[Bug 667269] Re: winbind crashes

[gmoore777]

similar to the stack trace in comment #6 above, winbind crashed today
(on machine bee64)  with this stack trace as gotten from gdb against the
core file left behind:

(gdb) bt
#0  0x00007f70d81fda75 in raise () from /lib/libc.so.6
#1  0x00007f70d82015c0 in abort () from /lib/libc.so.6
#2  0x00007f70da6a3460 in dump_core () at lib/fault.c:337
#3  0x00007f70da6b428e in smb_panic (why=<value optimized out>) at lib/util.c:1496
#4  0x00007f70da6a385d in fault_report (sig=11) at lib/fault.c:52
#5  sig_fault (sig=11) at lib/fault.c:75
#6  <signal handler called>
#7  get_cache (domain=0x0) at winbindd/winbindd_cache.c:139
#8  0x00007f70da616888 in resolve_username_to_alias (mem_ctx=<value optimized out>, domain=0x0, name=0x7f70db4aacb0 "gautam.naik", alias=0x7fffab4a4110) at winbindd/winbindd_cache.c:1019
#9  0x00007f70da609016 in normalize_name_map (mem_ctx=0x7f70db496180, domain=0x0, name=0x7f70db4aacb0 "gautam.naik", normalized=0x7fffab4a4110) at winbindd/winbindd_util.c:1442
#10 0x00007f70da6076ef in fill_grent_mem_domusers (domain=0x7f70db4b0d40, state=<value optimized out>, group_sid=0x7f70db46ced0, group_name_type=<value optimized out>, 
    num_gr_mem=<value optimized out>, gr_mem=<value optimized out>, gr_mem_len=0x7fffab4a44c8) at winbindd/winbindd_group.c:330
#11 fill_grent_mem (domain=0x7f70db4b0d40, state=<value optimized out>, group_sid=0x7f70db46ced0, group_name_type=<value optimized out>, num_gr_mem=<value optimized out>, 
    gr_mem=<value optimized out>, gr_mem_len=0x7fffab4a44c8) at winbindd/winbindd_group.c:578
#12 0x00007f70da607a9f in getgrsid_sid2gid_recv (private_data=0x7f70db46ceb0, success=<value optimized out>, gid=563) at winbindd/winbindd_group.c:866
#13 0x00007f70da600ba0 in process_loop (argc=<value optimized out>, argv=<value optimized out>, envp=<value optimized out>) at winbindd/winbindd.c:1115
#14 main (argc=<value optimized out>, argv=<value optimized out>, envp=<value optimized out>) at winbindd/winbindd.c:1440


I am running the latest winbind of 2:3.4.7~dfsg-1ubuntu3.6

So the problem still persists after upgrading to 2:3.4.7~dfsg-
1ubuntu3.6.

What other information do you want me to provide?

-- 
You received this bug notification because you are a member of Ubuntu
Foundations Bugs, which is subscribed to samba in Ubuntu.
https://bugs.launchpad.net/bugs/667269

Title:
  winbind crashes

Status in “samba” package in Ubuntu:
  Confirmed

Bug description:
  Binary package hint: samba

  Occasionally on various Ubuntu 64-bit LucidLynx systems (10.04.1), winbind goes into a state such that logging in fails.
  (we use ActiveDirectory for logging in. We do not use the Likewise-open product.)
  Although, it appears that there are 1 or more winbindd processes running, they are all dead/not_working/hung, etc.
  The only recourse to fix the problem, is that I log in with a local account that has sudo privilege, find and kill all the
  winbindd processes, and restart, nmbd, winbindd and samba. (not sure if I need to restart nmbd and samba, but
  I do it anyways.)

  This problem seems to happen more often on systems that are shared with other users.
  Some of those other users are connected from across the Wan via ssh.
  The problem seems to occur more often over the weekend. Meaning, Monday morning,
  is when we discover the machines are un-loginable with Active Directory account names.

  This problem does not happen with all 80+ machines. All machines are built out the same, 
  and are on approximately the same type of hardware. 

  This problem has existed since the onset of LucidLynx from what I can remember.
  The problem occurs frequently enough, that I had to set up a self-help web page for users with the 
  problem. From there I have a script that will log into the machine in question, 
  with a local sudo privileged account, kills all the winbind processes and restarts 
  nmbd, winbindd and samba.
  Even with using the local account, each command takes about 2 1/2 minutes to complete.
  It takes about 12 minutes to accomplish the "fixing" of the machine due to the # of commands that
  need to be run.

  I am assuming the core file under /var/log/samba/cores/winbindd/core
  will be relevant to the problem I am having.

  $ lsb_release -rd
  Description:    Ubuntu 10.04.1 LTS
  Release:        10.04

  
  $ apt-cache policy winbind
  winbind:
    Installed: 2:3.4.7~dfsg-1ubuntu3.2
    Candidate: 2:3.4.7~dfsg-1ubuntu3.2
    Version table:
   *** 2:3.4.7~dfsg-1ubuntu3.2 0
          500 http://us.archive.ubuntu.com/ubuntu/ lucid-updates/main Packages
          500 http://security.ubuntu.com/ubuntu/ lucid-security/main Packages
          100 /var/lib/dpkg/status
       2:3.4.7~dfsg-1ubuntu3 0
          500 http://us.archive.ubuntu.com/ubuntu/ lucid/main Packages