Tkdesk Segmentation Fault

Tkdesk Segmentation Fault

[Peter]

Hi,

In RH 7.3 the program tkdesk will not open any longer.

Only the message 'Segmentation fault' shows up immediately after giving the 
command tkdesk.

Why would that be and how could it be corrected?

Thanks & regards



-- 
Peter

-
To unsubscribe from this list: send the line "unsubscribe linux-newbie" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.linux-learn.org/faqs

Re: Tkdesk Segmentation Fault

[Ray Olszewski]

At 09:26 AM 11/5/02 +0800, Peter wrote:
>Hi,
>
>In RH 7.3 the program tkdesk will not open any longer.
>
>Only the message 'Segmentation fault' shows up immediately after giving the
>command tkdesk.
>
>Why would that be and how could it be corrected?


How basic is your question, Peter? I'll assume fairly basic.

A segfault occurs when a process attempts to access a location in memory 
that has not been assigned to it by the kernel. I don't know all the 
reasons a process may segfault (or maybe I do, but in that case I don't 
know that I know), but the ones I know of are --

         a programming error in the application (for example, attempting to 
read or write to an element of an array greater than the size of the array)

         a programming error in a shared library that the application uses

         a hardware problem, typically bad RAM, that causes a memory 
address to be corrupted

         a swap-space problem, causing an error similar to the hardware 
(RAM) problem above

Each source has a distinct (and, I trust, obvious, though not trivial) 
solution.

--
-------------------------------------------"Never tell me the odds!"--------
Ray Olszewski					-- Han Solo
Palo Alto, California, USA			  ray@comarre.com
-------------------------------------------------------------------------------

-
To unsubscribe from this list: send the line "unsubscribe linux-newbie" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.linux-learn.org/faqs

Re: Tkdesk Segmentation Fault

[Elias Athanasopoulos]

On Tue, Nov 05, 2002 at 09:26:22AM +0800, Peter wrote:
> Only the message 'Segmentation fault' shows up immediately after giving the 
> command tkdesk.
> 
> Why would that be and how could it be corrected?

Inspect the core file using gdb:

% gdb tkdesk ./core

Or use strace:

$ strace tkdesk

And see if you can figure where it crashed. If yes, then you should pro-
bably know why, too.

-- 
http://gnewtellium.sourceforge.net			MP3 is not a crime.	
-
To unsubscribe from this list: send the line "unsubscribe linux-newbie" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.linux-learn.org/faqs

Re: Tkdesk Segmentation Fault

[Peter]

Thanks!

On Tue, Nov 05, 2002 at 09:26:22AM +0800, Peter wrote:
>> Only the message 'Segmentation fault' shows up immediately after giving the 
>> command tkdesk.
> 
>> Why would that be and how could it be corrected?

I have to add from RH5.2 thru RH7.2 tkdesk always opened flawlessly.
It is only now with RH7.3 that I have the problem to which I have (foolishly) 
upgraded a few days ago.

Ray, my guess is that is has something to do with the new libraries. How would 
I find out which one. ldd tkdesk gives "not a dynamic executable".
I still have RH6.2 running and could transfer the libraries from there.

If I try to re-./.configure I get error:
checking version of Tcl/Tk... 8.3/8.3

TkDesk requires Tcl/Tk versions 7.5/4.1, 7.6/4.2., or 8.0.
Older versions are unlikely to work, while newer ones are not
yet supported

So I copied Tcl/Tk...8.0 from RH6.2 over to RH7.3 and I still get the same 
error.

eathan@otenet.gr said:
> Inspect the core file using gdb:
> % gdb tkdesk ./core 

There is no core file created.

eathan@otenet.gr said:
> Or use strace:
> $ strace tkdesk

This is the tail of the strace output:
mmap2(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 
0x40
016000
read(4, "#\t$Xorg: locale.dir,v 1.3 2000/0"..., 4096) = 4096
read(4, "LOCALE\t\t\tfr_BE.ISO8859-15\niso885"..., 4096) = 4096
read(4, ".TCVN\nvi_VN.viscii/XLC_LOCALE\t\tv"..., 4096) = 4096
read(4, "\t\t\tlv_LV.UTF-8\nen_US.UTF-8/XLC_L"..., 4096) = 4096
close(4)                                = 0
munmap(0x40016000, 4096)                = 0
open("/usr/X11R6/lib/X11/locale/C/XI18N_OBJS", O_RDONLY) = 4
fstat64(4, {st_dev=makedev(22, 67), st_ino=451076, st_mode=S_IFREG|0444, 
st_nlin
k=1, st_uid=0, st_gid=0, st_blksize=4096, st_blocks=8, st_size=333, 
st_atime=200
2/11/06-09:11:05, st_mtime=2002/04/19-11:11:08, st_ctime=2002/11/03-11:24:28}) 
=
 0
mmap2(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 
0x40
016000
read(4, "# CATEGORY(XLC|XIM|OM)\tSHARED_LI"..., 4096) = 333
read(4, "", 4096)                       = 0
close(4)                                = 0
munmap(0x40016000, 4096)                = 0
--- SIGSEGV (Segmentation fault) ---
+++ killed by SIGSEGV +++

It's Greek for me.

Regards
-- 
Peter

-
To unsubscribe from this list: send the line "unsubscribe linux-newbie" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.linux-learn.org/faqs

Re: Tkdesk Segmentation Fault

[Ray Olszewski]

At 10:42 AM 11/6/02 +0800, Peter wrote:
>On Tue, Nov 05, 2002 at 09:26:22AM +0800, Peter wrote:
> >> Only the message 'Segmentation fault' shows up immediately after 
> giving the
> >> command tkdesk.
> >
> >> Why would that be and how could it be corrected?
>
>I have to add from RH5.2 thru RH7.2 tkdesk always opened flawlessly.
>It is only now with RH7.3 that I have the problem to which I have (foolishly)
>upgraded a few days ago.
>
>Ray, my guess is that is has something to do with the new libraries. How 
>would
>I find out which one. ldd tkdesk gives "not a dynamic executable".
>I still have RH6.2 running and could transfer the libraries from there.

Peter -- I have absolutely no familiarity with tkdesk, sso all these 
comments are based on my more general understanding of Linux systems.

 From the ldd response, I'd infer that it is a script rather than an 
executable, presumably an tcl script from the beginning of its name. You 
might want to ldd tcl (or whatever the right name is -- see what 
interpreter the tkdesk script is invoking on its shebang line). But Tcl 
itself *seems* (I'm not certain because I do not have it installed) to 
depend only on libc6, and I'd be surprised if RH's install of libc6 is 
buggy (or its bugs would show up everywhere, this lib being so central to 
all apps).

You might also want to make sure that even after you install TCL 8.0, 
tkdesk uses it rather then the more recent version of Tcl that RH has 
installed.

As to the strace output you included, I can pretty much read it, but I 
don't see where or why it is segfaulting. Perhaps whoever suggested using 
strace will be better able to spot the exact problem than I am. The last 
statement executed -- the munmap -- appears correctly to be releasing the 
4096 bytes of memory that were allocated by mmap a few lines earlier, and 
the several reads right before it look OK too. So I'm missing it, whatever 
it is. Sorry.


>If I try to re-./.configure I get error:
>checking version of Tcl/Tk... 8.3/8.3
>
>TkDesk requires Tcl/Tk versions 7.5/4.1, 7.6/4.2., or 8.0.
>Older versions are unlikely to work, while newer ones are not
>yet supported
>
>So I copied Tcl/Tk...8.0 from RH6.2 over to RH7.3 and I still get the same
>error.
>
>eathan@otenet.gr said:
> > Inspect the core file using gdb:
> > % gdb tkdesk ./core
>
>There is no core file created.
>
>eathan@otenet.gr said:
> > Or use strace:
> > $ strace tkdesk
>
>This is the tail of the strace output:
>mmap2(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) =
>0x40
>016000
>read(4, "#\t$Xorg: locale.dir,v 1.3 2000/0"..., 4096) = 4096
>read(4, "LOCALE\t\t\tfr_BE.ISO8859-15\niso885"..., 4096) = 4096
>read(4, ".TCVN\nvi_VN.viscii/XLC_LOCALE\t\tv"..., 4096) = 4096
>read(4, "\t\t\tlv_LV.UTF-8\nen_US.UTF-8/XLC_L"..., 4096) = 4096
>close(4)                                = 0
>munmap(0x40016000, 4096)                = 0
>open("/usr/X11R6/lib/X11/locale/C/XI18N_OBJS", O_RDONLY) = 4
>fstat64(4, {st_dev=makedev(22, 67), st_ino=451076, st_mode=S_IFREG|0444,
>st_nlin
>k=1, st_uid=0, st_gid=0, st_blksize=4096, st_blocks=8, st_size=333,
>st_atime=200
>2/11/06-09:11:05, st_mtime=2002/04/19-11:11:08, 
>st_ctime=2002/11/03-11:24:28})
>=
>  0
>mmap2(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) =
>0x40
>016000
>read(4, "# CATEGORY(XLC|XIM|OM)\tSHARED_LI"..., 4096) = 333
>read(4, "", 4096)                       = 0
>close(4)                                = 0
>munmap(0x40016000, 4096)                = 0
>--- SIGSEGV (Segmentation fault) ---
>+++ killed by SIGSEGV +++
>
>It's Greek for me.



--
-------------------------------------------"Never tell me the odds!"--------
Ray Olszewski					-- Han Solo
Palo Alto, California, USA			  ray@comarre.com
-------------------------------------------------------------------------------

-
To unsubscribe from this list: send the line "unsubscribe linux-newbie" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.linux-learn.org/faqs

Re: Tkdesk Segmentation Fault

[Elias Athanasopoulos]

On Wed, Nov 06, 2002 at 10:42:13AM +0800, Peter wrote:
> eathan@otenet.gr said:
> > Inspect the core file using gdb:
> > % gdb tkdesk ./core 
> 
> There is no core file created.

% ulimit -c 1000000 

Rerun TkDesk and let it crash. You should have a core file now.

> eathan@otenet.gr said:
> > Or use strace:
> > $ strace tkdesk
> 
> This is the tail of the strace output:
> mmap2(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 
> 0x40
> 016000
> read(4, "#\t$Xorg: locale.dir,v 1.3 2000/0"..., 4096) = 4096
> read(4, "LOCALE\t\t\tfr_BE.ISO8859-15\niso885"..., 4096) = 4096
> read(4, ".TCVN\nvi_VN.viscii/XLC_LOCALE\t\tv"..., 4096) = 4096
> read(4, "\t\t\tlv_LV.UTF-8\nen_US.UTF-8/XLC_L"..., 4096) = 4096
> close(4)                                = 0
> munmap(0x40016000, 4096)                = 0
> open("/usr/X11R6/lib/X11/locale/C/XI18N_OBJS", O_RDONLY) = 4
> fstat64(4, {st_dev=makedev(22, 67), st_ino=451076, st_mode=S_IFREG|0444, 
> st_nlin
> k=1, st_uid=0, st_gid=0, st_blksize=4096, st_blocks=8, st_size=333, 
> st_atime=200
> 2/11/06-09:11:05, st_mtime=2002/04/19-11:11:08, st_ctime=2002/11/03-11:24:28}) 
> =
>  0
> mmap2(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 
> 0x40
> 016000
> read(4, "# CATEGORY(XLC|XIM|OM)\tSHARED_LI"..., 4096) = 333
> read(4, "", 4096)                       = 0
> close(4)                                = 0
> munmap(0x40016000, 4096)                = 0
> --- SIGSEGV (Segmentation fault) ---
> +++ killed by SIGSEGV +++

It's probably a parsing error. The script opens various files (such as
/usr/X11R6/lib/X11/locale/C/XI18N_OBJS) and tries to read them. I guess (but
that is just a guess) your upgrade from RedHat 7.1 to RedHat 7.3 had as result
to change the format/contents of these files, hence TkDesk can't parse them and
segfaults (if this is the case, it shouldn't happen if TkDesk was properly
programmed).

You'll find more if you inspect the core file with gdb (use the command
'where' to find where it crashed).

Elias

-- 
http://gnewtellium.sourceforge.net			MP3 is not a crime.	
-
To unsubscribe from this list: send the line "unsubscribe linux-newbie" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.linux-learn.org/faqs

Re: Tkdesk Segmentation Fault

[Peter]

Thanks!

eathan@otenet.gr said:
> % ulimit -c 1000000 
> Rerun TkDesk and let it crash. You should have a core file now. 

$ tkdesk
Segmentation fault (core dumped)

except it's nowhere to be found. Wait, here it is now with ending numbers: 
core.1298
One of the new tricks of RH7.3?

gdb core.1298 
This GDB was configured as "i386-redhat-linux"...core.1298: No such file or 
directory

cp core.1298 ~/core

gdb core
This GDB was configured as "i386-redhat-linux"..."/home/peter/core": not in 
executable format: File format not recognized

(gdb) where
No stack.

Looking at the core file I see a lot of talk about:
"his probably means that Itcl/Tcl weren't installed properly"
and
set msg "Can't find a usable init.tcl in the following directories: 
"
The end of core file is:

  set f [open pkgIndex.tcl w]
    puts $f $index
    close $f
    cd $oldDir
@|[\x01@(…Éué‹“0
@|Ó\x0f@L<
@\x01@…=
followed by a lot more of those weird signs.

ray@comarre.com said:
> You might also want to make sure that even after you install TCL 8.0,
> tkdesk uses it rather then the more recent version of Tcl that RH has
> installed.

I did that by changing the tcl/tlConfig.sh files.


ray@comarre.com said:
> You  might want to ldd tcl (or whatever the right name is -- see what
> interpreter the tkdesk script is invoking on its shebang line). 

ldd tcl... gives the same files in RH7.3 as in RH6.2.

If this is another case of strange happenings under RH7.3 then we might probe 
again in the dark or is it that the new tk/tcl of RH7.3 are at fault only. 
Could one reinstall the ones of RH7.2 w/o messing-up everything?

I hate to see a solid good program <tkdesk> go to waste on account of RH7.3.

Regards

-- 
Peter

-
To unsubscribe from this list: send the line "unsubscribe linux-newbie" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.linux-learn.org/faqs

Re: Tkdesk Segmentation Fault

[Elias Athanasopoulos]

On Thu, Nov 07, 2002 at 02:58:57PM +0800, Peter wrote:
> > % ulimit -c 1000000 
> > Rerun TkDesk and let it crash. You should have a core file now. 
> 
> $ tkdesk
> Segmentation fault (core dumped)
> 
> except it's nowhere to be found. Wait, here it is now with ending numbers: 
> core.1298
> One of the new tricks of RH7.3?

% gdb tkdesk ./core.1298
...
% (gdb) where

Elias

-- 
http://gnewtellium.sourceforge.net			MP3 is not a crime.	
-
To unsubscribe from this list: send the line "unsubscribe linux-newbie" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.linux-learn.org/faqs

Re: Tkdesk Segmentation Fault

[Peter]

There your are.

]$ gdb tkdesk ./core.1875

This GDB was configured as "i386-redhat-linux"..."/usr/local/bin/tkdesk": not 
in executable format: File format not recognized

Core was generated by `tkdesksh /usr/local/bin/tkdesk'.
Program terminated with signal 11, Segmentation fault.
-#0  0x0816e640 in ?? ()
(gdb) where
-#0  0x0816e640 in ?? ()
-#1  0x4007502a in ?? ()
-#2  0x400746f9 in ?? ()
-#3  0x08079c09 in ?? ()
-#4  0x080783e6 in ?? ()
-#5  0x0807823b in ?? ()
-#6  0x0807891c in ?? ()
-#7  0x08089544 in ?? ()
-#8  0x0807a185 in ?? ()
-#9  0x08079d4c in ?? ()
-#10 0x0804c921 in ?? ()
-#11 0x080768a0 in ?? ()
-#12 0x0804c8ec in ?? ()

I had to put the dash in front of # for the mailk to get out.
Does it say anything usefull?
Thanks for your patiences.
-- 
Peter

-
To unsubscribe from this list: send the line "unsubscribe linux-newbie" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.linux-learn.org/faqs

Re: Tkdesk Segmentation Fault

[Elias Athanasopoulos]

On Fri, Nov 08, 2002 at 10:13:07AM +0800, Peter wrote:
> I had to put the dash in front of # for the mailk to get out.
> Does it say anything usefull?
> Thanks for your patiences.

No. Sorry but I just realised that you have a problem with a Tcl/Tk 
script. Actually, you have to use gdb with the Tcl/Tk interpreter (not
sure which is that, it should be named in the first line of TkDesk).

Assuming the interpreter has been build with debugging info, it is called
'tcl' and it is in your path, do:

% gdb tcl
% (gdb) run /path/to/tkdesk

Normally, you'll get the segfault inside gdb, and using 'where' you'll get
the stack backtrace.

Elias

-- 
http://gnewtellium.sourceforge.net			MP3 is not a crime.	
-
To unsubscribe from this list: send the line "unsubscribe linux-newbie" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.linux-learn.org/faqs

need method of backing up files

[Shaggy Im-erbtham]

I use Slackware 7.0 with upgraded 2.4.18 kernel in a box which serves
as file server to 5 win9x clients.

Is there a method to perform incremental back-ups for every user, for
every day of the week.

Eg.  /home/peter
        /home/paul
        /home/mary

to be backed up everyday into day-of-the-week directories
so on Monday, files that were updated today would be copied to
each user's sub-directory in the monday directory.
/mnt/1monday/peter
/mnt/1monday/paul
/mnt/1monday/mary

and so on
/mnt/2tuesday/peter
/mnt/2tuesday/paul
/mnt/2tuesday/mary

Sometimes we need to track the daily changes that are made, that's
why.

I'm aware RAID is available but I think this is another thing altogether.

TIA
Shaggy

-
To unsubscribe from this list: send the line "unsubscribe linux-newbie" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.linux-learn.org/faqs

Re: need method of backing up files

[Paul Furness]

Well, there are so many ways of doing this that it's difficult to know
where to start!

There are two parts to what you want to do: scheduling when the backup
happens and doing the actual backup.

The scheduling is easy - use cron. man crontab will get you started on
that. One thing to be aware of is permissioning - if it's all on one
box, then root could run the backups fine, but you might not want to run
it as root, in which case a user like operator is used, but it runs the
actual backup binary SUID root.

Doing the backup:

A lot depends on how often you want to restore, and how you want to do
this - do you want to pull out individual files, or restore everything
in a give directory? How much data is there - is it necessary to
compress the backed up data?

If _ALL_ you want to do is copy files from one place to another once per
day, retaining permissions etc, then combine find with cpio. You can set
find to find only files which have been modified within the last day.
The commands you'd use (taking the first of the examples you give) would
be something like:

cd /home/peter
find . -depth -mtime -1 -print | cpio -pmudv /mnt/1monday/peter >
{Some_Log_File}



by using different options to cpio, you can also use this technique to
create a single archive file or go direct to a tape - say once per week
as a full backup. The man page for cpio should help you there.

 
If you want to simply create an archive of anything that changed since
the last backup, you could use dump. This is usually run once per week
as a full backup, and all the times in between as an incremental backup.
When run as an incremental, dump backs up all the files which have
changed since the last incremental backup of the same level. Again, look
at the man page for dump which explains it fairly well.


Paul.


On Fri, 2002-11-08 at 11:55, Shaggy Im-erbtham wrote:
> I use Slackware 7.0 with upgraded 2.4.18 kernel in a box which serves
> as file server to 5 win9x clients.
> 
> Is there a method to perform incremental back-ups for every user, for
> every day of the week.
> 
> Eg.  /home/peter
>         /home/paul
>         /home/mary
> 
> to be backed up everyday into day-of-the-week directories
> so on Monday, files that were updated today would be copied to
> each user's sub-directory in the monday directory.
> /mnt/1monday/peter
> /mnt/1monday/paul
> /mnt/1monday/mary
> 
> and so on
> /mnt/2tuesday/peter
> /mnt/2tuesday/paul
> /mnt/2tuesday/mary
> 
> Sometimes we need to track the daily changes that are made, that's
> why.
> 
> I'm aware RAID is available but I think this is another thing altogether.
> 
> TIA
> Shaggy
> 
> -
> To unsubscribe from this list: send the line "unsubscribe linux-newbie" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at http://www.linux-learn.org/faqs
> 
-- 
Paul Furness

Systems Manager

Steepness is an illusion caused by flat things leaning over.

-
To unsubscribe from this list: send the line "unsubscribe linux-newbie" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.linux-learn.org/faqs

RE: need method of backing up files

[Paul Kraus]

I did something similar with a backup script that copies only files that
change. However it does not take into account files that have been
deleted. You would then just need another simple script to run before
this one that would copy the backup directories to foldername-Tuesday
ect. If you use this you will need to make some minor changes to reflect
your environment. I have a bunch of other scripts that just call this
one. One for each machine I backup and one that is a batch of all the
machines I backup that I have sitting in cron.daily. 

Example: scriptname someuser somepassword computername computershare
pathwithinshare locationofpst

The last variable is for backing up outlook pst files but it can be
omitted as well as the pathwithinshare if you are backing up the root of
the share.

Let me know if you come up with a way to have it remove deleted files.

I am also working on a better way to handle logging.
Let me know if this helps.


The script in case you care :)
------------------------------
#Backup Script 
#------------
#Backup of Microsoft Windows Workstations using samba and cp.
#Script maps windows share to $mnt then copies the contents of share and
all 
#Directories beneath share. If file exists it is only over written if
source file is newer.
#Script unmounts drive and emails results to external server and local
account.
#Set variables
	#set locals
		dt=`date +%m.%d.%y\(%H.%M.%S\)`	
		externalmail=pkraus@pelsupply.com
		localmail=pdk
		msg=/backup/scripts/mailmessage

	#set login
		uid=$1
		pass=$2
		workstation=$3
		share=$4
		path=$5
		pst=$6

	#Set Paths
		source="/mnt/backup/${workstation}${path}"
		dest=/backup/$workstation
		mnt=/mnt/backup/$workstation
		log=/backup/logs/${workstation}.log.${dt}.txt
#Check to see if dest paths exists if not create
	if [ ! -d $dest ];then mkdir --parents "$dest";fi
	if [ ! -d $mnt ];then mkdir --parents "$mnt";fi

#tMount File Systew
	mount -t smbfs -o username=${uid},password=$pass
//${workstation}/$share $mnt > $log 2>/dev/null

#If mount return = 0 then copy if only newer then dest
	if [ $? -eq 0 ] ; then
		if [ -z "$5" ]; then
			cp --preserve --recursive --update "$source"/*
$dest 2>> $log
		else
			cp --preserve --recursive --update "$source"
$dest 2>> $log 
		fi

	#if pst location passed in then backup to backupdiretory/email
if /email missing add directory
		if [ -n "$6" ]; then
			if [ ! -d ${dest}/email ]; then mkdir --parents
"${dest}/email"; fi
			cp --preserve --recursive --update
"${mnt}$pst"*.pst "${dest}"/email 2>> $log
		fi	

		ds=`date +%m.%d.%y\(%H.%M.%S\)`
		subj="Backup log file for machine $workstation"
		
		echo "Backup started at $dt and ended at $ds" > $msg
		mutt -a $log -s "$subj" $externalmail $localmail < $msg
		rm -rf ${log}
		umount $mnt
	else
		echo $dt > $msg
		echo "Backup did not run for ${workstation}. Not
available!" >> ${msg}
		subj="Backup for machine $workstation FAILED!"
		mutt -a $log -s "$subj" $externalmail -c $localmail <
$msg
	fi

-----Original Message-----
From: linux-newbie-owner@vger.kernel.org
[mailto:linux-newbie-owner@vger.kernel.org] On Behalf Of Paul Furness
Sent: Friday, November 08, 2002 8:52 AM
To: Shaggy Im-erbtham
Cc: linux-newbie@vger.kernel.org
Subject: Re: need method of backing up files


Well, there are so many ways of doing this that it's difficult to know
where to start!

There are two parts to what you want to do: scheduling when the backup
happens and doing the actual backup.

The scheduling is easy - use cron. man crontab will get you started on
that. One thing to be aware of is permissioning - if it's all on one
box, then root could run the backups fine, but you might not want to run
it as root, in which case a user like operator is used, but it runs the
actual backup binary SUID root.

Doing the backup:

A lot depends on how often you want to restore, and how you want to do
this - do you want to pull out individual files, or restore everything
in a give directory? How much data is there - is it necessary to
compress the backed up data?

If _ALL_ you want to do is copy files from one place to another once per
day, retaining permissions etc, then combine find with cpio. You can set
find to find only files which have been modified within the last day.
The commands you'd use (taking the first of the examples you give) would
be something like:

cd /home/peter
find . -depth -mtime -1 -print | cpio -pmudv /mnt/1monday/peter >
{Some_Log_File}



by using different options to cpio, you can also use this technique to
create a single archive file or go direct to a tape - say once per week
as a full backup. The man page for cpio should help you there.

 
If you want to simply create an archive of anything that changed since
the last backup, you could use dump. This is usually run once per week
as a full backup, and all the times in between as an incremental backup.
When run as an incremental, dump backs up all the files which have
changed since the last incremental backup of the same level. Again, look
at the man page for dump which explains it fairly well.


Paul.


On Fri, 2002-11-08 at 11:55, Shaggy Im-erbtham wrote:
> I use Slackware 7.0 with upgraded 2.4.18 kernel in a box which serves 
> as file server to 5 win9x clients.
> 
> Is there a method to perform incremental back-ups for every user, for 
> every day of the week.
> 
> Eg.  /home/peter
>         /home/paul
>         /home/mary
> 
> to be backed up everyday into day-of-the-week directories
> so on Monday, files that were updated today would be copied to each 
> user's sub-directory in the monday directory. /mnt/1monday/peter
> /mnt/1monday/paul
> /mnt/1monday/mary
> 
> and so on
> /mnt/2tuesday/peter
> /mnt/2tuesday/paul
> /mnt/2tuesday/mary
> 
> Sometimes we need to track the daily changes that are made, that's 
> why.
> 
> I'm aware RAID is available but I think this is another thing 
> altogether.
> 
> TIA
> Shaggy
> 
> -
> To unsubscribe from this list: send the line "unsubscribe 
> linux-newbie" in the body of a message to majordomo@vger.kernel.org 
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at http://www.linux-learn.org/faqs
> 
-- 
Paul Furness

Systems Manager

Steepness is an illusion caused by flat things leaning over.

-
To unsubscribe from this list: send the line "unsubscribe linux-newbie"
in the body of a message to majordomo@vger.kernel.org More majordomo
info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.linux-learn.org/faqs

-
To unsubscribe from this list: send the line "unsubscribe linux-newbie" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.linux-learn.org/faqs

Re: need method of backing up files

[Mr. James W. Laferriere]

	Hello Shaggy ,  Look at rsync ,  works beautifully for me . JimL
	http://rsync.samba.org/index.html

On Fri, 8 Nov 2002, Shaggy Im-erbtham wrote:

> I use Slackware 7.0 with upgraded 2.4.18 kernel in a box which serves
> as file server to 5 win9x clients.
>
> Is there a method to perform incremental back-ups for every user, for
> every day of the week.
>
> Eg.  /home/peter
>         /home/paul
>         /home/mary
>
> to be backed up everyday into day-of-the-week directories
> so on Monday, files that were updated today would be copied to
> each user's sub-directory in the monday directory.
> /mnt/1monday/peter
> /mnt/1monday/paul
> /mnt/1monday/mary
>
> and so on
> /mnt/2tuesday/peter
> /mnt/2tuesday/paul
> /mnt/2tuesday/mary
>
> Sometimes we need to track the daily changes that are made, that's
> why.
>
> I'm aware RAID is available but I think this is another thing altogether.
>
> TIA
> Shaggy
>
> -
> To unsubscribe from this list: send the line "unsubscribe linux-newbie" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at http://www.linux-learn.org/faqs
>

--
       +------------------------------------------------------------------+
       | James   W.   Laferriere | System    Techniques | Give me VMS     |
       | Network        Engineer |     P.O. Box 854     |  Give me Linux  |
       | babydr@baby-dragons.com | Coudersport PA 16915 |   only  on  AXP |
       +------------------------------------------------------------------+

-
To unsubscribe from this list: send the line "unsubscribe linux-newbie" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.linux-learn.org/faqs

RE: need method of backing up files

[Paul Furness]

Yup, that should work nicely.

Couple of things you might want to add to it (or you may not, of course
:)  - see below

On Fri, 2002-11-08 at 14:35, Paul Kraus wrote:
> 
> Example: scriptname someuser somepassword computername computershare
> pathwithinshare locationofpst
> 
> 
> Let me know if you come up with a way to have it remove deleted files.
> 
Would you want to? How often does someone delete a file, then a couple
of days later ask you to restore it from the backup? I have something
similar going on where I simply copy stuff to the backup space, but
never delete anything. Well, not strictly; once per month it deletes
everything older than a month, using find:

find /{Backup_Place} -mtime +30 -exec rm {} >> {Log_File} \;

Old stuff is in the regular backups anyhow.


Oh, about logging: The approach I use is to log all activity and all
errors to the same file. This is because I find that if there _are_ any
problems, it's useful to have all the messages in chronological order.
Basically, every command in any of my system scripts ends in " >>
${Log_File} 2>&1 "  (actually, I usually do something like assign that
to a variable and stick that on each line).
At then end of the script, I usually put in something like:

mail -s "Backup script on ${Host} completed ok"
paul.furness@vil.ite.mee.com < ${Log}

and if there is a place in the script where things have gone wrong,
rather than just exitting, I put in something like:

mail -s "Backup script on ${Host} FAILED" paul.furness@vil.ite.mee.com <
${Log}


Hope some of this helps.

P.

> 
> 
> The script in case you care :)
> ------------------------------
> #Backup Script 
> #------------
.
.
.
> 		log=/backup/logs/${workstation}.log.${dt}.txt
> #Check to see if dest paths exists if not create
> 	if [ ! -d $dest ];then mkdir --parents "$dest";fi
> 	if [ ! -d $mnt ];then mkdir --parents "$mnt";fi

You might want to add something here that deals with a failure to make
the directory, and also logs the results of the mkdir command. Something
like:
 
 	if [ ! -d $dest ];then mkdir -v --parents "$dest" >> ${log};fi
	if [ ! -d $dest ]
		then 
		echo "Failure to make destination $dest >> ${log}
		#mail an error to someone?
		exit 1  # This ends the script if dest doesn't exist
	fi


........same for ${mnt}

It is possible to combine these into more compact code, but I think that
makes it harder to read later.

> 
> #tMount File Systew
> 	mount -t smbfs -o username=${uid},password=$pass
> //${workstation}/$share $mnt > $log 2>/dev/null

Is there any reason why you redirect errors to /dev/null rather than
appending to the log file? If it went wrong, you wouldn't know why.

> 
.
.
.
.

> 	fi


-- 
Paul Furness

Systems Manager

Steepness is an illusion caused by flat things leaning over.

-
To unsubscribe from this list: send the line "unsubscribe linux-newbie" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.linux-learn.org/faqs

RE: need method of backing up files

[Paul Kraus]

How can I add the machine name to the output. For instance...
This command writes to the log file that a file is in use and was not
backed up.

cp --preserve --recursive --update "$source"/* $dest 2>> $log

So if I have a variable called $machine and it equals the workstations
name how can I have it written before the log entry.

Example
Myworkstation: output from errors out.

Thanks in advance,
Paul Kraus

-
To unsubscribe from this list: send the line "unsubscribe linux-newbie" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.linux-learn.org/faqs

Re: Tkdesk Segmentation Fault

[lawson_whitney]

On Fri, 8 Nov 2002, Peter wrote:

>
> There your are.
>
> ]$ gdb tkdesk ./core.1875
>
> This GDB was configured as "i386-redhat-linux"..."/usr/local/bin/tkdesk": not
> in executable format: File format not recognized
>
> Core was generated by `tkdesksh /usr/local/bin/tkdesk'.

here is the program      ^^^^^^^^ that committed the fault.

> Program terminated with signal 11, Segmentation fault.
> -#0  0x0816e640 in ?? ()
> (gdb) where
> -#0  0x0816e640 in ?? ()
> -#1  0x4007502a in ?? ()
> -#2  0x400746f9 in ?? ()
> -#3  0x08079c09 in ?? ()
> -#4  0x080783e6 in ?? ()
> -#5  0x0807823b in ?? ()
> -#6  0x0807891c in ?? ()
> -#7  0x08089544 in ?? ()
> -#8  0x0807a185 in ?? ()
> -#9  0x08079d4c in ?? ()
> -#10 0x0804c921 in ?? ()
> -#11 0x080768a0 in ?? ()
> -#12 0x0804c8ec in ?? ()
>
> I had to put the dash in front of # for the mailk to get out.
> Does it say anything usefull?
> Thanks for your patiences.
>

gdb tkdesksh ./core.1875

--
---oops---



________________________________________________________________
Sign Up for Juno Platinum Internet Access Today
Only $9.95 per month!
Visit www.juno.com
-
To unsubscribe from this list: send the line "unsubscribe linux-newbie" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.linux-learn.org/faqs

Re: Tkdesk Segmentation Fault

[Peter]

 Thanks!

gdb tkdesksh ./core.1875
Core was generated by `tkdesksh /usr/local/bin/tkdesk'.
Program terminated with signal 11, Segmentation fault.
Reading symbols from /lib/libNoVersion.so.1...done.
Loaded symbols for /lib/libNoVersion.so.1
Reading symbols from /usr/X11R6/lib/libX11.so.6...done.
Loaded symbols for /usr/X11R6/lib/libX11.so.6
Reading symbols from /lib/libdl.so.2...done.
Loaded symbols for /lib/libdl.so.2
Reading symbols from /lib/i686/libm.so.6...done.
Loaded symbols for /lib/i686/libm.so.6
Reading symbols from /lib/i686/libc.so.6...done.
Loaded symbols for /lib/i686/libc.so.6
Reading symbols from /lib/ld-linux.so.2...done.
Loaded symbols for /lib/ld-linux.so.2
-#0  0x0816e640 in ?? ()
(gdb) where
-#0  0x0816e640 in ?? ()
-#1  0x4007502a in _XOpenLC () from /usr/X11R6/lib/libX11.so.6
-#2  0x400746f9 in XOpenIM () from /usr/X11R6/lib/libX11.so.6
-#3  0x08079c09 in OpenIM ()
-#4  0x080783e6 in GetScreen ()
-#5  0x0807823b in CreateTopLevelWindow ()
-#6  0x0807891c in TkCreateMainWindow ()
-#7  0x08089544 in TkCreateFrame ()
-#8  0x0807a185 in Initialize ()
-#9  0x08079d4c in Tk_Init ()
-#10 0x0804c921 in Tcl_AppInit ()
-#11 0x080768a0 in Tk_Main ()
-#12 0x0804c8ec in main ()
(gdb) 
(gdb) run tkdesksh
Starting program: /usr/local/bin/tkdesksh tkdesksh

Program received signal SIGSEGV, Segmentation fault.
0x0816e620 in ?? ()
(gdb) where
-#0  0x0816e620 in ?? ()
followed by above #1 to #12

(gdb) quit
The program is running.  Exit anyway? (y or n)

Now what does that all mean? Excuse my ignorance!

Regards




-- 
Peter

-
To unsubscribe from this list: send the line "unsubscribe linux-newbie" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.linux-learn.org/faqs

Re: Tkdesk Segmentation Fault

[Elias Athanasopoulos]

On Sun, Nov 10, 2002 at 12:24:22PM +0800, Peter wrote:
> -#1  0x4007502a in _XOpenLC () from /usr/X11R6/lib/libX11.so.6
> -#2  0x400746f9 in XOpenIM () from /usr/X11R6/lib/libX11.so.6
> -#3  0x08079c09 in OpenIM ()
> -#4  0x080783e6 in GetScreen ()
> -#5  0x0807823b in CreateTopLevelWindow ()
> -#6  0x0807891c in TkCreateMainWindow ()
> -#7  0x08089544 in TkCreateFrame ()
> -#8  0x0807a185 in Initialize ()
> -#9  0x08079d4c in Tk_Init ()
> -#10 0x0804c921 in Tcl_AppInit ()
> -#11 0x080768a0 in Tk_Main ()
> -#12 0x0804c8ec in main ()
> (gdb) 

Just a guess. Your X libs are incompatible with your Tcl/Tk.

Elias

-- 
http://gnewtellium.sourceforge.net			MP3 is not a crime.	
-
To unsubscribe from this list: send the line "unsubscribe linux-newbie" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.linux-learn.org/faqs

Re: Tkdesk Segmentation Fault

[Peter]

eathan@otenet.gr said:
> Just a guess. Your X libs are incompatible with your Tcl/Tk.

Thank you so much and Lawson for the lead!!!

You guessed it right. Copying libX11.so.6.1 into /usr/X11R6/lib/ made tkdesk 
coming up instantly.

Hurray!

Regards
-- 
Peter

-
To unsubscribe from this list: send the line "unsubscribe linux-newbie" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.linux-learn.org/faqs

Re: Tkdesk Segmentation Fault

[Heimo Claasen]

On a related note:
Some accident/repair event apparently had the effect to trash what
would be just one of the "Gdk" libraries.  (At least, this was the only
fault discovered hitherto after the repair.)

Indication is that just one application complains, with some
"Gtk-CRITICAL **" error, evidently some file is lacking.

Searching through the pertinent distribution CDs I could not find
anything apparenly relevant to reload/re-install.

What is the real name for those "Gtk-" libraries ?

// Heimo Claasen // <hammer at revobild dot net> // Brussels 2002-11-11
The WebPlace of ReRead - and much to read  ==>  http://www.revobild.net

-
To unsubscribe from this list: send the line "unsubscribe linux-newbie" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.linux-learn.org/faqs

Re: Tkdesk Segmentation Fault

[Elias Athanasopoulos]

On Tue, Nov 12, 2002 at 10:17:19AM +0000, Heimo Claasen wrote:
> Indication is that just one application complains, with some
> "Gtk-CRITICAL **" error, evidently some file is lacking.
> 
> Searching through the pertinent distribution CDs I could not find
> anything apparenly relevant to reload/re-install.
> 
> What is the real name for those "Gtk-" libraries ?

Don't bother. In most of the cases these are cases that haven't predicted
by the programmers and apparently drive into 'inoccent' error messages.

% ldd /path/to/gtkfoo

To see which libs are dynamically linked to your gtk foo application.

Elias

-- 
http://gnewtellium.sourceforge.net			MP3 is not a crime.	
-
To unsubscribe from this list: send the line "unsubscribe linux-newbie" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.linux-learn.org/faqs

Re: Tkdesk Segmentation Fault

[lawson_whitney]

On 12 Nov 2002, Heimo Claasen wrote:

> On a related note:
> Some accident/repair event apparently had the effect to trash what
> would be just one of the "Gdk" libraries.  (At least, this was the only
> fault discovered hitherto after the repair.)
>
> Indication is that just one application complains, with some
> "Gtk-CRITICAL **" error, evidently some file is lacking.
>
> Searching through the pertinent distribution CDs I could not find
> anything apparenly relevant to reload/re-install.
>
> What is the real name for those "Gtk-" libraries ?
>
> // Heimo Claasen // <hammer at revobild dot net> // Brussels 2002-11-11
> The WebPlace of ReRead - and much to read  ==>  http://www.revobild.net
>
You'll probably find them in the gnome-libs package.  That or gtk+, I
guess.
--
---oops---



________________________________________________________________
Sign Up for Juno Platinum Internet Access Today
Only $9.95 per month!
Visit www.juno.com
-
To unsubscribe from this list: send the line "unsubscribe linux-newbie" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.linux-learn.org/faqs

RE: need method of backing up files

[Paul Furness]

I don't know of a way to make it put the machine name at the beginning
of each line. 

As a partial solution, you could put the machine name into the log file
before the copy:

  echo ${machinename} >> $log
  cp --preserve ..... and so on.

I tend to put a lot of date stamps in log files as well as some
comments. It doesn't make much difference to the file size, and it can
be very useful information later, and also make the log file much easier
to read. I tend to do something like this:

  echo "Starting the backup of ${machinename} at `date`" >> $log
  cp --preserve .....

P.


On Fri, 2002-11-08 at 21:37, Paul Kraus wrote:
> How can I add the machine name to the output. For instance...
> This command writes to the log file that a file is in use and was not
> backed up.
> 
> cp --preserve --recursive --update "$source"/* $dest 2>> $log
> 
> So if I have a variable called $machine and it equals the workstations
> name how can I have it written before the log entry.
> 
> Example
> Myworkstation: output from errors out.
> 
> Thanks in advance,
> Paul Kraus
> 
> -
> To unsubscribe from this list: send the line "unsubscribe linux-newbie" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at http://www.linux-learn.org/faqs
> 
-- 
Paul Furness

Systems Manager

Steepness is an illusion caused by flat things leaning over.

-
To unsubscribe from this list: send the line "unsubscribe linux-newbie" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.linux-learn.org/faqs