From: Digestifier To: Linux-Admin@senator-bedfellow.mit.edu Reply-To: Linux-Admin@senator-bedfellow.mit.edu Date: Mon, 3 Oct 94 07:13:28 EDT Subject: Linux-Admin Digest #137 Linux-Admin Digest #137, Volume #2 Mon, 3 Oct 94 07:13:28 EDT Contents: Re: PCI vs. VLB (root) Hard Drive "sleep" program?? (Emarit Ranu) Routing table disappear :-(( (Ahmad Al-rasheedan) Re: Slip autoanswer (Gert Doering) Re: Signals howto wanted. (Pat O'Neil) Re: How do you see who logged in? (Bob Collie) Re: tar help (Bob Collie) Re: No Hostname (Patrick J. Volkerding) Re: Help INN on my Linux box! (Jerry Ablan) Re: HOW to transfer LARGE SINGLE file between Sun5.3 and Linux (David Fox) /sbin/hostname on 1.1.45- (Riku Saikkonen) Re: Assuring that a PPP link stays up... (Ralph Sims) Re: inetd seems to lock-up (Lee J. Silverman) Inn on a Linux box! (Nathan Stratton) Where can I get x11perf? (M. Shawn Easter) Re: What's failed after Bogomips (Christopher M. May) Computone MultiPort ??? (Tony Schwartz) xdm & xterm need help Re: 80x50 screen (Joel Storm) Setting up the BaseAddr for Mitumi (Jason Hong) Re: Ncurses signals broken? (Zeyd M. Ben-Halim) ---------------------------------------------------------------------------- From: root@mit.edu (root) Subject: Re: PCI vs. VLB Date: 3 Oct 1994 03:14:25 GMT Reply-To: jered@mit.edu Riku Saikkonen (riku.saikkonen@compart.fi) wrote: : Which system would be better to use for Linux (+X), a VLB motherboard : with a VLB display adapter, I/O card, and whatever, or a PCI system? Go with PCI for as many components as possible. PCI currently appears ot be the new PC bus standard. It performs better than VLB, and is prefered by manufacturers because it is easier to develop PCI cards than VLB ones. : I've heard that PCI would be somewhat faster; how much? Also, I've heard : that PCI isn't too stable yet, some cards won't work together; is this : right? PCI is faster, current 66 Mhz implementations are about 2x the speed of VLB, and PCI 2.1 will be 3 times the speed, and double the width. The PCI standard is pretty much stable, although the PCI and system BIOSes may not be. Card conflicts are not likely, but a conflict between card and BIOS may be. Almost all work, though. : Does the kernel support VLB or PCI IDE host adapters? Are they much : faster than ISA ones? The kernel can't tell the difference, really, between ISA, VLB or PCI. It currently supports the BusLogic BT946C SCSI host adaptor, and AHA27xxx/28xx/29xx support is not far behind. As for video, XFree86 3.1 (to be released this week, I'm told) supports several PCI video cards. I use the Diamond Stealth 64, which is based on the S3 Vision964 chip. One gotcha with PCI, DON'T GET THE OPTi CHIPSET! There is a bug with it (alas, I have the OPTi, and am trying to get it replaced.) such that bus performance will not be optimal. : A common PCI display adapter seems to be an 'AvanceLogic' (I don't know : of a model) card. Does XFree86 2.1.1 support this? Which chipset does it : have? Would I be better off with a Cirrus? (Which is better, a 5428 or a : 5434?) I haven't heard of this. You'll probably be safest with anything based on the S3 Vision964 chip. There's quite a list of these. : Does the kernel support the NCR SCSI-2 controller that is integrated in : some PCI motherboards? I found some NCR PCI drivers in drivers/scsi; do : they work with the integrated chip (unfortunately I've forgotten the : chip model number)? There exists an NCR53c810 driver for Linux, but it is a kernel patch. Although more expensive, I suggest the BusLogic card, it gives slightly better transfer rates, and is supported in the 1.1.18 kernel. (Make sure you recompile your kernel after install with ADAPTEC support DISABLED. Otherwise, the Buslogic will emulate the Adaptec, with worse performance than native mode.) Jered (jered@mit.edu) ------------------------------ From: drranu@lamar.ColoState.EDU (Emarit Ranu) Subject: Hard Drive "sleep" program?? Date: 2 Oct 1994 20:53:08 GMT I have a linux box that can be idle up to 12 hours at a time. I figure since the hard drive is not doing much, is there a program that will shut off the hard drive until it is needed again? I realize if there is such a thing "bdflush" would demolish it's purpose. If bdflush was changed to flush every hour or more, then maybe this could take wear off the hard drives? Answers/comments to both my concerns are more than welcome! -- -Emarit drranu@lamar.ColoState.EDU KG0CQ _._ __. _____ _._. __._ ------------------------------ From: asr@q8petroleum.com.kw (Ahmad Al-rasheedan) Subject: Routing table disappear :-(( Date: 2 Oct 1994 20:06:23 GMT I have managed to run PPP on linux; there is a small problem however. I use the ppp.go script to connect, I can only utalise the connection after I manually run routed. And after 4-5 minutes, the routing table is gone and I am left with lo0 as apposed to lo0 and ppp0. The remedy that I reached (sucks) is to kill the routed daemon and start it again. Now, doing this every 5 minutes is a lot of fun. I wonder if anybody came such a thing ?? would love to hear from their exp. I have followed the README in /usr/lib/ppp in filling out /etc/gateways and made it suitable to my hosts (work & home) Any tips ?? Thanx ------------------------------ From: gert@greenie.muc.de (Gert Doering) Subject: Re: Slip autoanswer Date: Sat, 1 Oct 1994 15:48:08 GMT bob4@slb.com writes: >> How do I get my modem to autoanswer the phone when it detects a ring ?? >At the risk of getting more abuse, I answered a similar question >the other day. My way is simply to turn on Auto Answer on the modem. >(Usually DIP switch 5) Well, definitely not "usually"... most modern modems don't have *ANY* dip switches at all. "Usually" would be "ATS0=1&W". The *exception* are USR and older Hayes modems that need to have a dip switch changed. >When I posted this the other day, I got a followup from someone saying >that was no good, because he didn't want the modem to answer the phone >when Linux was crashed. That was me. >I don't understand. I have a dialup system >here in the office which I have set up as a Slip server. The modem >is set to Auto Answer, and I don't have any problems. >Perhaps the person who replied before would care to expand his comments... Sure. Just imagine your slip server crashing, for whatever reason, leaving DTR high. OK, someone elses machine calls in. Your modem auto answers, he gets a CONNECT - and then? just a blank line, no welcome mesg... (remember, the machine is crashed!). If the callers machine is dumb, it retries the call... and retries... and retries... - imagine the callers phone bill. In addition, if you don't have the modem auto-answer the phone, but have a program answer incoming calls manually with ATA, you can do a lot other nice tricks with the modem, for example, answer only if the callers number (caller ID) matches a given list, log the CONNECT string, ... >BTW, you could try reading the modem manual ! (grin ..) That's always a good idea, agreed. gert -- Yield to temptation ... it may not pass your way again! -- Lazarus Long //www.muc.de/~gert Gert Doering - Munich, Germany gert@greenie.muc.de fax: +49-89-3243328 gert.doering@physik.tu-muenchen.de ------------------------------ From: poneil@infi.net (Pat O'Neil) Subject: Re: Signals howto wanted. Date: Thu, 29 Sep 1994 20:03:37 In article <367vs1$grq@panix.com> danw@panix.com (Dan Wold) writes: >From: danw@panix.com (Dan Wold) >Subject: Signals howto wanted. >Date: 26 Sep 1994 22:26:09 -0400 >I'm looking for an understanding on how signals work. None of my unix books >or accumulated FAQs, HOWTOs, etc, give much info on the topic. Can anyone >suggest any etexts or books that deal with the subject? O'Reilly and associates "Using C on the UNIX System" is a excellent reference to get you going on signals as well as several other topics. ------------------------------ From: collieb@iia.org (Bob Collie) Subject: Re: How do you see who logged in? Date: 2 Oct 1994 21:36:55 GMT Serge Solski u (sols7520@mach1.wlu.ca) wrote: : Early this morning, someone dialed into my Linux system and : logged-in. I know they were on for a while, because I could hear my : hard-drive thrashing away. I was too ill (some sort of flu thing) to : attend to my computer, but I thought I could see how they logged in later. : The only thing I could find, was that someone dialed in at 4:30 am. It : doesn't say who (I forget which file this was in.) Is there a file : somewhere that keeps track of who logs in and when? Try 'last' [this refers to the file /var/adm/wtmp or something very near to that]. It will tell you who has been on your machine since you last cleared wtmp. Bob Collie collieb@iia.org : -Mark : -- : "Key chuckles. 'If Skinny Puppy, in terms of the movie _Alien_, is a : chest-burster, then Doubting Thomas is more of a face-hugger,' he informs, : as if that were an explanation." : -Keyboard, Jan '92 ------------------------------ From: collieb@iia.org (Bob Collie) Subject: Re: tar help Date: 2 Oct 1994 21:38:55 GMT s010dls@alpha.wright.edu wrote: : I'm having trouble with tar. When I try to us it, all I get is: : tar: can't open /dev/rmt0 : No such device or address : Why does it say this? I looked for /dev/rmt0 and it's there. : Thanks. The default device to place entries from a tar file is the TAPE drive (/dev/rmt0) due to the fact that tar means Tape ARchive. To redirect it to the current directory, type tar -xvf and then the .tar file. Bob Collie collieb@iia.org ------------------------------ From: gonzo@magnet.mednet.net (Patrick J. Volkerding) Subject: Re: No Hostname Date: 1 Oct 1994 06:08:31 GMT In article <36gtob$okf@fs7.ece.cmu.edu>, Brad Matthew Garcia wrote: > >In article , cully@ritz.mordor.com (Robert J. Shmit) writes: >|> David - Morris (dwm@shell.portal.com) wrote: >|> : I have both /etc/HOSTNAME and /etc/hostname on my system and it >|> : still can't remember the hostname between boots. I issue: >|> : hostname gate >|> : from root (gate is the hostname) when I start and that fixes things. >|> >|> i had the same problem myself.. the problem is that slackware distribution >|> has a botched up hostname bin.. the new bin works fine.. > >Could you please tell me where to get the new bin? Do I need an upgraded >'a' disk set? Does anyone know? Yes, I do. The problem is not, and never was, that Slackware had a "botched up" hostname binary -- rather the bug was introduced by improper attempts to upgrade existing systems. In an older version of Slackware (I think 1.2.0), hostname could function as either hostname or domainname depending on the name it was called by. Therefore, domainname was simply a link to hostname. Some people tried to install a later version of tcpip.tgz which contained seperate hostname and domainname binaries without removing the old package first (like the UPGRADE.TXT says to). So, when domainname was untarred it followed the leftover link through and became the new 'hostname'. There's not much I can do about pilot error of this nature, unfortunately. The new version of Slackware doesn't even have a domainname binary. You change your domain name by editing /etc/hosts. This is the method recommended by the util-linux-1.10 documentation. (works great here :^) Hope this clears up that "bug" ;^) Pat ------------------------------ From: munster@MCS.COM (Jerry Ablan) Subject: Re: Help INN on my Linux box! Date: 2 Oct 1994 16:59:04 -0500 On Sat, 1 Oct 1994 19:34:13 GMT, Nathan Stratton spake thusly: : I am trying to set up my news on my linux box I had it working ok, but : now is has stoped working. If I look in /usr/adm/syslog I can see this error: : Oct 1 14:26:55 NovaNet innd: /usr/lib/news/history cant dbminit ME No : such file or directory : Can anyone tell me how to fix this problem, my news is backing up fast. : If you can help please send me mail as soon as possible. : -- : Nathan Stratton CEO, NovaNet, Inc. On-Line Communication Services. Sure looks like your history file got kilt. Just re-build it with the mkhistory script. -- Jerry ------------------------------ From: fox@graphics.cs.nyu.edu (David Fox) Subject: Re: HOW to transfer LARGE SINGLE file between Sun5.3 and Linux Date: 02 Oct 1994 14:41:05 GMT In article armb@setanta.demon.co.uk (Alan Braggins) writes: ] Or you can write a binary split. It's pretty short, so here's one ] someone else did earlier... The GNU version of split has an option for splitting binary files (of course!) -- David Fox xoF divaD NYU Media Research Lab baL hcraeseR aideM UYN ------------------------------ Subject: /sbin/hostname on 1.1.45- From: riku.saikkonen@compart.fi (Riku Saikkonen) Date: Sat, 1 Oct 94 00:50:00 +0200 > /sbin/hostname reston > /sbin/domainname cber.nih.gov >My system greets logins as: > cber.nih.gov (none) hostname has changed in net-tools-1.1.38. See 'man hostname' or 'hostname --help' for details on the new syntax. How to fix the problem: Remove the above commands and put just hostname reston.cber.nih.gov there. -=- Rjs -=- riku.saikkonen@compart.fi - IRC: Rjs GCS/L/M/TW/S -d+>++ H(+) s:- !g !p?>1+ !au a17 w+ v+(---)*>+++ C++>$ UL++++(A)>$ P+ L++>+++ 3 E>++ N+++>++ K- W+(++) M- !V po Y+>++ t/Tolkien+++ !5 !j R>+ G' tv-() b++(+) D++ B? e>+++ u+++@ h--! f+ !r>++ n+ !y+(*) ------------------------------ From: ralphs@halcyon.halcyon.com (Ralph Sims) Subject: Re: Assuring that a PPP link stays up... Date: 1 Oct 1994 03:59:23 GMT I spaketh wrongeth: >> Does anybody have a good, well thought-out system of assuring that a PPP >> link between two UNIX boxes stays up, through the use of crontab entries? > locus.halcyon.com:/pub/linux/ppp-on and ~ppp-off Those two files start and stop pppd. The actual file that keeps things alive is checkppp.sh. I've also placed checkslip.sh in there for folks are still SLIP'ing around. ------------------------------ From: lee@netspace.students.brown.edu (Lee J. Silverman) Subject: Re: inetd seems to lock-up Date: 02 Oct 1994 22:11:56 GMT Just so you know, I have a Linux box running Slackware 1.2, Binutils 1.5, and kernel version 1.1.37, and my Inetd hangs every once in a while as well. When this happens, I can connection to my web, gopher, ftp, and smail daemons, because they all run independently of inetd, but I cannot log in or rlogin. The problem usually seems to correct itself after an hour or two. If you have received *any* word of a fix I'd love to hear about it... my console often isn't accessible so when inetd is down there's no way to get into the machine... Thanks! -- Lee Silverman, Brown class of '94, Brown GeoPhysics ScM '95 Email to: Lee_Silverman@brown.edu Phish-Net Archivist: phish-archives@phish.net "Nonsense - you only say it's impossible because nobody's ever done it." ------------------------------ From: nstn@netcom.com (Nathan Stratton) Subject: Inn on a Linux box! Date: Sun, 2 Oct 1994 21:55:59 GMT I am trying to setup INN on my Linux box. I used inn1.4-linux-bin.tar off sunsite and untared it. I can not get it to run. When I do a innstat it just sits there. This is what my slslog look like: Oct 2 16:46:21 NovaNet innd: ME descriptors 256 Oct 2 16:46:22 NovaNet innd: ME outgoing 243 Oct 2 16:46:23 NovaNet innd: ME ccsetup control:12 Oct 2 16:46:23 NovaNet innd: ME rcsetup remconn:4 Oct 2 16:46:23 NovaNet innd: ME bad_newsfeeds no feeding sites Oct 2 16:46:24 NovaNet innd: ME internal no control and/or junk group I also tared my system on a 8 mill exabyte tape so I could make a bigger swap file, now I can not untar it. I get data reads errors and tar crashes. Is there a way I can untar the tape and have it not crash if it hist a read error? If you have any ideas please send me some mail as soon as possible. Thanks, -- Nathan Stratton CEO, NovaNet, Inc. On-Line Communication Services. ------------------------------ From: mse@crunch.hayward.ca.us (M. Shawn Easter) Subject: Where can I get x11perf? Date: Sat, 1 Oct 1994 02:56:43 GMT Can anyone tell me where I can get x11perf for X86 on linux? Thanks, Shawn. -- /~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~\ / /~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~\ \ / / M. Shawn Easter \ \ / / mse@crunch.hayward.ca.us \ \ /~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~\ | Get lean, mean, and strong. | +------------------------------------------------------------------+ ------------------------------ From: cmay@titan.ucs.umass.edu (Christopher M. May) Subject: Re: What's failed after Bogomips Date: 2 Oct 1994 00:26:08 GMT Eduardo Jacob Taquet (jtpjatae@bi.ehu.es) wrote: : I found that when booting Linux, just after bogomips line, (33.. Ok), i get : a failed that seems not to be related to anything. Does anybody know what is : this about? : Eduardo I believe that is a software probe for an adaptec 1542 scsi controller, in the scsi kernel in the slackware 2.0 distribution. -- -Chris May, Computer Science, University of MA, Amherst - Technical Assistant, P.C. Maintenance Lab ------------------------------ From: tony@teleport.com (Tony Schwartz) Subject: Computone MultiPort ??? Date: Sat, 1 Oct 1994 17:28:50 I have a Computone Multiport device with 16 serial ports. The software that comes with it supports SCO/UNIX/XENIX. Does anyone have any experience trying to run this under Linux ?? I am not sure what to look for as far as Unix compatibility with this card. Does it have to be Linux specific??? Any comments or input would be appreciated. Tony Schwartz Portland, OR ------------------------------ From: s010dls@alpha.wright.edu () Subject: xdm & xterm need help Date: Sat, 1 Oct 1994 07:05:41 GMT I have it set so that if you type startx, motif loads & opens a xterm window. I also have it set so that xdm does the same thing. However, when I log in through xdm and try to backspace in an xterm window (either the first one or any I start) it acts like a 'cancel' button, aborting the line and showing a fresh prompt. This doesn't happen with startx. The files xinitrc & Xsession are basically identical (except Xsession sources the ~/.profile file). I don't know how to fix it, help. Another problem: I have an option on the motif root window that runs an instance of xterm (looks like this): "New window" f.exec "xterm -sb -rv -ls &" This works as expected if I start X with startx. If I log in with xdm, the command works and I get a window, but it appears not to have source the startup files (/etc/profile & ~/.profile) since the prompt is bash# and none of the aliases make in those files are there. help again. :> ------------------------------ From: jbs30@eng.amdahl.com (Joel Storm) Subject: Re: 80x50 screen Date: Fri, 30 Sep 1994 19:26:02 GMT In article <1994Sep29.132613.4404@tudedv.et.tudelft.nl>, jakmouw@et.tudelft.nl (Erik Mouw) writes: |> In article , s010dls@alpha.wright.edu () writes: |> > Is there a way to use 80x50 text mode in Linux? I have a ATI mach32 |> > card, and I know it is capable of this. |> > |> > Thanks again. |> |> If you have LILO it's easy: |> Edit /etc/lilo.conf. You'll find something like: |> |> boot = /dev/hda1 |> compact |> image = /vmlinuz |> label = Linux |> ramdisk = 0 |> root = /dev/hda1 |> vga = normal |> ^^^^^^ |> |||||| |> If you change this in "vga = ask" LILO will ask which textmode you want. |> After you changed it run LILO (just type lilo) and the next time you |> boot LILO will ask about the VGA mode. It will display all the available |> VGA modes for your card. Just type a number to get it. Remeber the |> number and edit /etc/lilo.conf again. For the VGA setting use the number |> you used at boot time (if 80x50 is mode 1, then use vga = 1). Again run |> lilo to register the changes. Of course you have to be root to do all |> this stuff. |> |> If you don't have LILO it's a bit more difficult. Seems you have to use |> rdev... I don't have experience with it. Moral: man rdev :) |> I just set "vga = extended" (or maybe it's "expanded"). I expect it's documented in the LILO docs somewhere. -- ============================ Joel Storm: jbs30@eng.amdahl.com ------------------------------ From: hong@csulb.edu (Jason Hong) Subject: Setting up the BaseAddr for Mitumi Date: 1 Oct 1994 08:35:15 GMT I am using Mitsumi CD with base-addr = 0x340 and irq=11. I changed "mcd.h" file from 0x300, 10 to 0x340 and 11. However, whenever I start Linux, it still try to detect the CD from old address. I read through CD-ROM HOWTO but I couldn't find any mention on base-address. Also, I am getting an fsck error message after installation. Is there anything I have to do after installation to use Linux? Thank you, Jason -- +-------------------------------+-------------------------------+ | Documentary Photographer, | hong@csulb.edu | | Auto-Mechanics, | Academic Computing Services | | BS in Civil Eng. & | Cal. State Univ.@Long Beach | | Software Engineer | 310)985-7577 | +-------------------------------+-------------------------------+ ------------------------------ From: zmbenhal@netcom.com (Zeyd M. Ben-Halim) Subject: Re: Ncurses signals broken? Date: Sun, 2 Oct 1994 00:22:30 GMT In article <36hoqq$3ur@vixen.cso.uiuc.edu>, Joseph J Gebis wrote: > > I compiled ncurses 1.8, and it seems to work fine, except ^^^ >when you try to suspend a process that uses the ncurses library. >When you do this, the process simply hangs there; you can only >^C it. You have an OLD version, try using 1.8.5. Zeyd -- --- Zeyd M. Ben-Halim zmbenhal@netcom.com NCURSES is available from ftp.netcom.com:pub/zmbenhal/ncurses Current version is 1.8.5 ------------------------------ ** FOR YOUR REFERENCE ** The service address, to which questions about the list itself and requests to be added to or deleted from it should be directed, is: Internet: Linux-Admin-Request@NEWS-DIGESTS.MIT.EDU You can send mail to the entire list (and comp.os.linux.admin) via: Internet: Linux-Admin@NEWS-DIGESTS.MIT.EDU Linux may be obtained via one of these FTP sites: nic.funet.fi pub/OS/Linux tsx-11.mit.edu pub/linux sunsite.unc.edu pub/Linux End of Linux-Admin Digest ******************************