Linux-Misc Digest #265, Volume #26 Wed, 8 Nov 00 12:13:03 EST
Contents:
Re: Named started? (Raymond Doetjes)
Re: Bind capable of Dynamic DNS? (Raymond Doetjes)
Re: Linux/UNIX=Windows ([EMAIL PROTECTED])
Re: Why, ext2 don't need defrag (Johan Kullstam)
Re: Is this an IRQ problem? (Giacomo Catenazzi)
Re: What happens to the Linux system when there is little or no HD space (Lori
Holder-Webb)
HELP: Console text corruption (Stephen Cornell)
Re: Table of contents with TeTeX (Bob Tennent)
Re: HELP!! Partition Magic Error 105. PartitionInfo inside: Please look!! (Eric)
Re: Me tryin' convince my company to use Samba (Grant Edwards)
Re: KDE2 (JCA)
Re: HELP!! Partition Magic Error 105. PartitionInfo inside: Please look!! (Eric)
Re: samba and etc/printcap? (Wayne Pollock)
Re: HELP!! Partition Magic Error 105. PartitionInfo inside: Please look!!
([EMAIL PROTECTED])
Re: signal 11 and others while compiling (Scott Alfter)
Re: slocate eating up inodes? (Guillermo Labatte)
The significance of group "nobody" or user "nobody" (mike)
----------------------------------------------------------------------------
From: Raymond Doetjes <[EMAIL PROTECTED]>
Subject: Re: Named started?
Date: Wed, 08 Nov 2000 16:36:18 +0100
Which version of named are you running?
I saw this happen once on a RedHatbox with Version 4.3.something (NOT 9).
If you have a version younger then 4.3.9 or younger then 8.2.2p5
UPGRADE QUICKLY. Then your dies might be a thing of the past also.
Raymond
Wayne Pollock wrote:
> Suggest you check the error logs. Maybe named has some entries
> explaining why it dies.
>
> It sounds like your named runs ok for a while then dies unexpectedly.
> A temporary work-around it to not start named via the standard Linux
> method of an rc shell script. Instead, add an entry for it directly
> in /etc/inittab and use the "respawn" option. This way, init will
> automatically restart named any time it dies.
>
> But, check the error log (/var/log/messages) first!
>
> -Wayne Pollock
>
> Jay wrote:
> >
> > We are running a secondary DNS service on a Linux Red hat 6.2 OS
> > We had no problems initially but then we stopped getting authoritative
> > responses from this machine.
> > A colleague went into /etc/ and typed named and then all was fine.
> > Now twice this week we have not been getting an authoritative response from
> > the box, so we go into /etc/ and type named and nothing happens... apart
> > from the cursor dropping down a line and just flashing (without a prompt)
> > typing ps shows nothing to do with named.
> >
> > Is it not running?
> > If not then why?
> > If not then how do we stop it from stopping?
> >
> > Jay
------------------------------
From: Raymond Doetjes <[EMAIL PROTECTED]>
Subject: Re: Bind capable of Dynamic DNS?
Date: Wed, 08 Nov 2000 16:40:33 +0100
>From BIND8 and up it does suppoert DYnamic DNS.
They provide you with a tool called nsupdate to make these updates.
The nsupdate version 8 is better then the nsupdate version 9. The version 9 no
longer takes updates from file, (that stopped a few of my applications, I
quickly threw back the nsupdate version 8 on my system doing updates from file
to the BIND9 DNS server).
But what I don't understand is what you want. A DNS server has a fixed IP
address, otherwise your clients don't know where to find it if it keeps
changing,
Raymond
Micer wrote:
> I am not yet familiar with bind, and have a customer who is getting set up
> on Dynamic DNS. Is it possible to have a Linux-box DNS server do Dynamic DNS
> registration within itself whenever the IP address changes? (ie: if all you
> have is a dynamic IP rather than static).
>
> Micer
------------------------------
From: [EMAIL PROTECTED]
Subject: Re: Linux/UNIX=Windows
Date: Wed, 08 Nov 2000 14:45:41 GMT
Peter <[EMAIL PROTECTED]> writes:
> On Wed, 08 Nov 2000 02:55:06 GMT, [EMAIL PROTECTED] wrote:
> >"Mats Pettersson" <[EMAIL PROTECTED]> writes:
> >There are a whole lot of perspectives on what the priority of GUIs
> >are, and there most certainly _is_ a whole lot of attention that goes
> >into the goal of making GUIs essentially a tool "for ignorant people
> >who don't want to learn."
>
> Don't want to learn what?
>
> If you are saying they do not want to learn C or Perl or regular
> expressions, I agree.
>
> The advantages of a fast monochrome ASCII character based terminal
> approach could be demonstrated by using them to replace air traffic
> controller's radar screens.
Oh, goody. The sort of application that Every Computer User Needs.
Right?
No, wrong. The vast majority of us are incapable of the feats of
three-dimensional understanding that an ATC performs every minute of
their day.
ATC's go through _years_ of specialized training in order to know what
to do with that GUI.
I had some classmates in grad school that actually worked on an ATC
system project, but that represents two developers out of the hordes
that I have been acquainted with.
The fact that you'd pluck ATC radar screens out of the air and suggest
it to be a "typical" example of a computer application certainly
demonstrates your bias.
Similarly, "air superiority" fighters represent a tiny pimple on the
butt of the computing world where, even yet, there's probably still
more COBOL code in operation than there is code in all other languages
put together.
--
(concatenate 'string "cbbrowne" "@hex.net") <http://www.ntlug.org/~cbbrowne/>
"Never make any mistaeks." (Anonymous, in a mail discussion about to a
kernel bug report.)
------------------------------
Crossposted-To: comp.os.linux.system
Subject: Re: Why, ext2 don't need defrag
From: Johan Kullstam <[EMAIL PROTECTED]>
Date: 08 Nov 2000 09:51:48 -0500
"starbux" <[EMAIL PROTECTED]> writes:
> Hi,
> SGI's IRIX has that "journalized" feature...it needs no defrag.
afaik journaling is orthogonal to the fragementation issue.
journalling just serializes updates to the "metadata" (filesystem
pointers to the actual file data). journalling neither hurts nor
helps the way the file itself is written.
a contrary but i think valid viewpoint would be that with journalling
all writes are effectively fragmented since you have to keep shuttling
back and forth from journal to file as it is written. sometimes the
journal is kept on a different disk from the file data for this very
reason.
> there is a "defrag" utility for the linux fs, ext2. never used it tho...
> i read somewhere that linux next fs, ext3, will have that "auto-defrag"
> feature.
> unless...my memory is bad ;-)
i seem to recall something to the effect that the "phase tree" disk
state upkeep would admit a defragmenter to continuously work on an
active filesystem. ext3 doesn't use the phase tree, but a metadata
journal.
--
J o h a n K u l l s t a m
[[EMAIL PROTECTED]]
sysengr
------------------------------
Date: Wed, 08 Nov 2000 15:56:42 +0100
From: Giacomo Catenazzi <[EMAIL PROTECTED]>
Subject: Re: Is this an IRQ problem?
bill g wrote:
>
> Mandrake 7.2. After I run startx and the KDE desktop
> completely loads, in about 30 seconds my mouse and keyboard
> freeze.
this problem are common with gdm and X11. Try to remove the package gpm
or try to not load it.
> This happened after installing alsa and running
> sndconf. I've tried doing interactive boot and entering "N"
> when prompted for starting the sound module but it still
> freezes. Is there a way to check the IRQ's in console,
> and/or figure out and disable whatever is loading that
> causes the hang?
in /proc/interrupt there are listed the irq.
Normally the kernel checks for resource, thus there are no conflicts.
try CTRL-ALT-F1 to go in console mode (maybe is only a problem with X
drivers)
giacomo
> Thanks,
> Bill G
------------------------------
From: Lori Holder-Webb <[EMAIL PROTECTED]>
Subject: Re: What happens to the Linux system when there is little or no HD space
Date: Wed, 08 Nov 2000 08:52:37 -0600
Had this happen a while back - I wasn't maintaining my log files, and
the grew & grew & grew until my root parition was full. The next time I
signed on as user and tried to launch X, it wouldn't work and gave me
all sorts of really unhelpful messages. The errors I got made it look
like something was wrong with X itself, whereas the real problem was
that there wasn't enough space on the partition to create whatever temp
files (I guess) X and the environment needed to run.
Very confusing, I give it a 3 Hairs Pulled rating (out of 4) for Linux
Problems.
Can't recall where I got the notion to check the space on the disk, but
eventually I did and that's when I found that the root parition was 99%
full. Cleaned out the logs, everything fired up as it should have after
that.
mike wrote:
>
> Hi,
> I was wondering what happens if you are working on a Linux
> system and get low on hard disk space memory or run out of space.
> Does the system give you a warning message, do programs freeze or
> crash, etc?
> Thanks
> Mike
------------------------------
From: Stephen Cornell <[EMAIL PROTECTED]>
Subject: HELP: Console text corruption
Date: 08 Nov 2000 15:08:57 +0000
I'm having problems with Debian Potato installed on a rather old
Viglen Genie PC. Vi has problems in the console: the first few lines
of the file I'm editing are absent, spurious text appears, and the
text that is edited is not the text that appears to be under the
cursor. This problem also exists in xterms. However, the problem
doesn't seem to be limited to VI: similar problems happen in emacs,
and the console seems to become more and more corrupted, so that fewer
and fewer lines on the console actually scroll; eventually, the only
line on the console that refreshes is the bottom one, so that it is
completely unusable.
A problem that may be related: at boot time, a colour image of Tux
appears at the top left of the console, and this does not go away
completely at first.
Any ideas? Incidentally, Debian apparently does not use termcap -
only terminfo.
--
Stephen Cornell [EMAIL PROTECTED] Tel/fax +44-1223-336644
University of Cambridge, Zoology Department, Downing Street, CAMBRIDGE CB2 3EJ
------------------------------
From: [EMAIL PROTECTED] (Bob Tennent)
Subject: Re: Table of contents with TeTeX
Date: 8 Nov 2000 14:52:52 GMT
Reply-To: [EMAIL PROTECTED]
On Wed, 08 Nov 2000 12:10:31 -0200, Filipe Bonjour wrote:
>
>I'm not sure this is the place to post this, but my news server doesn't
>serve any TeX-related groups. I have a text (in the LaTeX2e report
>class) which I wrote some time ago under Solaris. Now I would like to
>print it with TeTeX 1.0.6-11 under Red hat 6.2, and it all works fine
>without having to change anything. All but one thing: the table of
>contents.
>
>For some reason, TeTeX only prints the parts and chapters in the toc,
>although the .toc file also lists the sections (and also although,
>under Solaris, LaTeX printed them). I tried fiddling with a few
>counters, but without success.
Should be controlled by the tocdepth counter. Set it to a value >= 1.
You may need to run LaTeX on your source file twice.
Bob T.
------------------------------
From: Eric <[EMAIL PROTECTED]>
Crossposted-To:
comp.os.linux,comp.os.linux.help,comp.os.linux.questions,comp.os.linux.setup,comp.os.linux.x
Subject: Re: HELP!! Partition Magic Error 105. PartitionInfo inside: Please look!!
Date: Wed, 08 Nov 2000 16:10:42 +0100
Reply-To: [EMAIL PROTECTED]
>
> You might be able to find the actual partition table by using gpart. Then
> you could use fdisk to rewrite the partition table. I don't think it will
> work to use linux fdisk to rewrite the table to extend it for windows,
> because windows will not recognize it unless you use dos fdisk which wipes
What do you mean with that? Windows will recognize partitions created
with linux (c)fdisk just fine. Ofcourse the partition ID must be one of
the types that windows knows of.
> out the data. Hopefully if you use gpart you will find what the table
> should be and fix the mess PM made.
Did PM make this mess? That's not very likely. I suppose the original
poster used the mandrake (re)partitioner. They should put a serious
warning on the screen before anyone can use it, 'cause it's notorious in
(corrupting) partition tables.
Perhaps the original poster can use GNU's parted to resize the
partitions, so they will start at cylinder boundary's again.
After that, don't use any other partitioning tool than PM. If you have
this tool, use it. It may not be free (so if you haven't bought it yet,
try parted first), but it's the best piece of non-free software you can
get (IMO).
>
> http://www.stud.uni-hannover.de/user/7620/gpart
>
> You will need to get the binary version to put on a floppy and also get
> Tom's linux to run a mini linux to use gpart. The url for tom's is on the
> gpart page. On tom's there is also rescuept which gives you info on the
> partition table, but I find it harder to use than what gpart gives. You can
> try both and compare.
The partition table is probably not corrupt, but PM is very strict on
what it does allow. I suppose mainly for compatability reasons (DOS
FDISK can only start/end a partition on a cylinder boundary)
So maybe the table is corrupt (but I couldn't tell from what I've seen),
could the original poster post the result of linux `fdisk -l /dev/hda`
please?
Eric
> Anita
------------------------------
From: [EMAIL PROTECTED] (Grant Edwards)
Crossposted-To:
comp.protocol.smb,linux.samba,comp.os.linux.network,mailing.unix.samba-technical,micosoft.public.windowsnt
Subject: Re: Me tryin' convince my company to use Samba
Date: Wed, 08 Nov 2000 15:14:23 GMT
In article <8u8lnm$oji$00$[EMAIL PROTECTED]>, Dietrich M. wrote:
>2 posibilities:
> 1: create a 100 MB file (dd if=... of=... bs=... count=...)
> and copy it to/from a mounted share of one of your clients
> 2: use smbmnt; it is very similar to ftp (get put)
Smbmnt just mounts a filesystem. Do you mean smbclient?
--
Grant Edwards grante Yow! LOU GRANT froze
at my ASSETS!!
visi.com
------------------------------
From: JCA <[EMAIL PROTECTED]>
Subject: Re: KDE2
Date: Wed, 08 Nov 2000 07:15:14 -0800
Jerry L Kreps wrote:
> JCA wrote:
>
> >
> > This is not strictly Linux but - I have just installed KDE2 in my
> > Linux box and I am most unimpressed. It not only doesn't look quite
> > as good as its 1.* counterpart, but the extra goodies added are not
> > all that great: Konqueror and Koffice are on par with free tools that I
> > already have, but they don't add all that much (if anything) extra.
> >
> > This is of course just my opinion, but I am tempted to get back
> > to 1.*, or to ditch it altogether in favor of Gnome.
> >
> > What a big disappointment.
>
> Differences of opinion are what makes the world go round...
You are right. I just wish I could share your enthusiasm about
KDE2 though....
------------------------------
From: Eric <[EMAIL PROTECTED]>
Crossposted-To:
comp.os.linux,comp.os.linux.help,comp.os.linux.questions,comp.os.linux.setup,comp.os.linux.x
Subject: Re: HELP!! Partition Magic Error 105. PartitionInfo inside: Please look!!
Date: Wed, 08 Nov 2000 16:23:03 +0100
Reply-To: [EMAIL PROTECTED]
Jacques Guy wrote:
>
> Nataraj Dasgupta wrote:
>
> > Forget Partition Magic.
> [...]
>
> Follow Nataraj's advice. Partition Magic is a useless
> piece of software. Time and again (how many? four
> times I think) I have been plagued by similar
> problems. Always, always, Partition Magic gave
> me horrendous error messages, no help whatsoever,
Yeah right. I know PM can abort with these error messages,
(I've had them myself too) but that doesn't make it bad software.
(I agree the error message aren't very clear)
But partitiontable's aren't easy. Every single OS on the market
that claim's to understand a PC's partitiontable has it's own
gotcha's in handling these tables. PM's tries to be so generic
in the table it creates/handles due to these (in)compatabilities.
If you would have used PM to create your table from start,
and not mess with the fdisk's from any OS, PM will resize, create
and delete all entries just fine. But if you have used an fdisk,
that's is more rich-featured than DOS fdisk (like the linux one)
PM may find it impossible to deal with your table.
If you would obey all the rules PM does (and they are too stringent, I
know)
you would never have any partition table problem to begin with.
Eric
> the prize going to an error message which was not
> in the manual! Boot from a Linux diskette, run
> Linux fdisk, go on from there. Avoid MS fdisk
> like the plague, unless you are keen to lose all
> your data. Once, 4-5 years ago, MS-DOS decided that
> I had no F: disk. The Norton Utilities (what a piece
> of shit!) told me to do a low-level formatting.
> So, stuffed for stuffed, I booted from a Linux
> diskette (I knew bugger all about Linux, just
> a little bit of Unix). Eventually, having found
> the right /dev/hdax to mount, I could read, edit,
> and copy the DOS files on my "lost" disk F:!
> From Linux. I eventually managed, all through
> Linux, to re-enable F: (forgot how, posted it
> on comp.linux.advocacy, but DejaNews no longer
> goes that far back).
Was your partition table correct?
The fact that you think it was, doesn't mean that it really was.
(You can use a partition from an incorrect table just fine
(depends on the error there is in the table) but that doesn't make it
correct
You will on the other hand have absolutely *no* guarantee that the data
on such a partition (and in fact on the entire disc!) will stay
uncorrupted
over time.
I suspect that at least 10% of every PC user out there that has more
than
one OS on his/her PC has an incorrect partition table. (I've seen it far
too often)
Eric
------------------------------
From: Wayne Pollock <[EMAIL PROTECTED]>
Subject: Re: samba and etc/printcap?
Date: Wed, 08 Nov 2000 10:54:28 -0500
This sort of error has nothing to do with samba, it is called
"stair-stepping" and if you use Red Hat printtool, there is a
checkbox to fix it. (I wonder what changes if you check that box?
I'm not in a position now to look.)
-Wayne Pollock
[EMAIL PROTECTED] wrote:
>
> I have gotten mine to print but only the first line of a file which is
> what I am trying to figure out. In the printcap file set
> if=/usr/local/samba/bin/smbprint: or where ever your smbprint script
> is located.
------------------------------
From: [EMAIL PROTECTED]
Crossposted-To: comp.os.linux,comp.os.linux.help,comp.os.linux.questions
Subject: Re: HELP!! Partition Magic Error 105. PartitionInfo inside: Please look!!
Date: Wed, 08 Nov 2000 16:29:42 GMT
In article <7vaO5.1094$[EMAIL PROTECTED]>,
[EMAIL PROTECTED] wrote:
> >
> >So what can I do to resize my Windows98 partition to the full
> >size of the HD without damaging the Win98 partition?
>
> You might be able to find the actual partition table by using
> gpart. Then you could use fdisk to rewrite the partition table.
> I don't think it will work to use linux fdisk to rewrite the
> table to extend it for windows, because windows will not
> recognize it unless you use dos fdisk which wipes out the data.
> Hopefully if you use gpart you will find what the table
> should be and fix the mess PM made.
>
> http://www.stud.uni-hannover.de/user/7620/gpart
>
> You will need to get the binary version to put on a floppy and
> also get Tom's linux to run a mini linux to use gpart. The url
> for tom's is on the gpart page. On tom's there is also rescuept
> which gives you info on the partition table, but I find it
> harder to use than what gpart gives. You can try both and compare.
>
> Anita
The URL fails :-( Guess somebody moved on.
--
Chuck Falconer ([EMAIL PROTECTED])
http://www.qwikpages.com/backstreets/cbfalconer/
Sent via Deja.com http://www.deja.com/
Before you buy.
------------------------------
From: [EMAIL PROTECTED] (Scott Alfter)
Subject: Re: signal 11 and others while compiling
Date: Wed, 08 Nov 2000 16:58:49 -0000
=====BEGIN PGP SIGNED MESSAGE=====
Hash: SHA1
In article <[EMAIL PROTECTED]>,
Jong-Min Park <[EMAIL PROTECTED]> wrote:
>Here is my situation:
>
>Whenever doing some heavy compilation (kernel, etc.) I get the
>following errors randomly
>
>signal 11 fault
>[...]
>The problem still persists. Is my CPU burned by the older motherboard?
>It used to be very hot until I replaced the motherboard. I had a good
>fan though.
There's a chance that overvoltage from your first motherboard cooked
something in the processor to where it's flaky under heavy load now. I also
have a 450-MHz K6-III, only it's on an FIC VA-503+ and has always run at
moderate temperature. (The board originally came with a 300-MHz K6-2; it's
about two years old at this point. All that was needed to get the K6-III to
work with it was a BIOS update.)
>How probable is it for the AMD CPU to cause these problems?
Not at all probable. The last time I had those errors was when I used a
ULSI 387SX with a couple of different processors, which was (as you can
probably guess) a long time ago.
_/_
/ v \
(IIGS( Scott Alfter (remove Voyager's hull number for email address)
\_^_/ http://salfter.dyndns.org
=====BEGIN PGP SIGNATURE=====
Version: GnuPG v1.0.2 (GNU/Linux)
Comment: For info see http://www.gnupg.org
iD8DBQE6CYZAVgTKos01OwkRAhS7AKCvTE+ZVuq+dj5lzbMkyTT5TH3kHACgm2Hh
rO+hljghYrLbFPPQYYnWrw4=
=5Qek
=====END PGP SIGNATURE=====
------------------------------
From: Guillermo Labatte <[EMAIL PROTECTED]>
Subject: Re: slocate eating up inodes?
Date: Wed, 08 Nov 2000 14:10:58 -0300
David Efflandt wrote:
> On Mon, 30 Oct 2000 09:07:24 -0300, Guillermo Labatte <[EMAIL PROTECTED]> wrote:
> >I discovered my linux box was crashing every other day due to inode
> >depletion.
> >My first though was to increase inode-max. I did it, then checked
> >inode-nr regularly. No matter how high I set inode-max, the number in
> >inode-nr (first value) continued rising.
> >So I created a log, sampling inode-nr values every two minutes. I
> >discovered the numbers went more or less the same all day, but at 4am
> >inode-nr increased in 14000 units every day. That is the time when
> >cron.daily scripts run. I modified the scripts in order to create a log
> >of inode-nr values (before and after the script execution). The culprit
> >seem to be the slocate script. After its execution inode-nr is increased
> >in 14000 units.
> >So the question is... Is somebody aware of a problem with slocate? Is
> >there a solution other than disable the script? Are other programs apart
> >from slocate affected?
> >
> >Setup info:
> >distribution: Mandrake 7.0
> >kernel version: 2.2.14
> >slocate version: Secure Locate v2.1 - Released October 19, 1999
> > (library dependencies)
> > libc.so.6 => /lib/libc.so.6 (0x00128000)
> > /lib/ld-linux.so.2 => /lib/ld-linux.so.2 (0x00110000)
>
> Just curious how you tell what inode-nr (mine is 12300 7266) and inode-max
> (mine 12288) are for? For my main box running Mandrake 7.0 24/7 with a
> generic 2.2.17 kernel, df shows 26% disk use and 12% inodes used for /,
> and 9% use and 1% inodes for /var. My laptop runs less often overnight
> and df shows / 91% used, but only 41% of its inodes used (/boot is using
> 1% of its inodes). I have never had an inode shortage even on a partition
> that was relatively full.
I don't know for sure what purpose inode-nr and inode-max have. I suppose they're
related to some data structure the kernel has to keep in memory, maybe some table
of "open" inodes (whatever that means). So I think you can have plenty of room in
your file system (df reports lots of free inodes), but you wouldn't be able to
access them if the kernel runs out of entries in that table. I'm just guessing
here, so it would be nice if someone explained how it really works.
What I know is that after removing slocate from the cron jobs, inode-nr stopped
rising (at least at the rate it used to), and the need to reboot disappeared.
> Are you running postfix? I never could figure out how to configure it and
> did not notice its jumbo log and saved cron msgs it could not send, until
> my /var was almost full. I removed postfix and installed sendmail
> instead, which works for me without any configuration.
>
I'm running sendmail too, but the problem wasn't a full file system. There was
plenty of space in all file systems. The problem seemed to be kernel-related.
> --
> David Efflandt [EMAIL PROTECTED] http://www.de-srv.com/
> http://www.autox.chicago.il.us/ http://www.berniesfloral.net/
> http://cgi-help.virtualave.net/ http://hammer.prohosting.com/~cgi-wiz/
------------------------------
From: mike <[EMAIL PROTECTED]>
Subject: The significance of group "nobody" or user "nobody"
Date: Wed, 08 Nov 2000 17:04:14 GMT
Hi,
I have seen the use of user "nobody" in /etc/passwd
and group "nobody" in /etc/group.
Also there is "wheel", "sync" "operator".
What is the significance of these and other named group / users?
Thanks
Mike
------------------------------
** 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: [EMAIL PROTECTED]
You can send mail to the entire list (and comp.os.linux.misc) via:
Internet: [EMAIL PROTECTED]
Linux may be obtained via one of these FTP sites:
ftp.funet.fi pub/Linux
tsx-11.mit.edu pub/linux
sunsite.unc.edu pub/Linux
End of Linux-Misc Digest
******************************
