Re: [Samba] net rpc user add produces duplicate SID

2010-01-21 Thread Sebastian Scholz
Sebastian Scholz gsscholz at gmail.com writes:

 
 
 I moved the domain SID and the local SID to the new pdc and filled the ldap
 directory with all users and machines from the old machine. I can access the 
 smb via smbclient, id user works and getent passwd, etc. 
 
 BUT when I try to add a new user with 
 # net rpc user add newusername 
 the user gets a SID which is already used by a machine account. The command
 responds with Faild to add user 'newusername' with: No such user.
 # pdbedit -L 
 reports this user and this machine account with ldapsam_getsampwsid: More 
 than one user with SID [S-1-5-21]. Failing. count=2. Deleting the user 
 with
 # net rpc user delete newusername 
 works.
 

Hi List

the same actually happens when I add a new machine to the domain with
net dom join


Can I change the SID/RID somehow after adding a new account or is there a 
way to tell net to start from a different RID?

Thanks
Sebastian

-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba


Re: [Samba] net rpc user add produces duplicate SID

2010-01-21 Thread Sebastian Scholz
Sebastian Scholz gsscholz at gmail.com writes:

 
 Sebastian Scholz gsscholz at gmail.com writes:
 
  
 
  I moved the domain SID and the local SID to the new pdc and filled the ldap
  directory with all users and machines from the old machine. I can access  
  the smb via smbclient, id user works and getent passwd, etc. 
  
  BUT when I try to add a new user with 
  # net rpc user add newusername 
  the user gets a SID which is already used by a machine account. The command
  responds with Faild to add user 'newusername' with: No such user.
  # pdbedit -L 
  reports this user and this machine account with ldapsam_getsampwsid: More 
  than one user with SID [S-1-5-21]. Failing. count=2. Deleting the user 
  with
  # net rpc user delete newusername 
  works.
  
 
 Hi List
 
 the same actually happens when I add a new machine to the domain with
 net dom join
 
 Can I change the SID/RID somehow after adding a new account or is there a 
 way to tell net to start from a different RID?
 
 Thanks
 Sebastian
 

Me again,

I hot fixed the problem for me by adding and deleting a dummy user a couple of
times:

# for i in `seq 0 50 ` ; do 
#  net rpc user add username  -U root%geheim
#  net rpc user delete username -U root%geheim
# done

but I still guess this is a bug in the net command.

Regards
Sebastian



-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba


[Samba] net rpc user add produces duplicate SID

2009-12-23 Thread Sebastian Scholz
Hi list,

The task is to move an old samba installation (ubuntu with samba 3.0 brachch) to
a new debian installation with samba 3.3.9. Both setups work as pdcs and have an
ldap backend which I try to move from one setup to the next.

In this new setup I tried to ldapsam:editposix = yes. Therefore I also installed
winbind.

I moved the domain SID and the local SID to the new pdc and filled the ldap
directory with all users and machines from the old machine. I can access the smb
via smbclient, id user works and getent passwd, etc. 

BUT when I try to add a new user with 
# net rpc user add newusername 
the user gets a SID which is already used by a machine account. The command
responds with Faild to add user 'newusername' with: No such user.
# pdbedit -L 
reports this user and this machine account with ldapsam_getsampwsid: More than
one user with SID [S-1-5-21]. Failing. count=2. Deleting the user via 
# net rpc user delete newusername 
works.

I know how to increase the used uid but I don't know to let winbind use the next
free RID.

Extracts from my smb.conf:
passdb backend = ldapsam
ldap suffix is set
ldap admin dn is set
ldap user/group/machine/idmap suffix are set
ldapsam:trusted = yes
ldapsam:editposix = yes
idmap uid = 1 - 2
idmap gid = 1 - 2


Hope anyone can help and merry christmas to the list
Sebastian

-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba


[Samba] Samba freezes server when writing 50MB MS Word document (sometimes)

2009-07-11 Thread Sebastian Scholz
Hi list

for a couple of days we have the problem that our samba pdc server (ubuntu
8.04
LTS [kernel 2.6.24-24] with 3.0.28a samba package) freezes sometimes when
someone saves a specific word document (about 50MB) from MS Word. It's not
always the same person. Saving the document - as nearly any huge MS Office
document - takes some time, about one minute (guessed).

First the server gets really slow. Some user programs on clients fail to
write.
The CPU usage of the smb process of the user writing the document goes up to
nearly 100%. After a while it's not possible to log on to the server - even
locally. It asks for a password but there won't be a prompt. Sometimes there
is
nothing on the console/screen. Hitting a button does not change a thing.
It's
not possible to kill the process. Another thing is that there are a couple
of
smbd processes for one user (see attached user smbd log).

When rebooting the server it does not shut down properly. I guess the
process
blocks the harddrive thus it cannot be unmounted. A ALT+Print+u and
ALT+Print+b
helps.

###

This is what the syslog says:

Jul  3 15:01:44 [..] BUG: soft lockup - CPU#1 stuck for 11s! [smbd:23017]
Jul  3 15:01:44 [..]
Jul  3 15:01:44 [..] Pid: 23017, comm: smbd Not tainted (2.6.24-24-server
#1)
Jul  3 15:01:44 [..] EIP: 0060:[f98c1e0c] EFLAGS: 0203 CPU: 1
Jul  3 15:01:44 [..] EIP is at xfs_iext_bno_to_ext+0x8c/0x1f0 [xfs]
Jul  3 15:01:44 [..] EAX: 00ba EBX: 2404 ECX: 00ba EDX: 00b5
Jul  3 15:01:44 [..] ESI:  EDI: 51c7 EBP: 0001 ESP: ee5298a4
Jul  3 15:01:44 [..]  DS: 007b ES: 007b FS: 00d8 GS: 0033 SS: 0068
Jul  3 15:01:44 [..] CR0: 8005003b CR2: b7f2f000 CR3: 3577d000 CR4: 06b0
Jul  3 15:01:44 [..] DR0:  DR1:  DR2:  DR3: 
Jul  3 15:01:44 [..] DR6: 0ff0 DR7: 0400
Jul  3 15:01:44 [..]  [f989af59] xfs_bmap_search_multi_extents+0x69/0x100
[xfs]
Jul  3 15:01:44 [..]  [f989b05f] xfs_bmap_search_extents+0x6f/0x150 [xfs]
Jul  3 15:01:44 [..]  [apic_timer_interrupt+0x28/0x30]
apic_timer_interrupt+0x28/0x30
Jul  3 15:01:44 [..]  [f98a4714] xfs_bmapi+0x1e4/0x16a0 [xfs]
Jul  3 15:01:44 [..]  [net_rx_action+0x12d/0x210] net_rx_action+0x12d/0x210
Jul  3 15:01:44 [..]  [run_timer_softirq+0x30/0x1e0]
run_timer_softirq+0x30/0x1e0
Jul  3 15:01:44 [..]  [tick_program_event+0x38/0x60]
tick_program_event+0x38/0x60
Jul  3 15:01:44 [..]  [__do_softirq+0x82/0x110] __do_softirq+0x82/0x110
Jul  3 15:01:44 [..]  [f98c83c9]
xfs_iomap_eof_want_preallocate+0x149/0x220 [xfs]
Jul  3 15:01:44 [..]  [f98c8bd3] xfs_iomap_write_delay+0x153/0x3e0 [xfs]
Jul  3 15:01:44 [..]  [f98c7fe7] xfs_iomap+0x277/0x510 [xfs]
Jul  3 15:01:44 [..]  [f98c1eda] xfs_iext_bno_to_ext+0x15a/0x1f0 [xfs]
Jul  3 15:01:44 [..]  [f98c8274] xfs_iomap+0x504/0x510 [xfs]
Jul  3 15:01:44 [..]  [f98e8393] __xfs_get_blocks+0x73/0x2b0 [xfs]
Jul  3 15:01:44 [..]  [f98e8628] xfs_get_blocks+0x28/0x30 [xfs]
Jul  3 15:01:44 [..]  [__block_prepare_write+0x228/0x400]
__block_prepare_write+0x228/0x400
Jul  3 15:01:44 [..]  [sg:__alloc_pages+0x5a/0x380] __alloc_pages+0x5a/0x380
Jul  3 15:01:44 [..]  [ext3:__grab_cache_page+0x97/0x1a90]
__grab_cache_page+0x97/0xc0
Jul  3 15:01:44 [..]  [ext3:block_write_begin+0x4d/0x590]
block_write_begin+0x4d/0xf0
Jul  3 15:01:44 [..]  [f98e8600] xfs_get_blocks+0x0/0x30 [xfs]
Jul  3 15:01:44 [..]  [f98e7da6] xfs_vm_write_begin+0x46/0x50 [xfs]
Jul  3 15:01:44 [..]  [f98e8600] xfs_get_blocks+0x0/0x30 [xfs]
Jul  3 15:01:44 [..]  [xfs:generic_file_buffered_write+0x117/0xcf0]
generic_file_buffered_write+0x117/0x630
Jul  3 15:01:44 [..]  [f98f0880] xfs_write+0x830/0x910 [xfs]
Jul  3 15:01:44 [..]  [f98ec6cc] xfs_file_aio_write+0x5c/0x70 [xfs]
Jul  3 15:01:44 [..]  [xfs:do_sync_write+0xd5/0x120]
do_sync_write+0xd5/0x120
Jul  3 15:01:44 [..]  [c0145fa0] autoremove_wake_function+0x0/0x40
Jul  3 15:01:44 [..]  [c0145fa0] autoremove_wake_function+0x0/0x40
Jul  3 15:01:44 [..]  [set_current_groups+0x25e/0x270]
set_current_groups+0x25e/0x270
Jul  3 15:01:44 [..]  [c019a2c0] do_sync_write+0x0/0x120
Jul  3 15:01:44 [..]  [vfs_write+0xb9/0x170] vfs_write+0xb9/0x170
Jul  3 15:01:44 [..]  [__capable+0x8/0x20] __capable+0x8/0x20
Jul  3 15:01:44 [..]  [sys_pwrite64+0x66/0x80] sys_pwrite64+0x66/0x80
Jul  3 15:01:44 [..]  [sysenter_past_esp+0x6b/0xa1]
sysenter_past_esp+0x6b/0xa1
Jul  3 15:01:44 [..]  ===
[.. this repeats x times ...]
[.. REBOOT / relevant boot messages ..]
Jul  3 15:33:04 [..] [  115.676800] Driver 'sd' needs updating - please use
bus_type methods
Jul  3 15:33:04 [..] [  115.676853] sd 0:0:0:0: [sda] 2927603712 512-byte
hardware sectors (1498933 MB)
Jul  3 15:33:04 [..] [  115.676863] sd 0:0:0:0: [sda] Write Protect is off
Jul  3 15:33:04 [..] [  115.676864] sd 0:0:0:0: [sda] Mode Sense: 06 00 10
00
Jul  3 15:33:04 [..] [  115.676879] sd 0:0:0:0: [sda] Write cache: enabled,
read cache: enabled, supports DPO and FUA
Jul  3 

[Samba] Re: Samba freezes server when writing 50MB MS Word document (sometimes)

2009-07-11 Thread Sebastian Scholz
Volker Lendecke Volker.Lendecke at SerNet.DE writes:
 
 On Sat, Jul 11, 2009 at 09:14:21AM +0200, Sebastian Scholz wrote:
  Hi list
  
  for a couple of days we have the problem that our samba pdc server (ubuntu
  8.04
  LTS [kernel 2.6.24-24] with 3.0.28a samba package) freezes sometimes when
  someone saves a specific word document (about 50MB) from MS Word. It's not
  always the same person. Saving the document - as nearly any huge MS Office
  document - takes some time, about one minute (guessed).
  
  First the server gets really slow. Some user programs on clients fail to
  write.
  The CPU usage of the smb process of the user writing the document goes up to
  nearly 100%. After a while it's not possible to log on to the server - even
  locally. It asks for a password but there won't be a prompt. Sometimes there
  is
  nothing on the console/screen. Hitting a button does not change a thing.
  It's
  not possible to kill the process. Another thing is that there are a couple
  of
  smbd processes for one user (see attached user smbd log).
 
 If it's not possible to kill smbd, even with -9, this really
 sounds like a kernel problem.
 
 Look for example at https://bugs.launchpad.net/linux/+bug/254326
 
 Volker


Hi Volker

thanks for your reply. 

You are right, it sounds more or less like that. The difference is that the
kernel was running for quite a while without any problems and we are using a
different board (Intel server board S3210SHLX with Core2Duo CPU).

Being Saturday today it was possible to install the lasted Ubuntu server kernel
update 8.04.3 (Linux xx 2.6.24-24-server #1 SMP Tue Jun 30 21:03:25 UTC 2009
i686 GNU/Linux). I tried to reproduce the fail but with just me in the office
and no real traffic it's hard to do so.

Maybe it just helped. I will wait and see.

Regards
Sebastian

-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba