Prev: [Samba] Winbind and getent issues
Next: Samba 3.5.2 : root and Domain Admins membership
From: Rob Moser on 11 May 2010 18:40
Hi folks,

# uname -srvo
Linux 2.6.18-194.el5xen #1 SMP Tue Mar 16 22:01:26 EDT 2010 GNU/Linux

# cat /etc/redhat-release
Red Hat Enterprise Linux Server release 5.5 (Tikanga)

# smbd --version
Version 3.3.8-0.51.el5

I'm trying to run a backup samba server using some custom in-house
scripts which I inherited and am stuck with. Essentially the primary
server periodically rsyncs a bunch of files over to the backup,
including a collection of the tdb files. When I try to connect to the
backup server, I get the following smbd panics in the logs:

-----
[2010/05/11 14:34:32, 0] lib/util_tdb.c:tdb_log(682)
tdb(/var/lib/samba/ntforms.tdb): tdb_reopen: file dev/inode has changed!
[2010/05/11 14:34:32, 0] lib/util.c:reinit_after_fork(1054)
tdb_reopen_all failed.
[2010/05/11 14:34:32, 0] printing/print_cups.c:cups_pcap_load_async(432)
cups_pcap_load_async: reinit_after_fork() failed
[2010/05/11 14:34:32, 0] lib/util.c:smb_panic(1673)
PANIC (pid 28994): cups_pcap_load_async: reinit_after_fork() failed
[2010/05/11 14:34:32, 0] lib/util_tdb.c:tdb_log(682)
tdb(/var/lib/samba/ntforms.tdb): tdb_reopen: file dev/inode has changed!
[2010/05/11 14:34:32, 0] lib/util.c:reinit_after_fork(1054)
tdb_reopen_all failed.
[2010/05/11 14:34:32, 0] smbd/server.c:open_sockets_smbd(774)
reinit_after_fork() failed
[2010/05/11 14:34:32, 0] lib/util.c:smb_panic(1673)
PANIC (pid 28995): reinit_after_fork() failed
[2010/05/11 14:34:32, 0] lib/util.c:log_stack_trace(1777)
BACKTRACE: 9 stack frames:
#0 smbd(log_stack_trace+0x1c) [0x2b49e45a126c]
#1 smbd(smb_panic+0x2b) [0x2b49e45a134b]
#2 smbd(cups_cache_reload+0x1b1) [0x2b49e4571d11]
#3 smbd(pcap_cache_reload+0xb8) [0x2b49e456e6f8]
[2010/05/11 14:34:32, 0] lib/util.c:log_stack_trace(1777)
BACKTRACE: 5 stack frames:
#4 smbd(reload_printers+0x25) [0x2b49e47ab3c5]
#0 smbd(log_stack_trace+0x1c) [0x2b49e45a126c]
#5 smbd(check_reload+0xe8) [0x2b49e43e7378]
#1 smbd(smb_panic+0x2b) [0x2b49e45a134b]
#6 smbd(main+0x1422) [0x2b49e47acee2]
#2 smbd(main+0x19bf) [0x2b49e47ad47f]
#7 /lib64/libc.so.6(__libc_start_main+0xf4) [0x2b49e7c3e994]
#3 /lib64/libc.so.6(__libc_start_main+0xf4) [0x2b49e7c3e994]
#8 smbd [0x2b49e43717c9]
#4 smbd [0x2b49e43717c9]
[2010/05/11 14:34:32, 0] lib/fault.c:dump_core(231)
[2010/05/11 14:34:32, 0] lib/fault.c:dump_core(231)
dumping core in /var/log/samba/cores/smbd
dumping core in /var/log/samba/cores/smbd
[2010/05/11 14:34:32, 2] printing/print_cups.c:cups_async_callback(544)
cups_async_callback: failed to read a new printer list
[2010/05/11 14:34:33, 0] lib/util_tdb.c:tdb_log(682)
tdb(/var/lib/samba/ntforms.tdb): tdb_reopen: file dev/inode has changed!
[2010/05/11 14:34:33, 0] lib/util.c:reinit_after_fork(1054)
tdb_reopen_all failed.
[2010/05/11 14:34:33, 0] smbd/server.c:open_sockets_smbd(774)
reinit_after_fork() failed
[2010/05/11 14:34:33, 0] lib/util.c:smb_panic(1673)
PANIC (pid 28996): reinit_after_fork() failed
[2010/05/11 14:34:33, 0] lib/util.c:log_stack_trace(1777)
BACKTRACE: 5 stack frames:
#0 smbd(log_stack_trace+0x1c) [0x2b49e45a126c]
#1 smbd(smb_panic+0x2b) [0x2b49e45a134b]
#2 smbd(main+0x19bf) [0x2b49e47ad47f]
#3 /lib64/libc.so.6(__libc_start_main+0xf4) [0x2b49e7c3e994]
#4 smbd [0x2b49e43717c9]
[2010/05/11 14:34:33, 0] lib/fault.c:dump_core(231)
dumping core in /var/log/samba/cores/smbd
-----

Uhm... no kidding the inode has changed? Its on a different machine!
From scrounging around the web, this looks remarkably similar to <a
href="https://bugzilla.samba.org/show_bug.cgi?id=5976">bug 5976</a>, but
a fix for that supposedly went into 3.3.0. And a reference to a
different patch for it going into 3.3.0 in <a
href="http://www.mail-archive.com/samba(a)lists.samba.org/msg98122.html">the
list archives for this list.</a> I installed from a non-source package,
so I can't check to see if either patch is there or not. Still, rather
than re-open a two-year-old bug on the presumption that I am the only
person that its broken for since, I thought I should check and see if
anyone had any insight into it first... any thoughts? Is it just broken
of these scripts to be trying to move tdb files across machines in the
first place? If I wipe them clean and just copy the configuration
files, will I keep all of my configuration?

Thanks for any advice,

- rob.
--
To unsubscribe from this list go to the following URL and read the
instructions: https://lists.samba.org/mailman/options/samba
 | 
Pages: 1
Prev: [Samba] Winbind and getent issues
Next: Samba 3.5.2 : root and Domain Admins membership