Re: Log monitoring GUI program
On Wed, 9 Sep 1998, George Bonser wrote: > On Thu, 10 Sep 1998, Richard L Shepherd wrote: > > > interface to manage/interpret/flag all the logging coming in. > > > > Has anyone heard of such a package? Especially a debianised one? > > > > There are two that I know of, XLogMaster and XLogMonitor. > > Try: > > http://www.gnu.org/software/xlogmaster/xlogmaster.html and > http://members.xoom.com/chaosmaker/linux/ > > I think XLogMaster is what you are really after, I am not sure if there is > a Debian package. Thanks, in fact I just found that there is one in slink, which is OK for me even though the rest of my system is hamm. It's version 1.4.1 whereas the latest release is 1.4.3, but I've opted to stick with the ease of install over the lates version for now ;-) R.
Log monitoring GUI program
I seem to remember reading sometime (in comp.os.linux.announce I think) about a package that was something like a GUI that provided log-file (e.g. syslog written stuff) monitoring features. I've got to the point of having several machines syslog-ing to one central machine (as well as keeping local copies) and would like to have a nice interface to manage/interpret/flag all the logging coming in. Has anyone heard of such a package? Especially a debianised one? TIA, Richard.
LRP offline
Anyone here know why the LRP (Linux Router Project) is offline at present? The URl is: http://www.psychosis.com/linux-router/ I desperately need to get to their pages - are there any mirrors? TIA -- 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) Phone: 07-838-4764 8<--->8 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Telnet Proxy anyone?
On Tue, 21 Apr 1998, Richard L Shepherd wrote: > Has anyone heard of a telnet-proxy package (especially for linux and > Debian of course)? > > We have some people who have (and want to keep) their subnet blocked for > offsite access (so they do all their WWW browsing via a WWWcache which can > then be billed for). I was wondering if there is something we can do like > this for telnet. Any ideas? Thanks for all the help here everyone. I have settled for the TIS firewall Tool-kit. It does everything I need and almost everything I want ;-) Combined with the transparent-proxy feature of the Linux kernel I can get it to intercept/redirect telnet connection attempts to the tn-gw proxy of the TIS fwtk. However I cannot get it to automatically connect to the remote destination after successfully authenticating the local user. The problem is how to find out where they wanted to go, without them having to type "c host" at the "tn-gw>" prompt. I can live with this 2 step procedure though, Cheers, 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) Phone: 07-838-4764 8<--->8 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Telnet Proxy anyone?
Has anyone heard of a telnet-proxy package (especially for linux and Debian of course)? We have some people who have (and want to keep) their subnet blocked for offsite access (so they do all their WWW browsing via a WWWcache which can then be billed for). I was wondering if there is something we can do like this for telnet. Any ideas? TIA 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) 8<--->8 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: no popclient?
On Mon, 20 Apr 1998, Remco Blaakmeer wrote: > On Mon, 20 Apr 1998, Richard L Shepherd wrote: > > > I see that popclient has been missing for sometime. I have a machine with > > popclient installed from some time ago (early days of bo). Just recently > > it has stopped working and I see no alternative package to get either the > > binary (popclient) or the equivalent (lightweight) functionality. > > > > The error I get when trying to pop off ANY server via my (rather old) > > popclient binary now is: > > > > popclient: querying mailserv.waikato.ac.nz at Mon Apr 20 15:23:43 1998 > > popclient: openuserfolder: open(): No such file or directory > > > > Now all my local mail readers work fine so there's no reason I know of why > > it's not happy anymore. The only shared lib it uses is libc5 (which I > > have). > > > > Any ideas on how to fix this or what else to use? > > Some time ago, when the popclient program was enhanced to support more > protocols than just POP, the name was changed to fetchmail. fetchmail is > in bo (stable) and also in hamm (frozen). Thanks, that's just what I needed. Thanks for the quick and helpful response everyone. 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) 8<--->8 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
no popclient?
I see that popclient has been missing for sometime. I have a machine with popclient installed from some time ago (early days of bo). Just recently it has stopped working and I see no alternative package to get either the binary (popclient) or the equivalent (lightweight) functionality. The error I get when trying to pop off ANY server via my (rather old) popclient binary now is: popclient: querying mailserv.waikato.ac.nz at Mon Apr 20 15:23:43 1998 popclient: openuserfolder: open(): No such file or directory Now all my local mail readers work fine so there's no reason I know of why it's not happy anymore. The only shared lib it uses is libc5 (which I have). Any ideas on how to fix this or what else to use? 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) 8<--->8 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Install disks for PowerPC
On Sun, 19 Apr 1998, Alain Toussaint wrote: > > where are they? I see a directory for disks-i386 and several other > > platforms (alpha, sparc and m68k) but no disks-powerpc. > > > > So how does one get started installing Debian on a PowerPC MAC, and where > > are the instructions? > > > > i don't think there will be a Debian for the Mac version of the powerPC as > long as these machine remain proprietary,there's version of linux for the > IBM powerPC line but the only linux distribution especially for the Mac > PPC is available at: Thanx for the reply. So how does one get started with the IBM PowerPC then if there are no disks-powerpc? 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) Phone: 07-838-4764 8<--->8 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Install disks for PowerPC
where are they? I see a directory for disks-i386 and several other platforms (alpha, sparc and m68k) but no disks-powerpc. So how does one get started installing Debian on a PowerPC MAC, and where are the instructions? TIA 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) 8<--->8 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Two-monitors
I have two video cards installed on my machine. I also have 2 monitors I'd like to have running. However I have not yet figured out how to get the system using the second video card. So far the only place I see any trace of the second one is in /proc/pci where it is listed as a PCI device. To possibly make life more difficult, both cards are the same brand - the only difference being the amount of RAM installed on each (4 vs 8MB). Does anyone have such a setup going? I plan to use x2x once I have them both going, but need to get over this slight hurdle first ;-) TIA 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) 8<--->8 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Two Monitors
Has anyone any experience in using two monitors on one machine. I'm thinking particularly of X. I can get a second video card (which I guess is mandatory or there'd be nowhere to plug the second one in right?). What I'm really interested in is getting the tow side-by-side and having X treat them as one big screen (similar to the way Macintosh's can do this). Is this possible? Or is my best bet to run up X :0 and X :1 and use to switch between them? I'd be interested to hear... 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) Phone: 07-838-4764 8<--->8 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: tulip driver
On Fri, 2 Jan 1998, Ian Keith Setford wrote: > > Hi- > > I am having no luck getting a new ethernet card up. I just installed a > new card with a DEC "tulip" chip and I can't get it working. Does anyone > have a "tulip" card working? If so, could you share with me how you got > it running? > > I am running 2.0.30 on a ppro 200. > > Thanks in advance! > > -Ian Try the latest tulip driver from: http://cesdis.gsfc.nasa.gov/linux/drivers/tulip.html I have found that it works significantly better than the one in the Debian kernel-source packages, and in some cases it works where the old ones simply don't! (e.g. DEC cards with the 21143 chip) 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) Phone: 07-838-4764 8<--->8 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Install debian with serial port as console?
On Fri, 2 Jan 1998, Larry Owens wrote: > Does anyone know if it is possible to install debian on a PC that does > not have a video card or keyboard? I would like to set up a linux > system that would use a modem attatched to a serial port as the only > console. I have used the serial-console patch on some 1.2/1.3 systems here running kernel 2.0.29 and 2.0.30 but have come across some problems that became more apparent as I installed more systems: 1. when the serial-console patch is in use the line in the file /etc/init.d/boot that is supposed to run modutils is mysteriously not executed, whereas when not using the serial-console patch it IS successfully executed! Solving this got enabled me to make more things modules rather than compiled in. 2. when doing an upgrade via dselect certain packages fiddled with /dev/console and /dev/tty0 which was contrary to the way prescribed for the serial-console to work. Also, when I found that the DEC Prioris XL6x00 series have a BIOS that supports console redirection I thought I'd try to install a system from scratch through a terminal connected to the serial port - however the Debian install disk just switched back to using the VGA screen after booting from the floppy, so I couldn't use it until I'd installed and recompiled the kernel with the serial-console patch applied. So the Debian install disk must specifically send it's output to the VGA screen, bypassing what the BIOS "thinks" the console ishmmm. And finally, when using the "console redirection" BIOS setting on these DEC machines, LILO fails to boot the machine from the hard-disk - only floppy will work. When attempting to boot from the HD it gets stuck at "LI", while disabling the BIOS setting will instantly fix the problem! How Bizare ;-) 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) Phone: 07-838-4764 8<--->8 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: BIND 8.1.1 and IP aliasing in Linux
On Thu, 18 Dec 1997, dA' Phucilage Phactory wrote: > > I have several Linux DNS servers, two of which have IP aliases on one > > interface. Recently I upgraded from BIND 4.9.6 to BIND 8.1.1. One of the > > subtle problems I have noticed is now the machines that have IP aliases > > don't automatically update a zone when sent a notify by the primary > > server. Under 4.9.6 they had no problems. > > > I've noticed that the debian package for bind 4.9.6 isn't to great > (no offense to whomever made it), so i downloaded bind-4.9.6-REL.tar.gz, > and manually did it myself. I have not and will not goto 8.x until they > resolve a lot of errors. Right now i am quite happy with 4.9.6, it works > as it should, and isn't a hassle what so ever. > > Now to your questions, could you possibly be forgetting to > increment your serial numbers in your databases? If so, that would > possibly explain lack of zone transfers to your secondaries. Also, setting > up a cronjob to pull records is a good idea. I have cron run > named-xfer with the correct parameters every 6 hours on my secondaries to > pull from myself. No sorry, I know the serial numbers are not the problem. If I do a 'ndc reload' then it does the zone xfer alright. As I said - the only difference between when it was receiving and processing successfully the notifies from primary servers and now (when it's not) is that now I've added an IP alias to one interface and let named auto-detect the interfaces (rather than specifying via 'listen-on' directives, which appears to be no problem if you believe the logs). BTW, I'm not using (and haven't been using) the debian BIND package(s) - I've built my own, but the system is otherwise debian 1.3.whatever (bo). But thanx for some interest... 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) Phone: 07-838-4764 8<--->8 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
BIND 8.1.1 and IP aliasing in Linux
I have several Linux DNS servers, two of which have IP aliases on one interface. Recently I upgraded from BIND 4.9.6 to BIND 8.1.1. One of the subtle problems I have noticed is now the machines that have IP aliases don't automatically update a zone when sent a notify by the primary server. Under 4.9.6 they had no problems. Points of interest: 1. One machine, which has always had aliases, stopped doing updates on receiving notifies as soon as it was upgraded to 8.1.1. 2. Another machine was doing the notifies fine under 8.1.1 UNTIL we added the IP alias, and now it too ignores them. 3. I haven't used any "listen-on" directives in /etc/named.conf, though the default behaviour of using all available ones seems to be working (and noted in the log file). Is this a documented problem with either Linux IP aliases or BIND 8.1.1? I'd be grateful to hear. 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) Phone: 07-838-4764 8<--->8 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
HP DeskDirect 10/100TX PCI LAN Adapter
Are there Linux drivers for this card? I'm about to get one (if there are drivers) and would like to hear of experiences... TIA 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) Phone: 07-838-4764 8<--->8 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Problem with booting after moving from 128 to 64M
On Fri, 12 Dec 1997, Daniel J. Mashao wrote: > I had my system working fine for about 2 months but lately I have been > experiencing problems. I initially thought it had to do with "dirty" > filesystems. Now I have excluded that possibility. > > Now I wanted to test memory (128M) by removing 64M at a time and changing > the lilo.conf mem statement. After I have done that the computer does not > boot anymore. It stops after line "VFS: mount root ...". If I replace all > 128M it boots fine, but it does not work with any combinations of 64 Megs > > Any clues? This looks very similar to what happened to me last week. I had a system that had 384MB RAM. I needed to downgrade it to 128MB. I changed the line in /etc/lilo.conf and built a boot floppy, BUT I forgot to change the script /usr/sbin/mkboot to reflect this change. The system would hang whenever it got to the "VFS: mount root ..." line during boot just as you describe. I had to: 1. boot off a debian rescue disk 2. manually mount the rootfs on /mnt (after switching to the shell on Alt-F2). 3. put a blank floppy in, and 4. cat /mnt/vmlinuz > /dev/fd0 5. dismount /mnt and reboot. This got my system back booting from a raw kernel on the floppy, but only seeing the default maximum of 64MB RAM. I then edited /usr/bin/mkboot to include: append = "mem=128M" and reran it. Then a reboot got me back to the 128MB RAM as intended. A bit of a hassle alright! 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) Phone: 07-838-4764 8<--->8 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Pentium II supported?
This may be a dumb question, but here goes: Do Pentium II chip machines run Linux? For instance the kernel build procedure (e.g. make xconfig) only has options for up to Pemtium Pro machines, which may be just a case of someone not having got-around-to putting the options in yet, or is there a bigger reason like "it doesn't go yet"? I am about to order a new machine and will get a Pentium II if Debian can run on it ;-) 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) Phone: 07-838-4764 8<--->8 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
dhcpd config
Can anyone tell me if they've got dhcpd going on Debian? I'm continually getting the message: dhcpd: No subnet declaration for eth0 in my logs when it tries to start, and it fails immediately. But I do have such a statement! What's missing? TIA 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) Phone: 07-838-4764 8<--->8 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Large buffers entry in free
On Mon, 3 Nov 1997, Joey Hess wrote: > > Is there a bug somewhere? > > Yes. This is fixed in the procps currently in hamm, version > 1.12.2.1. Are you using a hamm system that hasn't been updated for a > week or so? I didn't think bo's procps had this problem too.. Oh yes it is! I have several bo systems that I happily upgraded and all have this problem. Fortunately it doesn't break anything other than what you read, but I'd like it to be fixedespecially when I'm telling non-linux boffins around here how to see how much memory is used on my linux servers and this garbage shows up! 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) Phone: 07-838-4764 8<--->8 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Netatalk, changing passwords
I am using debian netatalk 1.4b2, and have used it for some time happily. Until now I have always had Mac users who also had interactive accounts, and so could change their passwords at the shell prompt. However I am now setting up a couple of servers for users who may never log in. I notice that when I look in the chooser at my servers (under apple-share) the "change password" button is shaded out and unuseable. Is it possible to fix this somehow? Is this what PAM is supposed to help in? I'm really ignorant about what/how PAM is/works. Cheers, 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) 8<--->8 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Year 2000 & Debian
Not sure if this has been thrashed out before: Is Debian (or Linux in general) year 2000 *safe*? I'm not even sure what that means precisely, but I'm responsible for finding out round here and wondered if it's been discussed on this group. 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) Phone: 07-838-4764 8<--->8 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Sambades on non-us site
On Wed, 15 Oct 1997, Matthew Tebbens wrote: > The Sambades package on the overseas site has permissions problems. > Anyone notice this ? Is it on purpose ? > > Matthew Yes, my mirror has failed to pick this up for the last 2 days because of this. I REALLY want it for about 3 servers here. 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) 8<--->8 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Debian Netatalk and Ethernet cards
On Thu, 9 Oct 1997, LeRoy D. Cressy wrote: > > Further to this problem I have experimented some more and have the > > following results, all on the one machine, using different cards, > > attempting to run netatalk on the same network: > > > > Card Description: Driver (Module):Results: > > - > > Intel EtherExpressPro 100 (builtin) eepro100BAD > > Digital 21140 tulip BAD > > Western Digital 8013EPC wd Excellent > > > > While all work fine for regular IP (the first 2 especially nice as they do > > 100Mbps) only the last one is consistently good for seeing the (large) > > appletalk network we have here...weird, annoying and frustrating... > On ot boxes that do not support apple talk, did you do a kernel compile > specifying in the net section to include apple talk in the kernel? I Er, if you read above, all those tests were on the same machine! (So they must have had appletalk in the running kernel.) I ALWAYS recompile the kernel to suit my setup. The problem seems to be in the multicast part of the ethernet card drivers...I'll await new versions for improvements. 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) Phone: 07-838-4764 8<--->8 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
"free" command stuffed
I have just updated my servers to the latest stuff in bo (arrived on my mirror on Saturday). The free command now gives weird output, like this: ns1# free total used free sharedbuffers cached Mem:386536 380196 6340 22472 202704 -/+ buffers: 1774923997940 Swap: 130748380 130368 69604 The second line figures do not add up with the first line---according to the second line I have over 4GB of RAM (I wish!). The other machines report similar crap. Also the Swap line has an extranumber at the end that I think doesn't belong! Have others noticed this? 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) Phone: 07-838-4764 8<--->8 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Debian Netatalk and Ethernet cards
On Thu, 9 Oct 1997, Richard L Shepherd wrote: > Further to this problem I have experimented some more and have the > following results, all on the one machine, using different cards, > attempting to run netatalk on the same network: > > Card Description: Driver (Module):Results: > - > Intel EtherExpressPro 100 (builtin) eepro100BAD > Digital 21140 tulip BAD > Western Digital 8013EPC wd > Excellent > > While all work fine for regular IP (the first 2 especially nice as they do > 100Mbps) only the last one is consistently good for seeing the (large) > appletalk network we have here...weird, annoying and frustrating... AND, further to this I have settled for a solution which goes like this... keep the IP going over the eepro100 (built-in), I have a Western Digital (old-cheapo) configured with 10.0.0.1 and netatalk running on it. Even though the IP number is bogus (I guess it just needs to be there to have the interface UP) the appletalk routing seems to work fine. Thus it can be a High-performance WWW server out the eepro100 and the Macs can upload there WWW pages via the WD card---not ideal but at least I didn't have to spend any more money... Cheers, 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) Phone: 07-838-4764 8<--->8 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Debian Netatalk and Ethernet cards
On Thu, 9 Oct 1997, Richard L Shepherd wrote: > I happily use netatalk on 3 debian machines using ethernet cards which all > use the wd driver module. Today I go to set it up on a DEC XLe6180 which > has a builtin Intel EtherExpress 100 card in it. The IP drivers work > great at both 10 and 100 Mbps. However Netatalk just isn't happy, it > can't see the appletalk network out there (well it kind-of worked once out > of many tries---picked an invalid net number and could aecho some other > machines...). > > I didn't think it was relevant which ethernet driver I was using, but > eventually, after comparing with the other machines I have, it all pointed > to the ethernet card and driver being the culprit! Sure enough someone > tells me that appletalk won't go on any Intel cards, is this true? If so, > are there any other cards people know of that I should stare clear of to? > I'm installing this server for lots of people with Macs to share files > with... Further to this problem I have experimented some more and have the following results, all on the one machine, using different cards, attempting to run netatalk on the same network: Card Description: Driver (Module):Results: - Intel EtherExpressPro 100 (builtin) eepro100BAD Digital 21140 tulip BAD Western Digital 8013EPC wd Excellent While all work fine for regular IP (the first 2 especially nice as they do 100Mbps) only the last one is consistently good for seeing the (large) appletalk network we have here...weird, annoying and frustrating... 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) Phone: 07-838-4764 8<--->8 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Debian Netatalk and Ethernet cards
I happily use netatalk on 3 debian machines using ethernet cards which all use the wd driver module. Today I go to set it up on a DEC XLe6180 which has a builtin Intel EtherExpress 100 card in it. The IP drivers work great at both 10 and 100 Mbps. However Netatalk just isn't happy, it can't see the appletalk network out there (well it kind-of worked once out of many tries---picked an invalid net number and could aecho some other machines...). I didn't think it was relevant which ethernet driver I was using, but eventually, after comparing with the other machines I have, it all pointed to the ethernet card and driver being the culprit! Sure enough someone tells me that appletalk won't go on any Intel cards, is this true? If so, are there any other cards people know of that I should stare clear of to? I'm installing this server for lots of people with Macs to share files with... 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) Phone: 07-838-4764 8<--->8 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
SCSI support
Hi, does anyone know if there is support for: Initio PCI 9100U Ultra Fast SCSI Host Adapter Kit within the standard debian kernel-source packages? Or any linux support for this card at all? TIA... 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) 8<--->8 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Best Ethernet Card
On 12 Sep 1997, Dale Martin wrote: > The other card I see mentioned a lot, but never see much as far as > recommendations, is the Intel EtherExpress 100. Anyone have any > thoughts on those? I have the "Intel EtherExpress 100 Pro", which is different from the regular one (requires a driver that is not currently distributed with Debian) and it goes fantastic. I tried it out on our Cisco Catalyst 5000 switch and got up to 90Mb transfers with a 100MB file. Even though the driver I have is only in alpha, we use it in production (on a major nameserver) and it has never put a foot wrong. 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) 8<--->8 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Max Process size
Can anyone tell me where in the kernel sources are specified any of the following: max size of a process max number of processes I once knew, but alas that was a long time ago...! I have kernel-source-2.0.29-7 installed 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) 8<--->8 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: why kernel 2.0.29 in Debian 1.3.1?
On Fri, 25 Jul 1997, Kevin Traas wrote: > > why Debian 1.3.1 includes kernel 2.0.29 and not 2.0.30 (that is in this > > Debian 1.3.1 but don't is preselect the first time that I install the > > system).? > > > > Is there any problem whith 2.0.30 kernel? > > Apparently, there was some of the 2.1.x code that was introduced into > 2.0.30 that caused some problems I've no idea what it was, maybe > others can comment; however, it caused the Debian Powers-That-Be to elect > to go with 2.0.29 as the default. However, 2.0.30 is still included in the > 1.3.x distribution. > > I've been using 2.0.30 on several machines without difficulty > Therefore, I've never investigated further as to what problems were present > in this newer version. > > 2.0.31 is supposed to be available within days. This is supposed to > resolve the 2.0.30 problems. Hmmm, I have 2 problems at present: 1. Cannot boot: originally got the 1FA: prompt. Changed some settings in the SCSI BIOS and then it worked (2.0.30 kernel). I recompiled the kernel one more time and from then on it always stops at "LI". I can't fix this one at all, even with 2.0.29. 2. With 2.0.30 I was witnessing a memory leak. However back with 2.0.29 it seems to be OK (or at least the leak is much slower!). So all-in-all 2.0.29 seems to be better. I guess the boot problem is lilo's rather than a specific kernel versionAny helpers here? 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) 8<--->8 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
latex2html and lyx
I have installed both lyx (0.10.7-3) and latex2html (96.1.h-3). I wish to write documents in lyx and reproduce them in HTML via latex2html. Trouble is: when lyx makes a "nice" latex file instead of "{\bfseries bla bla}" it produces "\bfseries bla bla\mdseries" which latex2html then munges into "bla bla" and then at the close of the environment closes with a "". Hence I get line breaks where I don't want them and bold continuing to the end of the environment! Is there a fix for this? Either in lyx or latex2html? 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) 8<--->8 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: samba and windows NT
On Tue, 22 Jul 1997, Syd Alsobrook wrote: > I'm using debian 1.3 with samba 1.9.16. > > When my main workstation was running Win95 I could connect fine now I'm > running NT 4.0 and I can browse the to my linux box but I'm getting a login > error that says: The account is not authorized to login from this station. I have a colleague here who can no longer mount, via samba, his home directory since he applied service-pack 3 to his NT 4.0 machine. It's a big problem here because suddenly lots of people can no longer access their samba volumes on our main WWW (unix) servers the way they used to. I don't yet know of a satisfactory solution. Does anyone else? 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) 8<--->8 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: 100 Mbit Ethernet
On Thu, 10 Jul 1997, dpk wrote: > I don't have much experience with having two ethernet cards in a machine, I've a little now that I've had to play with a few machines (one has 4 ethernet cards and 1 radio card!) > My ifconfig also says 10mb because the tulip driver first probes 10mb. > However, after boot up, once I start sending packets it says "Enabling > 100TX port" despite the fact it says 10mb on ifconfig. Also, my download > capabilities have shown it. ( On our internal network I can download up > to 1.3 MB/sec =), my scsi hard drive griding as fast as possible the whole Well I've got this problem solved now. It seems that the switch port I was testing on was stuffed. Moving to another one I get wonderful performance. On a 100Mb/s port I can get it working at both half and full duplex, with ftp transfers of 3.8 MB/s (which translates to about 30Mb/s, the other machine for the ftp is only on half duplex) on a 100 MB file which is excellent! The other card is still happy at 10Mb/s. And it's all autodetect too, no options needed. 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) 8<--->8 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: 100 Mbit Ethernet
On Tue, 17 Jun 1997, dpk wrote: > I actually did this today... I replaced my old bnc/twisted pair SMC card > with a 10/100 Switchable SMC which has the 21040 DEC chipset on it (tulip > driver). We are using a 100 Mbit hub and have noticed a great deal speed > increase with it. It seems as if the only thing that slows me down are my > scsi harddrives when downloading large files! > > They still use the same drivers for the corresponding chipsets. In my > case I didn't have to even recompile my kernel because the DEC 21040 and > the 21041 use the same driver. I hope you have the same success and > pleasure as I. Not quite. I have a machine with two DEC 21140 chip ethernet cards in it. I wish to have one at 100Mb and the other at 10Mb. I have the first one plugged into a port on a Cisco Catalyst 5000 switch. The switch detects 100Mb full-duplex, however ifconfig still says 10Mb and I get huge packet loss, presumably due to the speed mismatch. How do you get the cards to go at 100Mb? > On Wed, 18 Jun 1997, Richard L Shepherd wrote: > > > > This is just a quickie, and I'm sure it's an easy one too: > > > > Does Linux support any/all 100 Mbit capable ethernet cards? i.e. is it > > possible to use 100 Mbit ethernet with Linux? > > 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) 8<--->8 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: [Off-topic] AIC 7880 supported?
On 9 Jul 1997, Eloy A. Paris wrote: > sorry for this off-topic question. If someone provides me with the > mailing list I should subscribe to ask these kind of questions I will > be very grateful... > > I want to know if the AIC 7880 Ultra Wide SCSI adapter is supported > by Linux. Also, is there any advantage in using a Pentium II instead > of a normal Penitum or even a Pentium Pro? > > I am looking at a Dell PowerEdge 2200 machine that is going to be > a production file/Internet server. This machine has an AIC 7880 and a > Pentium II. I have just got a Digital Prioris Xl 6200 which has AIC 7880 on board (integrated) and it works fine. I just booted up with the Debian Rescue disk and never looked back. 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) 8<--->8 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Switching IP aliases
Well I have now written a script that does this quite nicely. Only Solaris still rebels to seeing the ARP updates. Further investigation shows that Solaris is breaking the rules here so I'll not bother any more with it. Email me if you have a need for this. Cheers. On Sat, 28 Jun 1997, Richard L Shepherd wrote: > I wish to have 2 nearly identical machines. Each will have it's own IP > number plus an IP alias. The idea is that I wish to occasionally take 1 > down and give it's IP alias over to the other machine, so that there is no > visible change to any clients. > > The problem is that machines on the localnet may have an entry in their > ARP tables with the ethernet address of the old machine at the time of the > changeover, and so be unable to reach the new holder of that IP number. > It is necessary for the new machine to send some packets with a source > address of the new alias it has taken on so that local hosts will see the > new ethernet address and update their ARP tables appropriately. > > The only way I've found to do this is to alter the routing so that the new > IP alias becomes the gateway to the localnet. (host routes are then added > for the primary IP address to restore that part of things) Outgoing > packets will now have the right source address. It sends some packets out > and then switches the routing back to normal so that the IP alias is ready > to be given back. This seems to work OK, but I'd really rather not stuff > with the routing, so: 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) 8<--->8 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Switching IP aliases
On Sat, 28 Jun 1997, Lindsay Allen wrote: > Richard, > > Would it fix things if you were to ping a non-existant host, thus causing > a broadcast from your box? Maybe ping to various subnets as required? Good idea. I now have two debian linux machines passing an IP alias between them. I get the one which has just taken on the alias to ping an unused IP number on the localnet, after setting the alias interface to be the gateway to the localnet. I then swap the routing back to normal. I'm observing what other machines on the localnet do with-respect-to their arp tables. These other machines are: AIX 4.2 Solaris 2.5 Win 95 Linux Digital Unix (OSF) Cisco router All of these happily update their arp caches as I switch the aliases EXCEPT for Solaris! It's brain-dead! It continues to cache entries for ages. I don't know what to do about that! > I do not have much experience here. I'm just learning to, with the help of some local network gurus. 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) 8<--->8 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Switching IP aliases
I wish to have 2 nearly identical machines. Each will have it's own IP number plus an IP alias. The idea is that I wish to occasionally take 1 down and give it's IP alias over to the other machine, so that there is no visible change to any clients. The problem is that machines on the localnet may have an entry in their ARP tables with the ethernet address of the old machine at the time of the changeover, and so be unable to reach the new holder of that IP number. It is necessary for the new machine to send some packets with a source address of the new alias it has taken on so that local hosts will see the new ethernet address and update their ARP tables appropriately. The only way I've found to do this is to alter the routing so that the new IP alias becomes the gateway to the localnet. (host routes are then added for the primary IP address to restore that part of things) Outgoing packets will now have the right source address. It sends some packets out and then switches the routing back to normal so that the IP alias is ready to be given back. This seems to work OK, but I'd really rather not stuff with the routing, so: Is there a better way? 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) 8<--->8 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
100 Mbit Ethernet
This is just a quickie, and I'm sure it's an easy one too: Does Linux support any/all 100 Mbit capable ethernet cards? i.e. is it possible to use 100 Mbit ethernet with Linux? I have no such machines (though I have some that drive 100 Mbit capable cards at 10 Mb, since that's what the nets they're on are at). I'm just checking to make sure there are 100 Mb drivers too... Please reply by email... 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) 8<--->8 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: gated file
I have now compiled gated (tried on another machine and it went no problems, thought they were the same configstrange), and got it running---good. I have installed it on our internet exchange and it listens to BGP nicely, including supernets---very good (no other OS's can do that, that we have). However: it does not seem to understand variable length subnet masks, as given to it by the BGP server? Instead it sets the BGP server as it's next hop for those nets. I would MOST like to solve this. Is it a problem with gated (I used "options VARIABLE_MASKS"), or is it a kernel feature I need to explicitly turn on? The way most of our other hosts cope with this problem is that they receive a redirect from the BGP server so that they get the correct route next time. However my linux machine doesn't seem to learn these redirect routes and continues to go the long way for these nets every single time! This is not fatal, just rather inefficient! This would not be an issue if the first problem was solved but, if it can't be, this would be a second best option ;-) Does anyone have solutions to these? PLease reply by email as well...TIA 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) 8<--->8 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
gated
Does anyone know of a good set of sources AND patch(es) (if necessary) for gated for linux? I mirror the gated sources here and have just tried building it on Debian 1.2, kernel 2.0.29 (from bo) using the patches for gated 3.6A2 I got some time ago from ftp.redhat.com. I guess it would be even better if gated was packaged for Debian (hint hint) but I'm happy to build it from sources IF it'l build! I have previously successfully done this (redhat 3.0.3, kernel 2.0.0) but now it seems broken (with my more up-to-date Debian). Anyone got pointers to better stuff than me? TIA 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) 8<--->8 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: IP-Accounting HOWTO?
On Thu, 22 May 1997, Heiko Schlittermann wrote: > On May 22, Benedikt Eric Heinen wrote > : I'd need to get a finer split in traffic list, I'd need to get an > : accounted list indicating each service and each new destination > : separately. Is there a package to do that yet? > > Try net-acct. There's a debian package out there. Nice thing, but > produces hge amounts of data. > > It logs > > date / protocol / source_ip:port / dest_ip:port / packets / bytes / user > > for every packet travelling on the connected ethernet segment (putting > the interface in promiscous mode). You can exclude / include specific > networks / addresses. But is it possible to exclude traffic based on source and/or destination IP separately? What we'd like to do is exclude any traffic where BOTH source and destination are local, and record all other traffic (e.g. for international traffic billing, where if either source or dest are non-local we want it counted, but not otherwise). My reading of the docs gave me the impression that one could exclude traffic involving certain nets but it wasn't fussy about what else the traffic involved...see what I'm getting at? 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) 8<--->8 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Terminal not powerful enough for SLang
On Fri, 23 May 1997, Craig Sanders wrote: > > > $ unset termcap > > > > Sorry to say but that doesn't do it either. Any other ideas? NB: It > > goes fine from a virtual console and an rxvt, just not an xterm. I'm > > guessing that it's the colors that an xterm is not capable of that is > > the problem. > > Karl was right, except that the variable to unset is "TERMCAP", not > "termcap". sh variables are case sensitive. Yep that works (should have guessed that bit myself!). I have also tried setting TERM to xterm-color and that looks real nice too. Thanx for all the help everyone! 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) 8<--->8 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Terminal not powerful enough for SLang
On Fri, 23 May 1997, Toens Bueker wrote: > On May 23, Richard L Shepherd wrote > > > > $ unset termcap > > > > Sorry to say but that doesn't do it either. Any other ideas? > > NB: It goes fine from a virtual console and an rxvt, just not an xterm. > > I'm guessing that it's the colors that an xterm is not capable of that is > > the problem. > > Try "export TERMCAP=" Thanx, that did it! 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) 8<--->8 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Terminal not powerful enough for SLang
On Wed, 21 May 1997, Karl M. Hegbloom wrote: > >>>>> "Richard" == Richard L Shepherd <[EMAIL PROTECTED]> writes: > > Richard> Does anyone know how to fix the problem of trying to use > Richard> a program (such as lynx or slrn) that uses SLang from > Richard> reporting the error: Terminal not powerful enough for > Richard> SLang and failing there? > > $ unset termcap Sorry to say but that doesn't do it either. Any other ideas? NB: It goes fine from a virtual console and an rxvt, just not an xterm. I'm guessing that it's the colors that an xterm is not capable of that is the problem. 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) 8<--->8 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Terminal not powerful enough for SLang
Does anyone know how to fix the problem of trying to use a program (such as lynx or slrn) that uses SLang from reporting the error: Terminal not powerful enough for SLang and failing there? I just did an update through dselect (using stable) and got a new version of lynx which told me to get SLang (which I did) but now when I start lynx I get the error message above. I remember getting this ages ago when trying to use slrn to read news. I never solved it then---just went to using pine instead! Now I want my lynx back ;-) Any takers...? 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) 8<--->8 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Other remote passwd servers?
On 18 May 1997, Manoj Srivastava wrote: > Hi, > > I make a living off Open Groups DCE (used to be OSF), and > billing it as an NIS replacement is akin to comparing a pocketknife > to a thermonuclear device. It would be arguably more accurate to say > DCE is nothing like NIS. Say what you like, our varsity IS going to use it as a replacement for NIS! (Maybe this says something about the admin here ;-)) It's not my doing...I'm just smallfry around here... 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) 8<--->8 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Other remote passwd servers?
On Fri, 16 May 1997, Karl Ferguson wrote: > Hi... > > Other than NIS itself, are there any other NIS-like administrations out > there? From and administrators point of view, NIS works great - but users > can simply examine the passwd file by using ypcat (sure, all be-it a > mangled passwd field) - is there any way to stop that? (Hence my question > for other NIS-like things). There's DCE, Distributed Computing Environment, from OSF. It is a very powerful system (so I've read) for tying multi-platform networks together under one authentication service. Our varsity's about to dive in and use it. Unfortunately I don't know if there is a Linux client for it. It is touted as "vendor neutral", and has a fairly impressive list of supported platforms: AIX AT&T GIS Cray Unicos . . . Digital Unix HP-UX Irix . . . SCO and on it goes. It also includes NT, Macintosh and other non-unixes. I'm really bummed out that Linux is not in this list (but maybe the list I have is incomplete?). Maybe, just maybe, the SCO version could be persuaded to operate in Linux via iBCS, but I've no idea what that entails... 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) 8<--->8 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Memory Gobbler
> At 09:20 AM 4/21/97 +1200, Richard L Shepherd wrote: > > >Yes I have read that too. However it does see the memory (when I put > >"mem=128M" on the boot line). I'm not sure that cache isn't the problem, > >though. It went so well for 3 weeks, then started to go downhill. This > >w/e it killed itself completely! I arrived yesterday afternoon to find > >the screen filled with messages "out of memory for bash" and other > >commands too. There was no response from any means, so I had to push the > >reset button and start againbummer! Well the answer to this has come in upgrading from 2.0.27 (as in rex) to 2.0.30 (as in hamm). I manually 'dpkg -i'ed the kernel source package from hamm and left the rest of the system as rex. So far no memory leaks (2 weeks that is...). Thanks for the interest anyway... 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) 8<--->8 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Wavelan cards and PCI ethernet
Well I solved this problem very simply: get a machine whose ISA slots actually work! After having successfully used the WD8003 card on 4 other machines, but not this one, I decided there must be something wrong with this one. So I tried another machine and it all went first pop! On Thu, 24 Apr 1997, Richard L Shepherd wrote: > I am trying to install a wavelan card into a linux box. The setup is: > Digital Celebris XL 5166 > Digital DE450 PCI Combo Ethernet Card (using tulip driver) > Western Digital (wd8003) 8-bit (el-cheapo) Ethernet card (using wd driver) > Built-In NCR SCSI Host (using 53c7,8xx driver) > Adaptec 2940 Ultr-Wide PCI SCSI (using aic7xxx driver) > (some PCI video card, who cares) > > The wavelan card I want to install is: > NCR Wavelan v1.04 (930406), IRQ 10, Port 300 > > I have tried having the drivers for the 2 ethernet cards, and the wavelan > card, as modules, but the wavelan card is not recognised initially and the > DE450 is assigned IRQ 10 (is that how PCI does it, just looks for a free > one?) and so I cannot load the wavelan driver anymore? > > I tried another tack: compiled the wavelan driver into the kernel and left > the other 2 as modules, sure enough it now assigns the DE450 to eth1 but > still uses IRQ 10, so I'm still stuffed. > > I noticed that the driver for the wavelan probes only at 0x390, whereas I > believe this card is at 0x300 so I altered the wavelan.c file to match > this, but still no go! 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) 8<--->8 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Wavelan cards and PCI ethernet
Hi, I am trying to install a wavelan card into a linux box. The setup is: Digital Celebris XL 5166 Digital DE450 PCI Combo Ethernet Card (using tulip driver) Western Digital (wd8003) 8-bit (el-cheapo) Ethernet card (using wd driver) Built-In NCR SCSI Host (using 53c7,8xx driver) Adaptec 2940 Ultr-Wide PCI SCSI (using aic7xxx driver) (some PCI video card, who cares) The wavelan card I want to install is: NCR Wavelan v1.04 (930406), IRQ 10, Port 300 I have tried having the drivers for the 2 ethernet cards, and the wavelan card, as modules, but the wavelan card is not recognised initially and the DE450 is assigned IRQ 10 (is that how PCI does it, just looks for a free one?) and so I cannot load the wavelan driver anymore? I tried another tack: compiled the wavelan driver into the kernel and left the other 2 as modules, sure enough it now assigns the DE450 to eth1 but still uses IRQ 10, so I'm still stuffed. I noticed that the driver for the wavelan probes only at 0x390, whereas I believe this card is at 0x300 so I altered the wavelan.c file to match this, but still no go! Could anyone help me here please? TIA, 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) 8<--->8 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Memory Gobbler
On Tue, 22 Apr 1997, Matt Lawrence wrote: > I had similar problems with the kernel on the install disks. If I let the > system sit for a day or so (386 with 8M), I would get a bunch of "Couldn't > get a free page" messages. After rebuilding the kernel for my hardware, > the message hasn't come back. I've also seen this message on a 486 with > 16M, it just took a couple of days to show up. I beleive that there is > some sort of a memory leak in the kernel as distributed. I have rebuilt a kernel for my own hardware (always do, often do it a couple of times while reorganising what gets compiled in and what's loaded as a module). So if there's a memory leak (which it seems there is, though it took some time to show up), it's not solved just by avoiding the distribution kernel. 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) 8<--->8 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Memory Gobbler
On Sat, 19 Apr 1997, Harmon Sequoya Nine wrote: > I was looking at how to recompile the kernel to get it to recognize memory > above 64M, > and it said you should have at least 512K of cache to do this... Yes I have read that too. However it does see the memory (when I put "mem=128M" on the boot line). I'm not sure that cache isn't the problem, though. It went so well for 3 weeks, then started to go downhill. This w/e it killed itself completely! I arrived yesterday afternoon to find the screen filled with messages "out of memory for bash" and other commands too. There was no response from any means, so I had to push the reset button and start againbummer! Another thing that might be relevant is what I read in "man bootparam". Linus says that some machines might reserve the top of RAM for BIOS cacheing (or something like that) and so to allow for this, I've booted up this time with "mem=126M". I thought that maybe I tramped all over some important memory beforemaybe I'm completely off track. I am trying to get some more cache, but it's proving difficult. 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) 8<--->8 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Memory Gobbler
I have 128MB RAM and 256KB cache. My machine has been up 32 days. For the first 3 weeks it went really nicely, didn't get anywhere near using any swap. However this week it's usage has got over 100MB (now up to 120MB, after deducting buffers and cached, i.e. the second row of "free" output) and there's plenty of swap being used! I just can't figure out what's eating it. I wrote a script to total the figures produced by "ps aux" in the VSZ and RSS columns. At present these come to: 192452 20612 respectively. To me this means there is only 20MB of RAM actually used by processes, and the 192MB of virtual memory is probably not correct because some is shared. So: what's eating it? If it's any help, here's the output of some commands: % free total used free sharedbuffers cached Mem:127756 126048 1708 17040304 3528 -/+ buffers: 122216 5540 Swap: 128484 27208 101276 % ps aux USER PID %CPU %MEM VSZ RSS TT STAT START TIME COMMAND daemon 210 0.0 0.0 832 0 ? SW Mar 17 0:08 rpc.portmap dennis 844 0.0 0.0 1452 0 pf SW Apr 16 0:00 bash dennis 880 0.0 0.8 3492 1032 pf SApr 16 0:16 ical dmneal 17887 0.0 0.0 1072 0 q4 SW Mar 27 0:00 csh nobody1586 0.0 0.0 100460 ? SApr 15 0:00 httpd nobody1587 0.0 0.0 100456 ? SApr 15 0:00 httpd nobody1588 0.0 0.0 100460 ? SApr 15 0:00 httpd nobody 21109 0.0 0.3 3328 440 q4 S N Apr 15 0:23 cached nobody 21110 0.0 0.0 816 0 q4 SWN Apr 15 0:00 dnsserver nobody 2 0.0 0.0 816 0 q4 SWN Apr 15 0:00 dnsserver nobody 21112 0.0 0.0 816 0 q4 SWN Apr 15 0:00 dnsserver operator 2132 0.0 0.0 1484 0 q6 SW Apr 16 0:00 bash operator 2245 0.1 0.7 4364 920 q6 SApr 16 4:09 wish -f /usr/bin/ex richards 1053 0.0 0.2 1484 300 2 S 09:07 0:00 bash richards 5020 0.0 0.3 1484 504 4 SMar 20 0:00 -bash richards 8248 0.0 0.7 2420 980 2 S 10:18 0:00 pine richards 9166 0.0 0.3 928 440 4 R 10:27 0:00 ps aux richards 31895 0.0 0.0 1488 0 pb SW16:41 0:00 bash root 1 0.0 0.0 81216 ? SMar 17 41:31 init root 2 0.0 0.0 0 0 ? SW Mar 17 0:03 kflushd root 3 0.0 0.0 0 0 ? SW< Mar 17 4:42 kswapd root 4 0.0 0.0 0 0 ? SW Mar 17 0:00 nfsiod root 5 0.0 0.0 0 0 ? SW Mar 17 0:00 nfsiod root 6 0.0 0.0 0 0 ? SW Mar 17 0:00 nfsiod root 7 0.0 0.0 0 0 ? SW Mar 17 0:00 nfsiod root13 0.0 0.0 78848 ? SMar 17 17:58 update root 196 0.0 0.2 1084 336 ? SMar 17 1:50 /sbin/syslogd root 198 0.0 0.0 964 0 ? SW Mar 17 0:01 klogd root 206 0.0 0.0 80048 ? SMar 17 0:00 /sbin/kerneld root 212 0.0 0.0 81224 ? SMar 17 0:13 inetd root 281 0.0 0.0 936 0 ? SW Mar 17 0:02 au root 291 0.0 0.0 1068 124 ? SMar 17 4:11 /usr/sbin/snmpd -f root 303 0.0 0.0 92476 ? SMar 17 1:52 /usr/sbin/rpc.nfsd root 305 0.0 0.0 88872 ? SMar 17 0:10 /usr/sbin/rpc.mount root 309 0.0 0.0 84488 ? SMar 17 13:49 /usr/sbin/atalkd root 319 0.0 0.0 89252 ? SMar 17 0:10 /usr/sbin/afpd root 322 0.0 0.0 828 112 ? SMar 17 1:22 /usr/sbin/cron root 329 0.0 0.0 200480 ? SMar 17 0:08 xdm root 331 0.0 0.0 804 0 1 SW Mar 17 0:00 getty root 335 0.0 0.0 804 0 5 SW Mar 17 0:00 getty root 336 0.0 0.0 804 0 6 SW Mar 17 0:00 getty root 843 0.0 0.0 1076 0 ? SW Apr 16 0:00 in.telnetd root 1428 0.0 0.0 274080 ? S 17:03 0:01 xterm root 1487 0.0 0.3 2740 420 ? S 17:04 0:01 xterm -bg LavenderB root 2131 0.0 0.0 1076 0 ? SW Apr 16 0:00 in.telnetd root 2443 0.0 0.0 283620 ? SApr 16 0:01 xterm root 2780 0.0 0.0 274064 ? SApr 14 0:01 xterm root 3537 0.0 0.1 1232 232 ? SMar 18 0:36 sendmail: accepting root 3766 0.0 0.0 848 0 ? SW Mar 20 0:00 papd root 4518 0.0 0.1 2044 208 ? SApr 11 0:01 -massiveduck.cc. root 7126 0.0 0.0 1048 0 q4 SWN Apr 9 0:01 RunCache root 9156 0.0 0.0 294832 ? SApr 14 0:15 xterm root 9161 0.0 0.3 1456 416 ? S 10:27 0:00 bash ./monitor.sh m root 9162 0.0 0.2 864 360 ? S 10:27 0:00 /bin/ping -q -i 1 - root 9163 0.0 0.2 896 272 ? S 10:27 0:00 fgrep % root 9164 0.0 0.2 808 268
RE: Linux in an NT domain
On Wed, 16 Apr 1997, Carpenter, Dean (MS Mail) wrote: > Yup - I think it's called smb-NT-verify.1.1.tar.gz. I don't remember > exactly where I got it, but here's the authors name from the > nt-linux.readme file ... > > Christopher Burke| Unique | Web: http://www.mindware.com.au > Mindware | Custom | Info: [EMAIL PROTECTED] > PO Box 9, Richlands | Innovative | E-Mail: [EMAIL PROTECTED] > Q. 4077, Australia | Development | Telephone:Australia 0412-696290 > > Check the website listed above - that has more information. Thanx, I found it, it's at: http://www.mindware.com.au/ftp/smb-NT-verify.1.1.tar.gz Cheers, 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) 8<--->8 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Linux in an NT domain
Does anyone know if Debian has packaged the package which allows a Linux box to participate in an NT domain? I read somewhere once that such a package exists, but am very vague as to the details of what it did. NB: it's not samba, it's do with sharing the user-space/passwords etc. Any ideas? 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) 8<--->8
Re: 'Frozen' boot disks
On Sun, 13 Apr 1997, Paul Wade wrote: > The reason I put this copy up was for testing. For the first round of > testing, people were picking them up from sites that only had modem > bandwidth. Yes, ftp.debian.org is the authoritative source. If you want to > play with the new toys before the store opens, you can pick them up at my > site. OK, that's cool then. I have no problems with that. I'll just keep doing what I am. I'm glad to have that clarified ;-) 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) 8<--->8
Re: 'Frozen' boot disks
On Sat, 12 Apr 1997, Paul Wade wrote: > > 4/6/97 I put a copy at ftp.greenbush.com, look in /pub/bodisks. The files > are dated by time of transfer, but they are the 4/4 set. > > > I have the same problem. I had thought it may be because I mirror a > > mirror (which in turn may not directly mirror ftp.debian.org) and so my > > mirror was effectively a few days behind. However I just checked on > > ftp.debian.org and it really is empty so > > > > Where are they then? So my next question then is: is not ftp.debian.org the authoritative source for the Debian Distribution? i.e. I seek to maintain a mirror of ftp.debian.org thinking that this will ensure I have an up-to-date Debian Distribution for my linux fans. Am I doing the right thing? 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) 8<--->8
Re: 'Frozen' boot disks
On Sat, 12 Apr 1997, Lamar Folsom wrote: > Is anyone else having trouble finding the disks for 1.3? I'd appreciate any > pointers. The 'bo/disks-i386/1997-04-04/' directory seems to be empty. I have the same problem. I had thought it may be because I mirror a mirror (which in turn may not directly mirror ftp.debian.org) and so my mirror was effectively a few days behind. However I just checked on ftp.debian.org and it really is empty so Where are they then? 8<--->8 Richard Shepherd ([EMAIL PROTECTED]) 8<--->8