NetBSD-1.4: something wrong?

NetBSD-1.4: something wrong?

Post by Emmanuel Dreyf » Fri, 14 May 1999 04:00:00



Hello

I upgraded my mac68k and my macppc to NetBSD-1.4. Since I did it, both
are crash city!

On the macppc, it does not boot anymore. I crashes just after the OF
loader messages. I thought it could be a problem with the kernel,
therefore I booted with the floppy, mounted my hard disk, and started to
recompile the kernel.

I can't do it! I get errors during compilation or link time. It's the
first time I see a NetBSD kernel haveing problem to compile.

On the mac68k, I installed in multiuser mode with tar -xvzp --unlink.
I have lots of daemonds crashing (ypserv, portmap, xntpd) with no
reason. I think it might be because I should not have unpacked in
multiuser mode

But I have more serious problems: perl does not work any more (It dumps
core), and the package does not compile.  Netatalk crashes too... I
don't understand what is going on: the generic kernel is supposed to
have 1.3.x backward compability, is'nt it?

Other problems (even more serious): my NAT does not work for one of my
machines (not translated outside), and the worst one: ftpd refuses any
login!

220 gizmo FTP server (Version: 7.1.0) ready.
Name (gizmo:manu): manu
331 Password required for manu.
Password:
421 Service not available, remote server has closed connection.
ftp: Login failed.

I'm thninking about reverting to 1.3.3...

--
Emmanuel Dreyfus.  
Pas de processeur Intel, pas de logiciels Microsoft:
Des programmes sains dans un ordinateur sain.

 
 
 

NetBSD-1.4: something wrong?

Post by Manuel Bouy » Tue, 18 May 1999 04:00:00



> Hello

> I upgraded my mac68k and my macppc to NetBSD-1.4. Since I did it, both
> are crash city!

> On the macppc, it does not boot anymore. I crashes just after the OF
> loader messages. I thought it could be a problem with the kernel,
> therefore I booted with the floppy, mounted my hard disk, and started to
> recompile the kernel.

> I can't do it! I get errors during compilation or link time. It's the
> first time I see a NetBSD kernel haveing problem to compile.

Sure, it should'nt happen. What kind or problem are these ? random
core dumps ?

Quote:

> On the mac68k, I installed in multiuser mode with tar -xvzp --unlink.
> I have lots of daemonds crashing (ypserv, portmap, xntpd) with no
> reason. I think it might be because I should not have unpacked in
> multiuser mode

No, shouldn't be a problem. I've done it several times.
How do they crash ? core dump ?

Quote:

> But I have more serious problems: perl does not work any more (It dumps
> core), and the package does not compile.  Netatalk crashes too... I
> don't understand what is going on: the generic kernel is supposed to
> have 1.3.x backward compability, is'nt it?

It should. But this can be related to the same problem you're seeing.

Quote:

> Other problems (even more serious): my NAT does not work for one of my
> machines (not translated outside), and the worst one: ftpd refuses any
> login!

> 220 gizmo FTP server (Version: 7.1.0) ready.
> Name (gizmo:manu): manu
> 331 Password required for manu.
> Password:
> 421 Service not available, remote server has closed connection.
> ftp: Login failed.

Is there something in the logs ?

--

--

 
 
 

NetBSD-1.4: something wrong?

Post by Emmanuel Dreyf » Tue, 18 May 1999 04:00:00


[kernel does not compile]

Quote:> Sure, it should'nt happen. What kind or problem are these ? random
> core dumps ?

No, syntax and linking errors. I unpacked the whole sources again, and
it now compiles without any problem. The only difference is that the
first time I overwrote the NetBSD-1.3.3 /usr/src/sys directory, and on
the second time I did erase it before unpacking syssrc.tgz.

[daemons crashing]

Quote:> No, shouldn't be a problem. I've done it several times.
> How do they crash ? core dump ?

Yes, massive core dumping! This issue is now also OK. I had to unpack
base.tgz again (it's the same base.tgz file!)
The only problem that still existed was ftpd and perl crashing. I fixed
the perl problem by recompiling it. Ftpd still crashes.

Quote:> > 220 gizmo FTP server (Version: 7.1.0) ready.
> > Name (gizmo:manu): manu
> > 331 Password required for manu.
> > Password:
> > 421 Service not available, remote server has closed connection.
> > ftp: Login failed.

> Is there something in the logs ?

in /var/mail/authlog

May 13 10:57:38 gizmo ftpd[10274]: FTP LOGIN FAILED FROM 157.159.10.10,
manu^?

Nothing else.
--
Emmanuel Dreyfus

 
 
 

NetBSD-1.4: something wrong?

Post by Manuel Bouy » Wed, 19 May 1999 04:00:00



> ftpd -C manu returns nothing at all

It returns a status: 0 or 1.
In csh or tcsh, the command would be:
ftpd -C manu
echo $status

--

--

 
 
 

NetBSD-1.4: something wrong?

Post by Emmanuel Dreyf » Wed, 19 May 1999 04:00:00



> In csh or tcsh, the command would be:
> ftpd -C manu
> echo $status


You have mail.
gizmo: {1} /usr/libexec/ftpd -C manu
gizmo: {2} echo $status
0

What does that mean?
--
Emmanuel Dreyfus

 
 
 

NetBSD-1.4: something wrong?

Post by Manuel Bouy » Wed, 19 May 1999 04:00:00



> No, syntax and linking errors. I unpacked the whole sources again, and
> it now compiles without any problem. The only difference is that the
> first time I overwrote the NetBSD-1.3.3 /usr/src/sys directory, and on
> the second time I did erase it before unpacking syssrc.tgz.

Ok this is because there where some old files left, which were used instead
of the new one.

Quote:

> [daemons crashing]
> > No, shouldn't be a problem. I've done it several times.
> > How do they crash ? core dump ?

> Yes, massive core dumping! This issue is now also OK. I had to unpack
> base.tgz again (it's the same base.tgz file!)
> The only problem that still existed was ftpd and perl crashing. I fixed
> the perl problem by recompiling it. Ftpd still crashes.

Well, I can't see a reason for this.

Quote:> > Is there something in the logs ?

> in /var/mail/authlog

> May 13 10:57:38 gizmo ftpd[10274]: FTP LOGIN FAILED FROM 157.159.10.10,
> manu^?

> Nothing else.

Ok, sounds like ftpd just refused the login.
Did you check all the ftpd config file, /etc/shells, /etc/ftpusers,
/etc/ftpchroot, /etc/ftpd.conf, ...
Did you try a 'ftpd -C manu' and checked the status ?

--

--

 
 
 

NetBSD-1.4: something wrong?

Post by Emmanuel Dreyf » Wed, 19 May 1999 04:00:00



> Ok, sounds like ftpd just refused the login.
> Did you check all the ftpd config file, /etc/shells, /etc/ftpusers,
> /etc/ftpchroot, /etc/ftpd.conf, ...
> Did you try a 'ftpd -C manu' and checked the status ?

ftpd -C manu returns nothing at all

           checkportcmd  all
           display       none
           maxtimeout    all    7200   # 2 hours
           modify        all
           modify        guest  off
           notify        none
           timeout       all    900    # 15 minutes
           umask         all    027
           umask         guest  0707

#       $NetBSD: ftpusers,v 1.4 1997/09/08 02:51:58 mikel Exp $
#
# list of users denied (or allowed) ftp access
# read by ftpd(8).
root    deny
altroot deny
toor    deny
uucp    deny
scroc   deny
leto    deny
hx      deny
tatoo   deny
newsmail        deny
ftpmail deny
anonymous       allow
*       allow
--
Emmanuel Dreyfus

 
 
 

NetBSD-1.4: something wrong?

Post by Manuel Bouy » Thu, 20 May 1999 04:00:00




> You have mail.
> gizmo: {1} /usr/libexec/ftpd -C manu
> gizmo: {2} echo $status
> 0

> What does that mean?

"read the doc" !
This means that user "manu" is allowed to ftp in. The problem is somewhere
else.

There is something else wrong in your setup - 1.4 ftp works for me.
Check permissions on directories and files, maybe that's it.

--

--

 
 
 

NetBSD-1.4: something wrong?

Post by Emmanuel Dreyf » Thu, 20 May 1999 04:00:00



> There is something else wrong in your setup - 1.4 ftp works for me.
> Check permissions on directories and files, maybe that's it.

All -rw-r--r--
Maybe a library related problem? I have lots of library related problems
since I upgraded...
--
Emmanuel Dreyfus

 
 
 

NetBSD-1.4: something wrong?

Post by Emmanuel Dreyf » Thu, 20 May 1999 04:00:00



> Maybe a library related problem? I have lots of library related problems
> since I upgraded...

Anyway, I cannot get ftpd outputing log messages anywhere. I added a
 ftp.*                  /var/log/messages
to /etc/syslogd.conf and sent a kill -1 to syslogd, and I still get no
informations from ftp when a login fail.

--
Emmanuel Dreyfus

 
 
 

NetBSD-1.4: something wrong?

Post by Manuel Bouy » Thu, 20 May 1999 04:00:00



> All -rw-r--r--
> Maybe a library related problem? I have lots of library related problems
> since I upgraded...

Maybe. What does 'ldd /usr/libexec/ftpd' says ?
Mine says:
antioche:/home/ftp/etc>ldd /usr/libexec/ftpd
/usr/libexec/ftpd:
    -lcrypt.0 => /usr/lib/libcrypt.so.0.0 (0x40023000)
    -lskey.0 => /usr/lib/libskey.so.0.0 (0x40029000)
    -lc.12 => /usr/lib/libc.so.12.40 (0x4002e000)

--

--

 
 
 

NetBSD-1.4: something wrong?

Post by Manuel Bouy » Thu, 20 May 1999 04:00:00



> Anyway, I cannot get ftpd outputing log messages anywhere. I added a
>  ftp.*                  /var/log/messages
> to /etc/syslogd.conf and sent a kill -1 to syslogd, and I still get no
> informations from ftp when a login fail.

Are you sure ftpd is started at all ?
Try to kill inetd and restart it with '-d'.

--

--

 
 
 

NetBSD-1.4: something wrong?

Post by Emmanuel Dreyf » Thu, 20 May 1999 04:00:00



> Are you sure ftpd is started at all ?

The client does ask me for my login and password. I assume that the
server must be asking him that. I am wrong?

Quote:> Try to kill inetd and restart it with '-d'.


ADD : ftp proto=tcp, wait.max=0.40, user.group=root.(null) builtin=0
server=/usr
/libexec/ftpd
registered /usr/libexec/ftpd on 6

(...)

someone wants ftp
accept, ctrl 5
6230 execl /usr/libexec/ftpd    <<seems to work.
6230 reaped, status 0xb         <<this is when ftpd reject my login

And, the client messages:
220 gizmo FTP server (Version: 7.1.0) ready.
Name (gizmo:manu): manu
331 Password required for manu.
Password:
421 Service not available, remote server has closed connection.
ftp: Login failed.
ftp: No control connection for command.
ftp>

--
Emmanuel Dreyfus

 
 
 

NetBSD-1.4: something wrong?

Post by Emmanuel Dreyf » Thu, 20 May 1999 04:00:00



> Maybe. What does 'ldd /usr/libexec/ftpd' says ?
> Mine says:
> antioche:/home/ftp/etc>ldd /usr/libexec/ftpd
> /usr/libexec/ftpd:
>     -lcrypt.0 => /usr/lib/libcrypt.so.0.0 (0x40023000)
>     -lskey.0 => /usr/lib/libskey.so.0.0 (0x40029000)
>     -lc.12 => /usr/lib/libc.so.12.40 (0x4002e000)

Nearly the same at mine:


/usr/libexec/ftpd:
        -lcrypt.0 => /usr/lib/libcrypt.so.0.0 (0x4026000)
        -lskey.0 => /usr/lib/libskey.so.0.0 (0x402c000)
        -lc.12 => /usr/lib/libc.so.12.40 (0x4032000)

--
Emmanuel Dreyfus

 
 
 

NetBSD-1.4: something wrong?

Post by Manuel Bouy » Fri, 21 May 1999 04:00:00



> The client does ask me for my login and password. I assume that the
> server must be asking him that. I am wrong?

No, you're rigth,

You said you  didn't get anything in /var/log/message, although you had
ftp.* redirected to it.
Could you start ftpd with -lld flags in inetd.conf and see if it helps ?

--

--

 
 
 

1. Something wrong: defunct eth0

Greetings.

I have a machine with two ethernet cards: eth0 with a static IP (lets
say 1.2.3.4), connected to the Internet, and eth1 with an internal
assigned IP (192.168.1.1). My ISP (for those who live in Portugal who
are reading, it's Netcabo) assigns this static IP through DHCP or
BOOTP (I really don't know anymore). I'm running Red Hat Linux 7.1 and
I use pump.

The problem is, just freshly after boot-up, I do a 'ps -A | grep eth'
and the following appears:

445 ?        00:00:00 eth0 <defunct>
446 ?        00:00:00 eth0
448 ?        00:00:00 eth1

The problem with this is that after a while the process eth0 simply
terminates, leaving me without Internet connection. I have to start
pump again manually, etc, etc...

I have iptables running, and with almost 100% sureness its because I'm
not accepting some kind of incoming connections from my ISPs server
that allow the checking of the IP, or something...

I have this rule in my firewall:
iptables -A INPUT -p udp --sport 68 --dport 67 -j ACCEPT

And the following is an excerpt of '/var/log/messages' just after the
bootup - everything looks fine to me!
Oct 31 18:42:14 myhost network: Setting network parameters:  succeeded
Oct 31 18:42:14 myhost network: Bringing up interface lo:  succeeded
Oct 31 18:42:15 myhost ifup: Determining IP information for eth0...
Oct 31 18:42:15 myhost pumpd[372]: starting at (uptime 0 days,
0:00:28) Wed Oct 31 18:42:15 2001
Oct 31 18:42:19 myhost pumpd[372]: configured interface eth0
Oct 31 18:42:19 myhost ifup:  done.
Oct 31 18:42:19 myhost network: Bringing up interface eth0:  succeeded
Oct 31 18:42:20 myhost network: Bringing up interface eth1:  succeeded

What am I missing here? What the heck is 'eth0 <defunct>' when I do
the 'ps'?
Best regards.

2. rz program not working!

3. Am I doing something wrong?? - (SAMBA)

4. (patch for Bash) disk-based associative array using gdbm(3)

5. HELP! Something wrong with root fs!

6. know a svgalib www-browser? (for the console)

7. HELP Something is very wrong

8. Linux thin client connecting to Windows 2003 terminal server

9. Something wrong with the AHA-2742 driver!?

10. Something is wrong.

11. Something terribly wrong with RPM or the latest Redhat updates

12. Did I do something wrong with RPM?

13. please help me. my program has something wrong.