controlled way you can telnet to the server and talk IMAPV41 directly. Telnet to it from the machine it's running on with: telnet localhost 143. You'll get a response like: * OK Courier-IMAP ready. Copyright 1998-2000 Double Precision, Inc. See COPYING for distribution information. Now, as for the basics, each IMAP command is tagged so the first token just needs to be some tag. The command we want to test is the 'login' command which takes the login name and password as extra tokens. So, you'll type in: 1 login user password and if all is good you'll get the response: 1 OK LOGIN Ok. But if the authentication fails you'll get back: 1 NO Login failed. This will tell you if it's really working or not since with Outlook you never know what the [EMAIL PROTECTED] it's doing. You can then just logout with: 2 logout Try these things out and let me know. Wesley A. Wannemacher wrote: Okay, I have all the test users to Maildir/ format, I also installed the courier-imap package that you suggested. But I still cannot seem to get the email to work properly. I tried the authshadow and the authpwd AUTHMODULES in the /usr/lib/courier-imap/etc/imapd.config file, nothing has worked. I will outline the basic steps below of what I have done and I will try to (conservatively) paste the contents of some of the config files for qmail. I think I must be missing something. Any help is greatly appreciated. 1. I downloaded and installed the following packages - qmail - daemontools 0.70 - ucspci 0.88 2. I configured everything and qmail appears to be sending/recieving email - I followed all the TEST.* files that came with the qmail source everything checked out. 3. I downloaded and built courier-imap - every time I try to connect with Outlook/Outlook Express I get error: There may be server or network problems or your timeout interval may be too short. Account:X.X.X.X Protocol: IMAP Server: X.X.X.X Port: 143 Where X.X.X.X is the IP address of the potential IMAP server. Are there any other tests I can run? What other output would be beneficial? courier- imap did not complain when it was build, nor does it complain when it runs. Wes Wesley Wannemacher [EMAIL PROTECTED] Network Administrator, Instructor Northwestern College http://www.nc.edu -Original Message- From: Jens B. Jorgensen [mailto:[EMAIL PROTECTED] Sent: Tuesday, June 13, 2000 11:00 AM To: Wesley A. Wannemacher Cc: debian-user@lists.debian.org Subject: Re: FW: IMAP + QMAIL That's a strong possibility. If you're using shadow and WU uses getpwnam rather than getspnam, it won't get the password and won't be able to authenticate. Does WU IMAP support PAM authentication? That would make your life easier if it can be built with PAM support since then the actual authentication checking is done is shared libraries. I've been a longtime qmail (with maildirs) user and I highly recommend the use of maildirs rather than the old mailbox files. The only imap server I've gotten to work satisfactorily for me is courier-imap. This is a maildir-only imap server which I found naturally enough on the qmail web site. The link is: http://www.inter7.com/courierimap/ Wesley A. Wannemacher wrote: I recently rebuilt a Linux box here at our office, I am trying to set the machine up to do email for us. I have successfully installed qmail, now I would like to get WU IMAP up and running. The software is building just fine, I even hacked the code to make it read the ~/Mailbox instead of the /var/spool/mail/$USER file. Now everytime I try to connect to the server with an IMAP client, I cannot authenticate. I think this has something to do with shadow, but I am not completely sure. Any help is appreciated. Wes Wesley Wannemacher [EMAIL PROTECTED] Network Administrator, Instructor Northwestern College http://www.nc.edu -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED] -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: ipchains latency
10-30 seconds for telnet? Even on a 386/33 this is just way too much delay to be accounting for in packet filtering rules. I would suspect something else, like ident checking which is waiting to time out and reverse-dns lookups timing out. Often these two things are used to gather info to log about who's getting into your machine. If they aren't there it can take a while for the lookup to time out. Check into that. Chris Brown wrote: Hello, Our company LAN is divided into two segments, and I have just finished implementing firewalling rules for the router in between them, to protect the inner network from the outside world. After meticulously designing an installing my ipchains rules, I was dismayed by the performance hit they incurred. Before installing the firewalling rules, connection latency between the networks was normally below ~50ms. telnet, ftp, and other logins took less than a second to return a login prompt. Now, after installing the rules, a connection across the firewall takes at least 10sec, occasionally taking over 30sec. Once the login is successful, latency isn't too bad, but still noticably worse - well over 200-300ms - when in a telnet session. The router is a 386/33 with 16MB of RAM and two ISA Ethernet cards. Is this an underpowered machine for firewalling? I shouldn't think this is the problem... Are there any errors that add to connection latency that I should be looking for in the firewalling rules? Thanks, Chris Brown [EMAIL PROTECTED] Seitz Technical Products Inc. * Chris Brown [EMAIL PROTECTED] !!! HELP FIGHT SPAM !!! Join; www.cauce.org See; spam.abuse.net, spamsucks.com, www.cm.org -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: FW: IMAP + QMAIL
That's a strong possibility. If you're using shadow and WU uses getpwnam rather than getspnam, it won't get the password and won't be able to authenticate. Does WU IMAP support PAM authentication? That would make your life easier if it can be built with PAM support since then the actual authentication checking is done is shared libraries. I've been a longtime qmail (with maildirs) user and I highly recommend the use of maildirs rather than the old mailbox files. The only imap server I've gotten to work satisfactorily for me is courier-imap. This is a maildir-only imap server which I found naturally enough on the qmail web site. The link is: http://www.inter7.com/courierimap/ Wesley A. Wannemacher wrote: I recently rebuilt a Linux box here at our office, I am trying to set the machine up to do email for us. I have successfully installed qmail, now I would like to get WU IMAP up and running. The software is building just fine, I even hacked the code to make it read the ~/Mailbox instead of the /var/spool/mail/$USER file. Now everytime I try to connect to the server with an IMAP client, I cannot authenticate. I think this has something to do with shadow, but I am not completely sure. Any help is appreciated. Wes Wesley Wannemacher [EMAIL PROTECTED] Network Administrator, Instructor Northwestern College http://www.nc.edu -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: xrdb
Rather than creating your own xsession file try editing /etc/X11/Xsession.options and make sure it has the line: allow-user-resources Then rename your .Xdefaults to .Xresources and then you can dump your .xsession. All should work fine then. Goeman Stefan wrote: Hello Everybody, I have put some things in a file .Xdefaults. I know that in order to make these settings active I have to do xrdb .Xdefaults. This is no problem. The problem is that I want this command to be executed automatically when I login in a gnome session. Therefore I have created the file .xsession as follows (PS: I use gdm): xrdb $HOME/.Xdefaults exec gnome-session This does not seem to work. Also, I have tried to put the xrdb $HOME/.Xdefaults line into the .bash-profile (or .bashrc) file. Now I first have to start a terminal window in order to activate the settings in .Xdefaults. My question is, where do I have to put xrdb $HOME/.Xdefaults in order to have this executed when I log in a gnome session ? Greetings, Stefan Goeman. -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: Question about a module ne
This happens because the ne modules relies on the 8390 module. It's best to use modprobe instead of insmod since it will resolve these dependencies for you. Alternatively, you can just first say 'insmod 8390' and then your insmod line. If you were trying to setup this card during installation of the operating system and it wouldn't take the options you passed on the command line this is because of a long-standing bug which was there in Hamm and I think even Bo. You'd think they'd fix it huh? When I got bit by this the second time it'd been so long I didn't even remember the problem. The solution in that scenario is indeed to exit to the command line and manually insert the module. If you're interested, the problem is that the modconf progam seems to add the file to /etc/modutils as it should but the either doesn't run update-modules to update /etc/modules.conf or update-modules fails. Anyway this should be enough information to get you going and then some. S enz Vargas Juan Carlos wrote: Yes, i have also used that: insmod ne io=0x280 irq=11 and the error message is: /lib/modules/2.2.12/net/ne.o: unresolved symbol ei_open /lib/modules/2.2.12/net/ne.o:unresolved symbol ethdev_init /lib/modules/2.2.12/net/ne.o unresolved symbol ei_interrupt /lib/modules/2.2.12/net/ne.o: unresolved symbol NS8390_init /lib/modules/2.2.12/net/ne.o: unresolved symbol ei_close ? I really dont know what that means. Thanks for your help. Juancasa On Tue, 6 Jun 2000, Ron Flory wrote: S enz Vargas Juan Carlos wrote: I am new to linux and i am installing my ne200 compatible net card. I dont know how to enter the parameters for this module to work. I have tried insmod ne i/o=0x280 irqval=11 did you use insmod ne io=0x280 irq=11 instead of what is shown above? are you certain your card is at this address and irq? I have seen printer ports and other cards interfere with ne2000 because the autoprobe has to scan such a wide area. somethimes this type of card is easy to sense, other times they can be very stubborn since there are so many 'clones' out there. ron -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: OT: dial-up - leased line
Yes, leased line signalling is different. You have to have a modem which supports leased line operation. I remember in the past an external modem I had which could do this. I don't recall the brand/make. Lehel Bernadt wrote: Hi folks, I'm thinking on switching from a dial-up connection to a leased line. Is there anything special my modem (3com usr 56k ext) has to know to do this (ie. to deal with the differences between a normal telephone line and the leased one) ? TIA, Lehel -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null
Re: Raid setup
This is a bigger issue than you surmise. RAID collections are activated after the system boots because information about the RAID set has to be fed to the RAID drivers. It's possible someone has devised a way of feeding this information to the kernel at boot time. These problems can be alleviated by a hardware raid controller. Jay Kelly wrote: Hello Group, I want to setup my potato to run a raid(mirror everything. I have a drive I took out of a win nt machine thats formated as ntfs. Once I install it in the linux box what will I need to do to get it running? I have never installed a second drive in Linux. How will I format it? Will I need to compile the kernel again? Thanks -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null
Re: dhcp
dhcp-client is indeed the package you want. Post the error messages you get during install. In order to run a gateway for your external interface you'll want to set up IP Masquerading and then have all your other boxes point to your linux box as their default gateway. In order to get IP Masq. going you'll need to get your kernel loading the masq. modules and then configure the masquerading using ipfwadm or ipchains depending on your kernel. Please look at the IP Masquerading HOWTO for details: http://www.linux.org/help/ldp/howto/IP-Masquerade-HOWTO.html John Plummer wrote: Hi, Sorry to bother you with another newbie question but hours searching man, info, HOWTO's and FAQ's wasn't too revealing. I have a PC with both a Netgear and a 3Com Etherlink II nic cards. The Netgear is attached to a cablemodem and uses DHCP. The 3Com is fixed IP for the internal LAN. Both work under NT but I'm trying to switch to Debian. Loaded ver. 2.1 from the CD with dselect settings stable main contrib. Added the 3Com during installation and later using modconf added a tulip module for the Netgear. This was implied from an email thread about the Netgear NIC. The problem is first how to get DHCP up and running so as not to have to switch back to the dark side to browse and do email. Running dpkg --list 'dhcp*' listed a few dhcp entries, e.g. dhcp, dhcp-client, dhcp-beta, dhcpd, etc. But both apt-get install dhcp-client and using dselect (probably incorrectly) erred out. Can anyone point me to some docs on doing this? The second needed HOWTO would be setting up the gateway for the LAN through the 3Com. At this point, even being pointed to the appropriate info, man or HOWTO would be appreciated. First brush with OS has been impressive. Brings back the feeling of potential felt when discovering Sun's OS with Openwindows years ago. Except this time the pieces are there and affordable. jcp -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: Transfer data between two comps without network
I highly recommend you check out the Ecra VXA-1. Check out www.ecrix.com. I've got one on order right now. It's also Linux certified (I forget who does that Linux hardware certification thingy but they did it). The specs were more than enough to get me. They have a special offer right now for first-time US buyers (sorry Dariush). Here's the excerpt from my VXAdata newsletter: SPECIAL OFFER FOR FIRST-TIME U.S. BUYERS Now you can evaluate the VXA drive -completely- free for 30 days. And, if you decide to keep it, get big savings. Internal Kit, only $399, External Kit, only $499. Offer good only in the US, until May 31, by going to: http://www.ecrix.com/eval I'm thinking of finally getting a nice tape drive for home. Dariush Pietrzak wrote: Welcome, my problem is that I have to transfer large amount of data (20~50 Gigs) daily. And it can't be done via network due to 'secret' nature of that data. I considered IDE disk put in hot-swap bay, but I found that's not the best way to do that: i got system on scsi disc, compiled ide-disk support as module and when I want to remove ide-disc i unmount it, rmmod the module then swap the discs, modprobe ide-disk, mount it. That scheme works ... but sometimes it fails.. and when it fails I have to reboot the system to be able to mount ide disc. that situation is unacceptable. Does anyone have any suggestions on this? (data should be moved via some physical way, not using network as that's what bossess fear the most, zip drives could be nice, but they B are too small, streamers seems to be to slow ) regards, Dariush Pietrzak -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: file transfers to Win2K
Additionally you could use smbclient. This allows batch transfers as well and you wouldn't have to run an FTP server on the target machine. [EMAIL PROTECTED] wrote: Hiya what would be the best way to routinely (ie. daily) transfer 500MB worth of image files from a linux system to a Windows 2000 or NT? would ftp be the best choice or is it prone to bad behaviour if unsupervised? Zane -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: moving filesystem
The lost+found directory is created for use by fsck. When fsck runs if it finds orphaned fragments and things like that it will put them into lost+found, creating a file for each cohesive chunk. So, you probably don't need to copy it from /home to /images. [EMAIL PROTECTED] wrote: Hi i am moving /home to an existing file system, /images, either by : cp )a /home /images or find /home )xdev | cpio )padm /images My concern is the lost+found directory... should i retain it on the /images filesystem or delete it or what? thanx Zane -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: moving filesystem
Oh, well I wouldn't worry too much about that. Just let it copy and then delete the directory afterwards. If there's already a lost+found in the target directory then don't sweat it. Just let the thing copy it. It won't make a difference. [EMAIL PROTECTED] wrote: then how would i refrain from copying the lost+found file whilst using the cp -a or cpio commands? Jens B. Jorgensen [EMAIL PROTECTED] on 27/04/2000 04:05:47 To: Zane Drysdale/Diagnostic Labs/64 cc: debian-user@lists.debian.org Subject: Re: moving filesystem The lost+found directory is created for use by fsck. When fsck runs if it finds orphaned fragments and things like that it will put them into lost+found, creating a file for each cohesive chunk. So, you probably don't need to copy it from /home to /images. [EMAIL PROTECTED] wrote: Hi i am moving /home to an existing file system, /images, either by : cp !)a /home /images or find /home !)xdev | cpio !)padm /images My concern is the lost+found directory... should i retain it on the /images filesystem or delete it or what? thanx Zane -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED] -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: ssh passphrase
Well, I know what you're talking about with the agent sticking around. I don't have any solution for you but this certainly sounds like something that's needed. Russell Coker wrote: On Wed, 12 Apr 2000, Jens B. Jorgensen wrote: That's what ssh-agent is for. You run ssh-agent and it will output environment variable for a unix domain socket. Then you run ssh-add and type in your passphrase. The ssh-agent caches your key and access is limited to your user (permissions on the unix socket). This is not secure enough for some of course. Thanks Ben and Jens for your advice on this issue. I have now got ssh-agent working with support for X and non-X logins (/etc/profile checks whether $DISPLAY is set to determine which alias to setup for ssh-add). Now I have a problem though, sometimes a session gets killed without the .logout running and the ssh-agent keeps running. This is a problem as the machine in question could potentially be accessed by an untrusted person and the ssh-agent contains the root password. What I would like to do is have the ssh-agent timeout after some time of inactivity and/or a specified period of time. Another thing I would like to do is have a password get removed from the ssh-agent after a period of time. Has anyone worked on any of these issues? Does anyone have any code that may help? If no-one else has done any of this then I intend to write some support for this myself. Russell Coker Russell Coker wrote: Is it possible to have the ssh client read the pass-phrase for an authorised key from an environment variable? What I want to do is: export PASS=`ssh-askpass` for n in $MACHINES do ssh $n command done unset PASS Or something similar. Basically I want to login to 30 machines and run some command but without having to enter my pass-phrase 30 times. I know I could use expect (and will if no-one has a better suggestion). But I'm sure there is a better way (why else would ssh-askpass exist?). -- My current location - X marks the spot. X X X -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: ssh passphrase
That's what ssh-agent is for. You run ssh-agent and it will output environment variable for a unix domain socket. Then you run ssh-add and type in your passphrase. The ssh-agent caches your key and access is limited to your user (permissions on the unix socket). This is not secure enough for some of course. Russell Coker wrote: Is it possible to have the ssh client read the pass-phrase for an authorised key from an environment variable? What I want to do is: export PASS=`ssh-askpass` for n in $MACHINES do ssh $n command done unset PASS Or something similar. Basically I want to login to 30 machines and run some command but without having to enter my pass-phrase 30 times. I know I could use expect (and will if no-one has a better suggestion). But I'm sure there is a better way (why else would ssh-askpass exist?). -- My current location - X marks the spot. X X X -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: NIC
You need to add: ether=0,0,eth1 to your kernel params. If you're using LILO to boot you can add the line: append = ether=0,0,eth1 after the 'image=' line for your kernel (there are lots of these in Corel, so just add the 'append = ' line after each of them. Chris Mason wrote: What's the easiest way to add a second NIC to my Corel Linux box. The first one was auot-installed during the installation of the OS, but now I have to add another PCI NIC to allow me to setup a firelwall. THanks Chris Mason Box 340, The Valley, Anguilla, British West Indies Tel: 264 497 5670 Fax: 264 497 8463 USA Fax (561) 382-7771 Take a virtual tour of the island http://net.ai/ The Anguilla Guide Find out more about NetConcepts www.netconcepts.ai bwz*mq -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: Compaq 5300
I'm not sure quite what you mean. Do you mean you don't have any room on your hard disk to install linux? You'll need to have some room because even the base debian is to small to run off a single floppy. If you don't want to create a new partition for linux (which is the standard way of doing things) I think you can still do it using UMSDOS which allows you to create your linux system within an existing DOS filesystem. I've never done this and couldn't give you any more details about this. Did you take a look at this web page: http://www.debian.org/releases/slink/i386/install ? The Millers wrote: Dear Wonderful Deb_Ian People, I would like to move forward with exploring a Linux system using a Compaq LTE 5300 laptop. The method by which I can load the kernel/OS is limited to diskette. What would be the best way way to proceed from this point using Debian's Linux? Thank You Very Much for Your Time, Jerry Miller -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: Trying to autostart mySQL
Are you sure 3 is the default runlevel? Check: grep default /etc/inittab My debian system defaults to runlevel 2. That would explain your system not starting. Chris Mason wrote: I'm trying to sutostart mySQL at boot up on my Corel Linux machine. I have put the mysql.server script in /etc/init.d and I have aput a symbolic link to it in /etc/rc3.d name S90mysql but it doesn't run on boot. I can start it by typeing /etc/rc3.d/S90mysql start no problem. Any ideas? Chris Mason Box 340, The Valley, Anguilla, British West Indies Tel: 264 497 5670 Fax: 264 497 8463 USA Fax (561) 382-7771 Take a virtual tour of the island http://net.ai/ The Anguilla Guide Find out more about NetConcepts www.netconcepts.ai bwz*mq -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: Trying to autostart mySQL
runlevel is a parameter to init. init is the first (and only) process started directly from within the kernel itself. Starting init is the last thing the kernel does. You can pass a parameter to the kernel for it to pass to init, which is the runlevel. If no runlevel is passed then init relies on the default runlevel in /etc/inittab. init also uses /etc/inittab to decide what other things to run (which includes all the rc scripts). Chris Mason wrote: Bingo! That did it. What is runlevel? Chris Mason Box 340, The Valley, Anguilla, British West Indies Tel: 264 497 5670 Fax: 264 497 8463 USA Fax (561) 382-7771 Take a virtual tour of the island http://net.ai/ The Anguilla Guide Find out more about NetConcepts www.netconcepts.ai bwz*mq -Original Message- From: Jens B. Jorgensen [mailto:[EMAIL PROTECTED] Sent: Thursday, March 30, 2000 11:03 AM To: [EMAIL PROTECTED] Cc: Debian Users Subject: Re: Trying to autostart mySQL Are you sure 3 is the default runlevel? Check: grep default /etc/inittab My debian system defaults to runlevel 2. That would explain your system not starting. Chris Mason wrote: I'm trying to sutostart mySQL at boot up on my Corel Linux machine. I have put the mysql.server script in /etc/init.d and I have aput a symbolic link to it in /etc/rc3.d name S90mysql but it doesn't run on boot. I can start it by typeing /etc/rc3.d/S90mysql start no problem. Any ideas? Chris Mason Box 340, The Valley, Anguilla, British West Indies Tel: 264 497 5670 Fax: 264 497 8463 USA Fax (561) 382-7771 Take a virtual tour of the island http://net.ai/ The Anguilla Guide Find out more about NetConcepts www.netconcepts.ai bwz*mq -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED] -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: which driver for Netgear FA310TX
I used the one that came with the card (actually, I downloaded it from the web site) and that worked best for me. matt garman wrote: Hello: I've been having some problems setting up my home network, and I was wondering if my card is not working correctly. I have a Netgear FA-310TX, revision D2. Which driver is best for this card: the linux/tulip.c driver that comes on the FA-310tx's drivers disk, or the new tulip driver as supplied by linux kernel 2.2.14 or the old tulip driver (also in linux kernel source)? Thanks again, Matt -- Matt Garman, [EMAIL PROTECTED] I was just reading the interview with Korn in _Guitar_World_, and one of the guitarists said they don't play guitar solos because they've been done. Well, I guess that's true if you stick with what's been done. But you have to look beyond that; there's a lot more left to say on the guitar. -- Warren Haynes of Gov't Mule -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: MS Proxy
M$ Proxy Server uses a proprietary (surprise!) protocol for the regular M$ proxy server clients. However, Proxy Server also supports Socks4 but this has to be turned on (I think by default it isn't turned on). With socks4 you can use the proxy server just fine, in Netscape you can set this in the proxy preferences, with other command-line stuff you can use runsocks. Install the socks4 client package and read the man pages. David Gisborne wrote: I am stuck behind a Microsoft Proxy Server. Is there any way I can access the internet from behind the proxy? Thanks -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: funky samba stuff
Whoa. This sounds like a bug I ran into long ago. I got it when I upgraded to potato but kept my 2.0.X kernel. The problem has to do with a library call samba uses to set gid (just like the log suggests). Does this sound like your situation? I fixed it by modifying the source. The configure script checks for the existence of this library call and uses it if it exists. It did on my potato system but didn't work. I fixed it by munging the source after configure was run. It would have been easier to just run 2.2.X but sometimes you just have to go the hard way. Hamish Moffatt wrote: I just upgraded a linux server here from hamm to potato. Some funky messages from samba; smbclient -L reggie -U hamish says tree connect failed: 0 after I enter my password. In the log, [2000/03/23 00:45:51, 1] smbd/server.c:main(643) smbd version 2.0.6 started. Copyright Andrew Tridgell 1992-1998 [2000/03/23 00:45:51, 1] smbd/files.c:file_init(216) file_init: Information only: requested 1 open files, 246 are available. [2000/03/23 00:45:53, 0] lib/util_sec.c:assert_gid(72) Failed to set gid privileges to (-1,1000) now set to (0,0) uid=(0,0) [2000/03/23 00:45:53, 0] lib/util.c:smb_panic(2456) PANIC: failed to set gid Any ideas what this means and how I fix it? It's running from inetd. Hamish -- Hamish Moffatt VK3SB [EMAIL PROTECTED] [EMAIL PROTECTED] -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: is there an HTTP C library?
Try libwww. You can get it from: http://www.w3.org/Library/ [EMAIL PROTECTED] wrote: On Thu, Mar 16, 2000 at 12:38:44PM -0800, Sean 'Shaleh' Perry generated a stream of 1s and 0s: On 15-Mar-2000 [EMAIL PROTECTED] wrote: What C/C++ libraries are available to get web pages from within a C Know nothing about it, but I bet the Apache project has something like that. -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: SMTP password
Oswald Buddenhagen wrote: Hello! I have a problem: my ISP ask me for username and password to send e-mail by smtp. With Netscape is easy to configure it, but I don't know how to do that with smail or sendmail. Can somebody help me? the problem is: smtp has NO password at all (AFAIK). the point is, you have to do POP before SMTP to authorize yourself, which netscape does normally. This is not true. SMTP can have the AUTH extension to allow user/password authentication. This mechanism is used by many ISPs and is implemented in mail SMTP servers (including recent sendmail versions). when using an mta (e.g. sendmail) you have to do a fetchmail before sending any e-mail. on one box i configured i simply put a fetchmail in the ppp-up script, that is, fetchmail is called directly after dialing in. a problem arises, if you stay on-line that long, that your authorization times out - then you would have to fetchmail before sending e-mail again. but i thing it is a good idea to run fetchmail as a daemon as long as you are on-line, so it checks for mail and holds your smpt-authorization open every, say 10, minutes. in short: put a fetchmail -d 600 in your /etc/ppp/ppp-up put a fetchmail --quit in your /etc/ppp/ppp-down (i'm not sure if the files are in /etc/ppp/ on debian, as i set up a suse box.) P.S. Sorry, terrible English. not that bad - at least not worse than mine :-) -- Hi! I'm a .signature virus! Copy me into your ~/.signature, please! -- Linux - the last service pack you'll ever need. -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: samba
Timothy C. Phan wrote: Hi, I did a grep usera /etc/samba/smbpasswd and the usera is in this file. However, this smbpasswd file has the 0700 permission, is this the problem? Probably not, it needs to have this permission. Who is the owner of the file. The usera also try to change the samba password using smbpasswd and here is the error message: read_socket_with_timeout: timeout read. read error = Connection reset by peer. machine 127.0.0.1 rejected the session setup. Error was : code 0. Failed to change password for usera This sounds like you've got a problem with /etc/hosts.allow. You're connecting to the service and then it's dropping you. Make sure that /etc/hosts.allow is set to allow you to connect to the netbios-ssn service. -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: ppp: frame with bad fcs
[EMAIL PROTECTED] wrote: ...something munging bytes on the way through, is what I assume. I can load the same pages on my Win95 laptop with no problem with the same ISP, so it appears to be Linux somehow that is doing this. Last night I recompiled the kernel (a potato system, I had been using the default 2.2.14 kernel), and this seems to have relieved the problem somewhat -- stuff downloads very slowly from the problem sites, but I don't get the timeouts that I was getting. No error messages are now evident. Maybe there aren't error messages but do an ifconfig on the interface and take a look at the error counts there. My guess is you're still getting the errors. You need to look closer at configuring your modem more likely than trying to do stuff with the software. Maybe I need to examine the packets that are getting sent over to me? It would be good to know exactly what bytes are screwed up but this would be very difficult to ferret out. Probably not worth the time. fcs is Frame Check Sequence. PPP over an async link uses HDLC-like framing for each packet. The FCS is used to detect errors in packets which are received. Such an error could happen on a link which is not error-corrected. On a link which is error-corrected this probably indicates something is munging bytes on the way through. This could happen because of things like software flow-control, which inserts XON/XOFF bytes into the stream to tell the other side to stop sending data. Make sure your modem is configured to only do hardware flow control. [EMAIL PROTECTED] wrote: I am having trouble recieving files via PPP from certain sites. I checked the debugging log, and I am repeatedly getting a two-line error message from those sites: kernel: ppp: frame with bad fcs, length = 994 kernel: ppp: bad frame, count = 994 Can anyone tell me what this means? It seems that my PPP connection receives a bad frame, and keeps requesting that same frame over and over again (I get the same message repeated throughout the log). Thanks, Steve Martin -- Jens B. Jorgensen [EMAIL PROTECTED] -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: samba
You need to add your user on the potato box to the smbpasswd file if it isn't there. As root, say: grep myusername /etc/samba/smbpasswd If your user isn't in there, then run: smbpasswd -a myusername If it is in there you can just run as yourself, the user whose home directory you wish to access: smbpasswd The reason you have to give your password again is that like unix and Lan Manager (Windows Networking, whatever you want to call it) use different hashing algorithms for passwords. Timothy C. Phan wrote: hi, I have just install samba on my potato box. Is there an addition setup that I need to do to browse my home directory on the potato box? Here is the sample setup: On potato box: user login: usera passwrd : passwda home dir : /home/usera On NT box: I also have the same use and same password. Now, from the NT box, I'd like to map this home directory on the potato to a drive where I can read/write file. Anything in the /etc/samba/smb.conf file need to be modified to do this? Currently, everytime I click on the workgroup on the NT explore, it kept asking me for a user/passwd and I kept getting invalid user/passwd. Thanks --- tcp -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: ppp: frame with bad fcs
fcs is Frame Check Sequence. PPP over an async link uses HDLC-like framing for each packet. The FCS is used to detect errors in packets which are received. Such an error could happen on a link which is not error-corrected. On a link which is error-corrected this probably indicates something is munging bytes on the way through. This could happen because of things like software flow-control, which inserts XON/XOFF bytes into the stream to tell the other side to stop sending data. Make sure your modem is configured to only do hardware flow control. [EMAIL PROTECTED] wrote: I am having trouble recieving files via PPP from certain sites. I checked the debugging log, and I am repeatedly getting a two-line error message from those sites: kernel: ppp: frame with bad fcs, length = 994 kernel: ppp: bad frame, count = 994 Can anyone tell me what this means? It seems that my PPP connection receives a bad frame, and keeps requesting that same frame over and over again (I get the same message repeated throughout the log). Thanks, Steve Martin -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: nmbd refuses connection
What machine is 192.1.1.25? Is that the machine gateway? If so then is nmbd actually running? It might be configured to run from inetd.conf if it is then check out /etc/hosts.allow and make sure that it's setup to allow connections. I'm guessing nmbd probably isn't running for whatever reason and isn't set up to run from inetd.conf. It could be possible you've got a bad smb.conf file. Patrick Kirk wrote: I can't get samba to actually share. Even on the server I get, this: gateway:/etc/samba# smbclient -L gateway Client started (version 2.0.5a). resolve_name: Attempting lmhosts lookup for name gateway0x20 Connecting to 192.1.1.25 at port 139 error connecting to 192.1.1.25:139 (Connection refused) Connection to gateway failed gateway:/etc/samba# Anyone recognise this error? Patrick -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: Technical problem -- please help.
Jérôme Loisel wrote: Hi! I am having a rather technical problem with my GNU/Linux system. I have tried really hard to resolve this on my own, and am out of luck. The sad thing is I'm not running Debian, but RedHat... However, I don't have commercial support, and the people on this list feel like The Most Likely to Actually Be Able to Help Me(TM)... So please help. The RedHat 6.1 installer hiccupped on my mouse, so I had to do the text install. No problem, but the hosts were not configured properly. The /etc/hosts and related files were empty. I filled them as best I could with localhost entries, scrupulously following man pages. But I feel something is still wrong. I almost always use X. Well, eventually, after 5 minutes or one day, I can no longer start new apps. I get an error. (Luckily, it is happening now, and I am not even kidding, so here is a transcript of terminal 1) AUDIT: [date]: 1515 X: client X rejected from local host AUTH name: MIT-MAGIC-COOKIE-1: ID: -1 This authorization is X authorization. Look at the man page for xauth. And trying to start an app from a terminal yields the following results: Xlib: connection to :0.0 refused by server Xlib: Client is not authorized to connect to Server kedit: cannot connect to X server :0 I read the man pages on security and such. I put ALL in /etc/hosts.allow. I tried erasing the .Xauthority file to see if it wasn't corrupt or anything. Nothing works. I keep having those problems. So here I am, restarting X I don't know how many times a day. And RedHat can't bring itself to at least host a redhat-help mailing list... or if it does, mention it on their page. In any case, please help me. This is painful. I can provide additional information to whoever wants it. Thanks, and happy coding. Jérôme Loisel ___ Get 100% FREE Internet Access powered by Excite Visit http://freeworld.excite.com -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: Remote printing with lprng
Symbolic links work fine across separate filesystems, NFS or otherwise. Hard links cannot be made to files outside the filesystem where the file linked to resides. I would guess rather that the link failure refers to an inability to connect (or a connection which terminates). Look at the lpd man page to see how access is controlled to the printer. Perhaps try telneting to port 515 on the machine to see if you get connected. Joseph A. Martin wrote: Hello, I have a system set up for my family that is networked with the rest of my systems. I want their machine to be able to print to my machine. I have lprng installed on their system and a printcap entry setup. This printcap entry works for network printing from my laptop also running lprng. However, when I try to print from their machine lpq shows status as Link failure when sending j. I have a suspicion as to the cause of the error. I seem to remember that lpr will create a link from the real file to a file in the spool directory. BUT their home directory is NFS mounted from my server. IIRC links don't work across disparate file systems. How can I tell lprng to copy the file rather than create a link? Am I off balance in my analysis here? later, joseph -- the LaterDude ICQ: 52640402 [EMAIL PROTECTED] http://www.ice-works.com/personal/LaterDude/ All opinions expressed are my own and not necessarily those of my employer unless otherwise noted. -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: Remote printing with lprng
I'm not too sure. If you print to a remote printer then the file must be sent over the network. I'm not sure why the daemon would complain then about the file not being linked. Joseph A. Martin wrote: Hello, I found part of the problem. I added my brothers' computer to /etc/hosts.lpd and it gave them access. However I have another problem. After they print a message is sent by the lpd on my server back to the user account on their machine. The message says: Your printer job ((stdin)) was not printed because it was not linked to the original file. The job did print, but what do I need to modify in order to stop that error message from being mailed. later, joseph On Wed, 02 Feb 2000, Jens B. Jorgensen wrote: Symbolic links work fine across separate filesystems, NFS or otherwise. Hard links cannot be made to files outside the filesystem where the file linked to resides. I would guess rather that the link failure refers to an inability to connect (or a connection which terminates). Look at the lpd man page to see how access is controlled to the printer. Perhaps try telneting to port 515 on the machine to see if you get connected. Joseph A. Martin wrote: Hello, I have a system set up for my family that is networked with the rest of my systems. I want their machine to be able to print to my machine. I have lprng installed on their system and a printcap entry setup. This printcap entry works for network printing from my laptop also running lprng. However, when I try to print from their machine lpq shows status as Link failure when sending j. I have a suspicion as to the cause of the error. I seem to remember that lpr will create a link from the real file to a file in the spool directory. BUT their home directory is NFS mounted from my server. IIRC links don't work across disparate file systems. How can I tell lprng to copy the file rather than create a link? Am I off balance in my analysis here? later, joseph -- the LaterDude ICQ: 52640402 [EMAIL PROTECTED] http://www.ice-works.com/personal/LaterDude/ All opinions expressed are my own and not necessarily those of my employer unless otherwise noted. -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED] -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- the LaterDude ICQ: 52640402 [EMAIL PROTECTED] http://www.ice-works.com/personal/LaterDude/ All opinions expressed are my own and not necessarily those of my employer unless otherwise noted. -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: forwarding X11 over ssh
Well, it may have failed if you ssh'd into the box and then su'd. If this is what you did then ssh to the box as directly as root. Joseph A. Martin wrote: Hello, I maintain a Linux system at our office. Quite frequently I administer this system remotely via ssh. Today I wanted to try running an X app over the SSH link. I passed the '-X' option to ssh which the man page claims will forward X connections. However when I try to launch an X client from the remote system the client says that the X11 connection rejected because of wrong authentication. As I understand it ssh is supposes to handle all of that for me. What do I need to change? How can I manually set up the correct authentication? I appreciate the help. later, joseph -- the LaterDude ICQ: 52640402 [EMAIL PROTECTED] http://www.ice-works.com/personal/LaterDude/ All opinions expressed are my own and not necessarily those of my employer unless otherwise noted. -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: Advanced hard disk mirroring!
, Onno -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: mod_ssl
You need to use the openssl tools. You don't need to user Verisign or other CA but without doing that people will get pop-up boxes the first time they visit your site about the certificate being signed by an unknown party. If they choose to keep the certificate they won't be bothered again. I think the mod_ssl docs have an example on how to generate the certificate. Mario Olimpio de Menezes wrote: Hi, Do you know how can I generate a certificate for a client? Is this possible with the mod-ssldebian package? Is there a way to generate such certificate without recurring to VerignSign or other CA? This is intended for a private system (Intranet via Internet). I'm using potato. Thanks, []s, Mario O.de MenezesMany are the plans in a man's heart, but IPEN-CNEN/SP is the Lord's purpose that prevails http://curiango.ipen.br/~mario Prov. 19.21 http://www.revistalinux.com.br -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: s3 trio 3d (86C356) won't work in svga
G96 Subsection Display Depth 8 Modes 640x480 800x600 1024x768 1280x1024 ViewPort0 0 EndSubsection Subsection Display Depth 16 Modes 640x480 800x600 1024x768 1280x1024 ViewPort0 0 EndSubsection Subsection Display Depth 24 Modes 640x480 800x600 1024x768 1280x1024 ViewPort0 0 EndSubsection Subsection Display Depth 32 Modes 640x480 800x600 1024x768 ViewPort0 0 EndSubsection EndSection -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: mgetty external sportster
On thing you might want to set if it isn't already is ATB1. This tells the modem to use a fixed serial rate, not one that's adjustable based on connection. Since you showed that mgetty was able to communicate with the modem for setting it up it's a little weird that when rings come in it gets that garbage. Oh, the [0d] [0a] are CR and LF, respectively. BTW, why are you setting the port speed to 38400? Why not just set to 115200? aphro wrote: I have an internal modem(not sure what brand) that works fine, but i want to switch to an external incase the modem needs a hard reset we dont have to shut the machine down. Its been sooo many years since i used an external modem so im kinda rusty. sofar, i got it to communicate with the sportster (both minicom and mgetty) i can get minicom to establish a terminal connection and transfer a file(by connecting with another computer and hitting 'ATA' when i saw the 'RING') the file transferred fine so i assume the modem is good. but when mgetty tries to answer i get this: 01/03 10:50:45 yS0 waiting for line to clear (VTIME), read: [0d][0a] 01/03 10:50:46 yS0 removing lock file 01/03 10:50:46 yS0 waiting... 01/03 10:50:46 yS0select returned 1 01/03 10:50:46 yS0 checking lockfiles, locking the line 01/03 10:50:46 yS0 makelock(ttyS0) called 01/03 10:50:46 yS0 do_makelock: lock='/var/lock/LCK..ttyS0' 01/03 10:50:46 yS0 lock made 01/03 10:50:46 yS0 wfr: waiting for ``RING'' 01/03 10:50:46 yS0 got: x[fe][00][80]x[00][ff][80][80][f8]x[e0][80][80][80][f8][00]x[fe][00][80]x[00][ff]x[fe][00][80]x[00][ff][80][80][f8]x[e0][80][80][80][f8][00]x[fe][00][80]x[00][ff] 01/03 10:50:56 yS0 mdm_read_byte: read returned -1: Interrupted system call 01/03 10:50:56 yS0 wfr: timeout waiting for RING 01/03 10:50:56 yS0 wfr: rc=-1, drn=0 01/03 10:50:56 yS0 huh? Junk on the line? 01/03 10:50:56 yS0 removing lock file and mgetty clearly can communicate with the modem : 01/03 11:33:23 yS0 mgetty: experimental test release 1.1.18-Sep11 01/03 11:33:23 yS0 check for lockfiles 01/03 11:33:23 yS0 checklock: stat failed, no file 01/03 11:33:23 yS0 locking the line 01/03 11:33:23 yS0 makelock(ttyS0) called 01/03 11:33:23 yS0 do_makelock: lock='/var/lock/LCK..ttyS0' 01/03 11:33:23 yS0 lock made 01/03 11:33:24 yS0 tio_get_rs232_lines: status: RTS CTS DSR DTR 01/03 11:33:24 yS0 lowering DTR to reset Modem 01/03 11:33:24 yS0 tss: set speed to 38400 (017) 01/03 11:33:24 yS0 tio_set_flow_control( HARD ) 01/03 11:33:24 yS0 waiting for line to clear (VTIME), read: 01/03 11:33:24 yS0 send: ATZ[0d] 01/03 11:33:24 yS0 waiting for ``OK'' 01/03 11:33:24 yS0 got: ATZ[0d] 01/03 11:33:24 yS0CND: ATZ[0d][0a]OK ** found ** 01/03 11:33:24 yS0 send: [0d] 01/03 11:33:24 yS0 waiting for line to clear (VTIME), read: [0d][0a][0d] 01/03 11:33:25 yS0 removing lock file 01/03 11:33:25 yS0 waiting... i've been going through documentation on the modem and on the AT command set, but so far this is as far as i've gotten. looks like those [0d][0a][0d] are escape codes? that could be screwing the modem up. im not sure of the modem's model# but it is a 28.8 sportster external, it has a bunch of dip switches on it: 1 On - data terminal ready normal **SET** off - data terminal ready override 2 Up - verbal result codes **SET** down - numeric result codes 3 Up - supress result codes down - display result codes **SET** 4 up - echo offline commands **SET** down - no echo, offline commands 5 up - auto answer on first ring, or higher if specified in NVRAM down - auto answer off **SET** (i have tried both ways with same results) 6 up - carrier detect normal **SET** down - carrier detect override 7 up - load NVRAM defaults **SET** down - load factory defaults 8 up - dumb mode down - smart mode **SET** (tried both ways, dumb mode the modem does not respond at all) any help would be greatly appreciated! [EMAIL PROTECTED] nate [mailto:[EMAIL PROTECTED] ]-- Vice President Network Operations http://www.firetrail.com/ Firetrail Internet Services Limited http://www.aphroland.org/ Everett, WA 425-348-7336http://www.linuxpowered.net/ Powered By:http://comedy.aphroland.org/ Debian 2.1 Linux 2.0.36 SMPhttp://yahoo.aphroland.org/ -[mailto:[EMAIL PROTECTED] ]-- 11:25am up 136 days, 23:21, 3 users, load average: 2.03, 1.66, 1.58 -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: python startup files
I dunno about .pythonrc.py. The docs I read said to put startup code in ~/.python-startup.py and this is what I use successfully. In order to modify python's search path from python code you just need to manipulate the sys.path list, eg.: import sys sys.path.append('path-to-append') In order to augment the python search path from an environment variable just set PYTHONPATH (colon-separated list just as the PATH variable). This is in the man page. Did you read the fine manual? Douglas Bates wrote: According to some python documentation I have read I should put commands that I want executed in every interactive python session into the file `~/.pythonrc.py'. I have done that but these commands do not seem to be executed when I start python on a Debian GNU/Linux 2.1 system with ||/ NameVersionDescription +++-===-==- ii python-base 1.5.1-7An interactive object-oriented scripting lan Is there another convention for python startup with this package? Also, what is the recommended form for adding to the python search path? I believe there is an environment variable to set. Please cc: me on replies. I am unable to keep up with the amount of traffic on this list. -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: questions about DHCP
Salman Ahmed wrote: JBJ == Jens B Jorgensen [EMAIL PROTECTED] writes: (1) What Ethernet driver should I select for this NIC ? I got the details of the HW from Windows NT which the machine already has on it. Not sure if its a PCI or ISA NIC. Has anyone heard of this NIC ? JBJ Dunno. Sorry. I selected one of the Digital XXX modules (I think it was de4x5.o) during the install and it was detected fine. (2) How do I setup the system to use DHCP once the ethernet card is setup properly ? I never setup a machine to use DHCP but I am assuming that it isn't too tricky. JBJ Just install the dhcpcd package and in your JBJ /etc/init.d/network put in a line to start up this daemon with JBJ eth0 as an argument. JBJ I couldn't get DHCP to work using the dhcpd package. On our network, most of the DHCP clients are Win9*/NT workstations. I tried a number of things in the /etc/dhcpd.conf file but still couldn't get DHCP to work. In the end, I tried using the dhclient script: dhclient eth0 and it worked fine. What should I be putting in the /etc/dhcpd.conf file ? /etc/dhcpd.conf belongs to the dhcp package which is the dhcp *server*, not the client. Stick with dhclient, it'll work fine for you. Now, if you're wondering about the /etc/dhclient.conf then you don't need to worry. I just leave mine empty since the program's defaults seem to work fine. Just do this: touch /etc/dhclient.conf Can someone please post (or email to me directly) a copy of their DHCP client configuration file ? Our network admins only know about Windows so they haven't been any help to me in terms of giving info about DHCP configuration. The only thing that they have told me is: (1) our default domain (I knew that already!) (2) the IP address of our DNS server (I knew that too from looking Windows Control Panel) I am still running slink, but will upgrade to potato soon. But I would like to have DHCP activated on boot time and don't like having to type dhclient eth0 every time. Type this as root: echo '/sbin/dhclient eth0' /etc/init.d/network Then it'll be started during the boot sequence at the appropriate time. -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: questions about generic Ethernet drivers, DHCP
Salman Ahmed wrote: I finally got a second box to install Linux on today. Unfortunately its not the Compaq Deskpro EN I had hoped for. Its an older Digital 5000 PC with the following specs: PII 233 Adaptec AHA-2940 UW SCSI Adapter 2 Seagate 2Gb SCSI HDs Matrox Mystique video card (2Mb) Digital Semiconductor EB143 based 10/100 evaluation ethernet card (1) What Ethernet driver should I select for this NIC ? I got the details of the HW from Windows NT which the machine already has on it. Not sure if its a PCI or ISA NIC. Has anyone heard of this NIC ? Dunno. Sorry. (2) How do I setup the system to use DHCP once the ethernet card is setup properly ? I never setup a machine to use DHCP but I am assuming that it isn't too tricky. Just install the dhcpcd package and in your /etc/init.d/network put in a line to start up this daemon with eth0 as an argument. I would like to get the machine up and running with slink, and then plan to upgrade it to potato and kernel 2.2.13 once networking is configured properly. Thanks for any info. -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: Netscape blows away preferences.js file when disk is full
Well, there probably isn't much you can do about this. Everything I've seen with Netscape indicates that it regenerates the preferences.js each time it runs. This means that when it's shutting down it probably opens the file for writing without the append bit. If the file system is full at this point the calls to write to the file probably just fail. You could probably report this to netscape but there's little they could (or would, I imagine) do to fix this. You just need to keep your disk from filling up. Why not write a wrapper script for netscape that will calculate available disk space when before it starts and the write to this same preferences file the preference to limit netscape's cache allowance size to something less than what's available. This should fix the problem unless you're creating other files while running netscape. Bob Horvath wrote: I am seeing a recurring problem that is quite annoying. My disk space is at a premium. If I am doing something and the disk fills, I first get a Error saving bookmarks error. Netscape doesn't crash, but upon stopping it and restarting it, all of the preferences are gone. The preferences.js file is blown away to have a length of zero. I have to go in and reconfigure all of the mail and news servers, as well as any other preferences in there. Anyone else seen this? How would I report this to Netscape? -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: Debian and QMail?
The only possible gotcha I could tell you about is that you'll have to build the package. Actually this is very, very easy because the src packages (debian hasn't been granted the right to distribute qmail in binary) build the binary packages and then you just install those. It's really easy. Robert L. Harris wrote: I currently have a RedHat6 mailserver. I'm considering blowing it away this weekend and installing Debian on it. I use qmail since it works really well for serving currently and I like it's claim to security. Are there any gotcha's I should know about first? Robert :wq! --- Robert L. Harris| Low quality in a product happens. Senior System Engineer |That doesn't mean it's right and at RnD Consulting.| and defintely doesn't mean it should \_ be accepted. Require quality. http://www.rnd-consulting.com/~nomad DISCLAIMER: These are MY OPINIONS ALONE. I speak for no-one else. FYI: perl -e 'print $i=pack(c5,(41*2),sqrt(7056),(unpack(c,H)-2),oct(115),10);' -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: Kernel SCSI for Adaptec 1505
A new Adaptec 1505 scsi card? Well, this card could hardly be new. Anyway! This card is incapable of auto-detect. You *must* supply kernel (or module if you're loading as a module) options to tell the driver what IO, IRQ, etc. you want to use. For instance, in my /etc/lilo.conf file I've got: append = aha152x=0x340,9,7,1,1,0,100,1 This is because mine's compiled into the kernel. If I were loading as a module then I'd have a file /etc/modutils/aha152x and inside this file: options aha152x aha152x=0x340,11,7,1,1,0,100,1 This works just dandy for me. Bill Leach wrote: First, while subscribed to a number of debian lists, I am not currently subscribed to debian-user so please reply to me or cc me. Thank you. I have tried to build (yet another new kernel) for a machine that has the now usual pri/sec ide and a new Adaptec 1505 scsi card. The card is recognized by the machine's bios. The kernel builds ok and the module is created but will not load. The errors that I get are: modprobe: Can't locate module aha152x.o frodo:/lib/modules/2.2.13/scsi# modprobe aha152x /lib/modules/2.2.13/scsi/aha152x.o: init_module: Device or resource busy /lib/modules/2.2.13/scsi/aha152x.o: insmod /lib/modules/2.2.13/scsi/aha152x.o failed /lib/modules/2.2.13/scsi/aha152x.o: insmod aha152x failed It seems that this is the sort of response that is received when a card is not found but the modules. Anyone have any hints, ideas, etc.? -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: boot disk hangs on my puta
How long did you wait after it hung. Sometimes it can sit there for quite a while in that scsi-detection part but it does boot after a minute or so. Andrew Walker wrote: The boot disk seems to work on my new puta but it seems to hang on my old one it seems to be at that point where it trys to detect scsi, is there any way to tell it that there is no scsi on my system? -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: E-mail (POP) not working
This smells suspiciously like a permissions (tcpd) problem. Check /var/log/daemon.log. If tcpd is the problem then there should be a line like: Dec 8 18:11:53 chilin qpopper[28923]: refused connect from X.X.X.X Also of course have a look at /etc/hosts.allow and /etc/hosts.deny to see how things are configured there. Paulo Henrique Baptista de Oliveira wrote: Hi all, I have working in a ISP that POP stopped to work for the clients with modem. Altought I can telnet port 110 the clients with modem receive the following error Error: Your server closed the connection. This could be problem with server, network or large inativity period. I have installed exim and qpopper (both running in inetd). What else I can test? -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: Can not boot from HD after install
What's the contents of your /etc/lilo.conf? Do you have boot=/dev/hda? Why not also send us a printout of the partition setup from (c)fdisk. Do you have the bootable flag set on /dev/hda2? Matthew Denson wrote: I posted this to the newsgroup linux.debian.user yesterday. But now I'm subscribed to this list. I apologize if you've seen this before. Good evening all, I have just completed the installation of Debian/Linux 2.1 on my old Compaq laptop. When I finished it would not boot from the hard drive. The error I got was Non-System disk or disk error Replace and press any key when ready I insert the Boot Floppy and it boots fine (albeit real slow). Could someone give me a remedy to this? More detailed info: This computer has a retrofitted ~800 MB hard drive and in it's previous incarnation (up until last night) it was a DOS machine which had the On-Track Disk Manager software installed to translate the big disk for DOS. The debian installation balked at the partition step, it said the disk was factory fresh or screwed up. I assumed it might be the Disk Manager partition voodoo, so... I did not have the disk for Disk Manager so I did not properly uninstall it, but rather simply repartitioned the drive with DOS fdisk. Same message from the installation program so I shelled out and ran cfdisk directly. This worked great. I partitioned the drive the way I wanted with a ~768MB hda2 at the beginning of the drive and a ~8MB swap hda1 at the end. The rest of the installation script went smoothly until the reboot when I discovered the above. I have a nagging suspicion that the hda2 needs to be reduced or that the boot record still has the Disk Manager strap in it. I'm going to set it aside in case I have to start over. Any guidance would be appreciated. TTFN, Matthew Denson -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: Adding a lot of users in a single operation.
The trouble is that none of the utilities (that I know of) allow you to specify the password on the command line. Perhaps you can do it with pipes, like the following. Ok, let's say you have a file users.txt with a list of user names you want to add, one per line. for i in $(cat users.txt) ; do pwd=$(makepasswd) ; echo $i $pwd userpass.txt ; adduser --disabled-password --gecos $i $i ; cpwd=$(perl -e 'print crypt('$pwd', '$i');') ; echo -e '/^'$i':/ s/:!:/:'$cpwd':/\nw\n' | ed /etc/passwd ; done That should do it. (See, you had to user perl after all.) Mind you the above is all one line, you'll have to fix the breaks inserted by my mail-reader and yours. Good practice dictates you save your /etc/passwd file before completing the above. Also, the above assumes you are not using shadow passwords. In any case the above should be adaptable to shadow passwords too if you can see what I'm doing. Jose L Gomez Dans wrote: Hi! I was wondering whether it would be possible to add a number of users with a single command. Ideally, there would be a call to pwgen (or something like that!) and the passwords would be held in a file. Any ideas (apart from do it in perl!)? Jose -- Jose L Gomez Dans PhD student Radar Communications Group Department of Electronic Engineering University of Sheffield UK -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: Can not boot from HD after install
In your lilo.conf change: boot=/dev/hda2 to: boot=/dev/hda That should work. Matthew Denson wrote: OK here are the files. /etc/lilo.conf -- boot=/dev/hda2 root=/dev/hda2 install=/boot/boot.b map=/boot/map vga=normal delay=20 image=/vmlinuz label=Linux read-only Partition Table for /dev/hda ---Starting--- EndingStart Number of # Flags Head Sect Cyl ID Head Sect CylSector Sectors -- - - 1 0x0001 781 0x82 31 63 788 1574496 16128 2 0x80110 0x83 31 63 780 63 1574433 3 0x00000 0x000000 0 4 0x00000 0x000000 0 Thanks. Matthew Denson -Original Message- From: Jens B. Jorgensen [mailto:[EMAIL PROTECTED] Sent: Thursday, December 02, 1999 10:34 AM To: Matthew Denson Cc: debian-user@lists.debian.org Subject: Re: Can not boot from HD after install What's the contents of your /etc/lilo.conf? Do you have boot=/dev/hda? Why not also send us a printout of the partition setup from (c)fdisk. Do you have the bootable flag set on /dev/hda2? Matthew Denson wrote: I posted this to the newsgroup linux.debian.user yesterday. But now I'm subscribed to this list. I apologize if you've seen this before. Good evening all, I have just completed the installation of Debian/Linux 2.1 on my old Compaq laptop. When I finished it would not boot from the hard drive. The error I got was Non-System disk or disk error Replace and press any key when ready I insert the Boot Floppy and it boots fine (albeit real slow). Could someone give me a remedy to this? More detailed info: This computer has a retrofitted ~800 MB hard drive and in it's previous incarnation (up until last night) it was a DOS machine which had the On-Track Disk Manager software installed to translate the big disk for DOS. The debian installation balked at the partition step, it said the disk was factory fresh or screwed up. I assumed it might be the Disk Manager partition voodoo, so... I did not have the disk for Disk Manager so I did not properly uninstall it, but rather simply repartitioned the drive with DOS fdisk. Same message from the installation program so I shelled out and ran cfdisk directly. This worked great. I partitioned the drive the way I wanted with a ~768MB hda2 at the beginning of the drive and a ~8MB swap hda1 at the end. The rest of the installation script went smoothly until the reboot when I discovered the above. I have a nagging suspicion that the hda2 needs to be reduced or that the boot record still has the Disk Manager strap in it. I'm going to set it aside in case I have to start over. Any guidance would be appreciated. TTFN, Matthew Denson -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED] -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: OT: Network map solutions
You should check out xfig. It's not exactly Visio but may be the best tool in this case. Marc Mongeon wrote: Sorry, this is a bit off-topic, but I thought it might be a quick and painless way to reach people who have already tackled a problem I'm facing. I'm looking for a way to create rather nice-looking network topology maps. I figure a solution will involve an easy-to-use object-oriented drawing program with a library of network objects (routers, switches, links, clouds, etc.) that is also extensible, so I can add my own ob- jects. I'm thinking of something Visio-like, but given the quality of open-source software I've worked with in the past, I'm certain I can find something even better. Has anybody found a utility that fits this description? I played around a little bit with a drawing program called dia, but its object lib- rary was minimal. Thanks in advance for any help you can offer. Marc -- Marc Mongeon [EMAIL PROTECTED] Unix Specialist Ban-Koe Systems 9100 W Bloomington Fwy Bloomington, MN 55431-2200 (612)888-0123, x417 | FAX: (612)888-3344 -- It's such a fine line between clever and stupid. -- David St. Hubbins and Nigel Tufnel of Spinal Tap -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: scsi bus and tape backup questions
Bill wrote: greetings all: i have a couple of questions i hope someone can answer for me. to start: i have a sparc station and linux (debian 2.1, 2.0.35 kernel) 1st question - re: scsi my system works pretty good. i am only curious as to why the sparc scsi bus locates the internal system hard drive as target 3 when i have external hard drives, as it is physically closer to the system. i.e. esp0: FAST chip is fas100a (family=0, version=2) ESP: Total of 1 ESP hosts found, 1 actually in use. scsi0 : Sparc ESP100A-FAST scsi : 1 host. Vendor: SEAGATE Model: ST15230N Rev: 0638 Type: Direct-Access ANSI SCSI revision: 02 Detected scsi disk sda at scsi0, channel 0, id 0, lun 0 Vendor: SEAGATE Model: ST15230N Rev: 0638 Type: Direct-Access ANSI SCSI revision: 02 Detected scsi disk sdb at scsi0, channel 0, id 2, lun 0 Vendor: SEAGATE Model: ST1480 SUN0424 Rev: 8628 Type: Direct-Access ANSI SCSI revision: 02 Detected scsi disk sdc at scsi0, channel 0, id 3, lun 0 Vendor: HPModel: C1533ARev: 9503 Type: Sequential-Access ANSI SCSI revision: 02 scsi : detected 3 SCSI disks total. sda and sdb are in an external box (scsi id 2) whereas the small internal is sdc. i am just curious why the sparc does not find and id the internal first, i.e. target 0 or 1; sda. ??? Well, the disks are iterated by SCSI ID, which is usually set by jumpers on the drive to a number between 0 and 7. Physical distance is irrelevant and not really knowable by the SCSI adapter or the OS. For some historical reason Sun boxes have always (all the one's I've ever encountered, from Sparc 2s up through Ultras) had their internal drives with the OS installed as SCSI ID 3. 2nd question: re: tape backups. i am trying to finess mt, dump, and restore to do the job for us. they work o.k., except that there are a few main information bits that seem to be missing. i went through the mans on each of these, and i can't see anywhere an operation that will tell me how much space is left on the tape. GNU mt version 2.4.2 Package: dump Version: 0.4b4-13 Unfortunately there isn't really a way for the device or the software to know how much space is left on the tape. The drive can't find this out. I've done the same kind of thing as you. I've written a perl script which takes care of keeping track what's on the tape (the script labels the tape by keeping a record at the end of what dump was made when) and puts multiple backups on the same tape so I only have to change it once a week. I'm sending this along in a separate email and you may do with it what you wish. 3rd question: cdrom drive: i also have a cdrom drive ii want to put on the system. whenever i add it to the chain, however, i get a scsi bus interrupt; reset infinite loop when i try to boot. here's how my system looks as above for the hard drives. external case with hard drives: id 2 plus: hp tape backup: scsi id 4 this configuration works fine. but when i add the cdrom to the end of the chain, i have the problem. the last device is terminated properly. with respect to scsi types; sparc - (fast scsi 2) -- external drives (scsi) -- hp tape (scsi) -- cdrom (sun - fast scsi 2) i tried setting the cdrom as device 5 and 6, neither worked. however, the nvram probe-scsi util did find them all properly and it looked fine in that setting. it was only when i went to boot when it went into an infinite loop. could it be a driver problem? i have the sg and st modules loaded. I doubt it's a driver problem but how far does the boot get? What's the last thing that happens. I'm still going to guess that this is a termination problem. Is this an external cdrom in its own enclosure? Is your chain too long (can't remember the limit offhand). thanks for any assistance with this. Sure. Good luck. -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: class 1 modem
You can also use efax. luis wrote: hello which package i muss use to send faxes with a class 1 modem i already can use minicom with it thanks a lot -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: secure pop3 via ssh
You can do it easily with stunnel as well. Here's my line from inetd.conf: pop3s stream tcp nowait root/usr/sbin/stunnel /usr/sbin/stunnel -p /etc/ssl/certs/imap4.pem -l /usr/sbin/cucipop -Yq You need to create the certificate using the openssl package's openssl program. Mario Olimpio de Menezes wrote: Hi, How can I set a secure pop3 server using ssh? I read the man page for fetchmail and there is an example of a secure conection via ssh. How do I implement the server? Thanks, []s, Mario O.de MenezesMany are the plans in a man's heart, but IPEN-CNEN/SP is the Lord's purpose that prevails http://curiango.ipen.br/~mario Prov. 19.21 http://www.revistalinux.com.br -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: Another Crazy DNS Question... ;-)
Bob Nielsen wrote: On Thu, Nov 04, 1999 at 12:01:29AM -0500, William T Wilson wrote: Yes, you can point A records from two domains to the same IP address. You can also use a CNAME record. I'm not sure when one approach would be preferred over the other. This is true however the rfc (not sure what number to look at) I believe says that the hostname mentioned in an MX record shouldn't be a CNAME. I should also mention that as far as how your mailserver appears to other machines, when a mailserver is _sending_ mail out it first announces its identity via the HELO or EHLO commands. What identity the host presents itself as to other SMTP hosts is configurable in all the mail server packages I've dealt with. That said, mail server packages that I've dealt with will only allow you configure it to present itself as a single identity. This doesn't affect mail clients which would get mail via POP/IMAP of course. -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: Password encryption
Most likely because it uses some kind of regular DES which isn't strong enough to fall under export controls. Sami Dalouche wrote: Hi everyone, Today, I've just realized that the passwd package uses encryption. The problem is that I wonder why it's not in the non-US section. Every package that uses encryption seems to be in non-US, so why isn't passwd in non-US too ? Bye, sami -- E II A NN N LL II NN N U U X X DD D E__B__B II A___A N N N LL II N N N U U XX DD D E B B II A A N N N LL II N N N U U XX E II A A N NN L II N NN U X X -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: proxyarp
You can use the arp command to make your machine answer arp requests for whatever address you'd like. The arp command is in the netbase package so you should already have it. For example, if I wanted my machine to answer arp requests for the host 192.9.202.22 and my ethernet address is 00:C0:4F:4F:F5:59 and I want to answer request on the eth0 interface then the command would be: arp -i eth0 -s 192.9.202.22 00:C0:4F:4F:F5:59 pub esl wrote: Help, please. How do I enable proxyarp on my ethernet interface? I know how to do this in PPP. Thanks. E.LL -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: Debian and Microsoft proxy
No, the M$ proxy server is not capable of ip masq, at least not in the linux sense of ip masq. As you've noticed, using the M$ proxy requires their own proprietary client software (suprised?). M$ does not release specs of how to talk to their proxy server. That said, I can assure you that the M$ proxy server does support socks4 protocol. This must be enabled by your systems administrator. Once this is done you can install the socks4-client package. This package includes a runsocks script which you can use to run basically any program 'socksified'. This will work (I myself have a debian machine which accesses the net through M$ proxy server). David Cureton wrote: Hi all, I have access to the net via microsoft proxy which is capable of ip masq. However to used it on wintel boxes a client need to be installed. Once installed this which works fine. Is it possible to access the net (ip masq. not proxy web content) via the MS proxy server using debian. Is it enought to point to the ms proxy box as my gateway. I will experiment when I have some time, Has anyone else tried? Comments?. -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: eth1 can't be loaded
Take a look at the output of dmesg and see if there are any other messages being generated which might indicate the problem. Perhaps it wants an IRQ it can't get? Have you installed any new hardware into the machine? Patrick Kirk wrote: Hi all, Does this look like a hardware failure on a NIC? I'm getting a little desperate now as the kids will be most upset if they can't browse the Teletubbies sites over the weekend! enterprise:/lib/modules/2.0.36/net# ifconfig eth1 10.0.0.25 netmask 255.255.255. 0 broadcast 10.0.0.25 SIOCSIFFLAGS: Resource temporarily unavailable enterprise:/lib/modules/2.0.36/net# cat /etc/modules # /etc/modules: kernel modules to load at boot time. # # This file should contain the names of kernel modules that are # to be loaded at boot time, one per line. Comments begin with # a `#', and everything on the line after them are ignored. # An entry named `auto' will cause the system to start kerneld immediately. # Kerneld then loads modules on demand. `noauto' disables kerneld completely. #auto ne2k-pci tulip ip_alias psaux -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: How to Access 2nd Serial Port Cisco 675?
I would just try using minicom or seyon. Both these programs will allow you to make a terminal connection to the device and type whatever commands you wish. The second serial port ordinarily will be accessed as /dev/ttyS1. Art Lemasters wrote: How can I access my Cisco 675 modem/router through the second serial interface? ...anyone have experience with the 675 or at least with access any device through the second serial port? Art -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: DHCP problem....
I would add: if you want to be able to ping the machine just like you can windows boxes then you'll need the Linux version of the software which makes this mechanism (Lan Manager a.k.a. Windows Networking) work. This package is called samba and is available as a couple different debian packages. Doing 'apt-get install samba' should get you there. Note that you may need a script which will update /etc/hosts. I'm not sure if the current dhcp package will do this and samba will not work otherwise. Phil Brutsche wrote: A long time ago, in a galaxy far, far way, someone said... I just installed DHCP client on Debian slink machine. It works almost fine it just does not pick up the host name :(. It is seen on the network by its IP address only. If I ping it by the name from the local console it responds that the host name is not known (I removed its static IP entry from the hosts file). Any suggestions will be appreciated :) This is how I read your message: 1) A computer running slink called elvis (for example) uses DHCP to get it's IP# and related information. 2) Someone else on the network tries to ping elvis, and fails 3) Someone on elvis tries to ping elvis, and fails If this is the case, welcome to DHCP :) The reason #2 and #3 fails is because there is no way for another system to find out elvis' IP#, either because the name 'elvis' wasn't found in /etc/hosts (or the equivalent), or in DNS, as you found out. Your solutions are: 1) convince the local system administrator(s) to use Dynamic DNS. This allows the DHCP server to update the DNS server's IP# database 2) convince the local system administrator(s) to give you a static IP# somehow, either through DHCP by MAC address or an outright static IP# (no DHCP involved) 3) use one of the free dynamic hosting services available on the internet; an example is www.dhs.org -- -- Phil Brutsche [EMAIL PROTECTED] There are two things that are infinite; Human stupidity and the universe. And I'm not sure about the universe. - Albert Einstein -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: DHCP problem....
Yes, ok. Well the first thing is you want to use the newer client, (dhcp-client-beta if you're running slink, dhcp-client if you're running potato). Using the newer client is necessary because this client does not go into the background until it gets an address (or times out, but let's hope/assume that doesn't happen). The older client did go into the background with the unfortunate result that samba and other network daemons that want to know what your address is and want to look it up by host name would be started right away before the dhcp client had time to get an address. The next thing is to modify /etc/dhclient-script. This script is run at various times by the dhcp client daemon in order to accomplish certain things like setting up routes. We need to add code to it so that when dhcp acquires an IP address it will update the /etc/hosts file to make the machine's host name resolve to it's IP address properly. I've included my /etc/dhclient-script which you can copy from. J.M. \Jersey\ Miszczyk wrote: Hello Phil, I am not that much worried about the other machines on the network, since they are Win systems and I am installing Samba on my machine but.. my machine does not even know itself :(. If I put some sort of static IP address into hosts file this IP address is used when I try to ping say, elvis and the error message that network is unreachable shows on my screen. If I ping the IP address which I pick up from running ifconfig for example, everything works fine. I can force network admin to give me a static address but this is besides the point lol... now the matter of pride is to solve the problem if you know what I mean :))). Jens Jorgensen mentiond about a script updating hosts file automatically... would you have any idea it should look like? :) Best regards and thank you for help :) Jersey -- From: Phil Brutsche[SMTP:[EMAIL PROTECTED] Sent: Wednesday, October 13, 1999 12:36 AM To: [EMAIL PROTECTED] Cc: Debian-User Subject:Re: DHCP problem A long time ago, in a galaxy far, far way, someone said... I just installed DHCP client on Debian slink machine. It works almost fine it just does not pick up the host name :(. It is seen on the network by its IP address only. If I ping it by the name from the local console it responds that the host name is not known (I removed its static IP entry from the hosts file). Any suggestions will be appreciated :) This is how I read your message: 1) A computer running slink called elvis (for example) uses DHCP to get it's IP# and related information. 2) Someone else on the network tries to ping elvis, and fails 3) Someone on elvis tries to ping elvis, and fails If this is the case, welcome to DHCP :) The reason #2 and #3 fails is because there is no way for another system to find out elvis' IP#, either because the name 'elvis' wasn't found in /etc/hosts (or the equivalent), or in DNS, as you found out. Your solutions are: 1) convince the local system administrator(s) to use Dynamic DNS. This allows the DHCP server to update the DNS server's IP# database 2) convince the local system administrator(s) to give you a static IP# somehow, either through DHCP by MAC address or an outright static IP# (no DHCP involved) 3) use one of the free dynamic hosting services available on the internet; an example is www.dhs.org -- -- Phil Brutsche [EMAIL PROTECTED] There are two things that are infinite; Human stupidity and the universe. And I'm not sure about the universe. - Albert Einstein -- Jens B. Jorgensen [EMAIL PROTECTED] #!/bin/sh # dhclient-script for Linux. Dan Halbert, March, 1997. # Updated for Linux 2.[12] by Brian J. Murrell, January 1999. # No guarantees about this. I'm a novice at the details of Linux # networking. # Notes: # 0. This script is based on the netbsd script supplied with dhcp-970306. # 1. ifconfig down apparently deletes all relevant routes and flushes # the arp cache, so this doesn't need to be done explicitly. # 2. The alias address handling here has not been tested AT ALL. # I'm just going by the doc of modern Linux ip aliasing, which uses # notations like eth0:0, eth0:1, for each alias. # 3. I have to calculate the network address, and calculate the broadcast # address if it is not supplied. This might be much more easily done # by the dhclient C code, and passed on. # 4. TIMEOUT not tested. ping has a flag I don't know, and I'm suspicious # of the $1 in its args. update_hosts () { if sed -e '/^[0-9]\{1,3\}\.[0-9]\{1,3\}\.[0-9]\{1,3\}\.[0-9]\{1,3\}[[:space:]]\+.*'$(/bin/hostname)'.*/ s/^[0-9]\{1,3\}\.[0-9]\{1,3\}\.[0-9]\{1,3\}\.[0-9]\{1,3\}/'$1'/' /etc/hosts /etc/hosts.new.$$ then /bin/cp /etc/hosts /etc/hosts.old mv /etc/hosts.new.$$ /etc/hosts else rm /etc/hosts.new
Re: fax page size
If the fax pages appear stretched then the conflict is most likely due to a confusion between 'normal' and 'fine' resolution of the fax you're sending. That is, perhaps the fax is formatted at 'normal' resolution (204 x 98 dpi) but the fax software thinks it's been formatted at 'fine' resolution (204 x 196 dpi), and tells the fax receiver so. You see, the resolution of the document is a negotiated parameter in the fax transmission. [EMAIL PROTECTED] wrote: Hi i am using mgetty/sendfax to send faxes but unfortunately it sends the faxes in a stretched format... Where abouts can i change it so that it faxes as A5? thanx -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: Callback
callback is generally accomplished through an extension to ppp. I don't think that the linux pppd daemon has support for this (at least not version 2.3.5 that I'm using). A quick look at the README on the upstream maintainer's site doesn't seem to indicate support for this either: ftp://cs.anu.edu.au/pub/software/ppp/README A. Barry wrote: Hi, I am a new converter to linux os, ending the misery of windows os. I managed will till now to install debian linux and managed to have ppp to work fine for some connections. My main isp requires a callback authentication to connect, and I could not fined any mention of callback in all the docs I have, including the HOW-TOs. Any help. A. Barry -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: ppp broken, log says: Receive serial link is not 8-bit clean
It's more likely that some further authentication is necessary, i.e. PPP itself is not being started automatically upon connect but instead there's some other prompt which must be negotiated. The best way to diagnose this is to use seyon or minicom (or other terminal program) to manually dial the number and see what the remote side does. David Karlin wrote: Hello, Upon returning home from a trip, my slink box is no longer able to connect to my ISP. Up until the time I left, it had been working fine. I was even able to telnet (ssh) into the system for the first few days I was away. Fortunately/unfortunately, my win95 system is able to connect as before, with no problem. Here is a snippet from /var/log/ppp.log: Sep 29 19:11:56 champagne chat[27653]: expect (CONNECT) Sep 29 19:11:56 champagne chat[27653]: ^M Sep 29 19:12:16 champagne chat[27653]: ATDT9494614^M^M Sep 29 19:12:16 champagne chat[27653]: CONNECT Sep 29 19:12:16 champagne chat[27653]: -- got it Sep 29 19:12:16 champagne chat[27653]: send (\d) Sep 29 19:12:17 champagne pppd[27652]: Serial connection established. Sep 29 19:12:18 champagne pppd[27652]: Using interface ppp0 Sep 29 19:12:18 champagne pppd[27652]: Connect: ppp0 -- /dev/ttyS1 Sep 29 19:12:18 champagne pppd[27652]: sent [LCP ConfReq id=0x1 asyncmap 0x0 magic 0xb74c60e2 pcomp accomp] Sep 29 19:12:45 champagne last message repeated 9 times Sep 29 19:12:48 champagne pppd[27652]: LCP: timeout sending Config-Requests Sep 29 19:12:48 champagne pppd[27652]: Connection terminated. Sep 29 19:12:48 champagne pppd[27652]: Receive serial link is not 8-bit clean: Sep 29 19:12:48 champagne pppd[27652]: Problem: all had even parity Sep 29 19:12:48 champagne pppd[27652]: Hangup (SIGHUP) Sep 29 19:12:48 champagne pppd[27652]: Exit. My ISP has me connect through a Compuserve access number because they have none of their own in my local calling area. I suspect that something has changed at the Compuserve end of the line while I was away, and that this thing is something that MS DUN can deal with automatically, but which must be configured manually in linux. I further suspect that some change was made in the parity setting, as indicated by /var/log/ppp.log. Which file must I edit in order to get ppp to work with even parity? Is this some thing that can be fixed by editing my modems initialization string, or must I edit some other configuration file? Please enlighten me. Thanks. P.S. Please cc: me as I'm a digest subscriber. -- David Karlin mailto:[EMAIL PROTECTED] Reduced to using Windows 95 :-( -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: pptp for linux - where can it be
With a newer version of ppp (=2.3.6, such as the one in potato) you can specify a pty connection. Using this with ssh will allow you to do this. Look through the readme and the man page for this newer version to see how to do this. Alex V. Toropov wrote: Hi, all I remember that some where I read about organizing pptp link between to linux-box over Internet. Can some one figure me where can iI obtain this info ? -- Jens B. Jorgensen [EMAIL PROTECTED]
Re: imap
Jaldhar H. Vyas wrote: On Fri, 24 Sep 1999, Jens B. Jorgensen wrote: Ah, but the latest version *does* support maildirs (I'm using it). Actually I did compile my own version to apply one patch that the maintainer hadn't yet applied. The bug caused a file to not be removed if a folder is deleted (not too big of a deal). The next version will have this patch. The thing I had trouble with was the pickiness of it. It wants your inbox maildir to be ~/mail/Maildir and doesn't seem to find it otherwise. *sigh* I really screwed up with this ~/ vs ~/mail thing didn't I? The question is at this point is it worth it to change back or will this just make the other half of the users mad at me? :-) Actually, now that I know that's what it expects I like this better. This way new folders appear under ~/mail as well and don't clutter up the top level of one's home dir. It'd be nice if there were a note in the man page to this effect though 8-). -- Jens B. Jorgensen [EMAIL PROTECTED]