Greetings,

Until a few days ago, xadsd daemon was working just fine and we were adding computers to our domain with much gusto. When we reached 53 computers (I don't really know if computer count is really relevant here) the daemon began to crash.

Quick inspection in /var/log/messages shows that

Oct 8 16:50:18 gwdom kernel: xadsd[16773]: segfault at 0000000000000018 rip 00002aaaab399b60 rsp 000000004687b030 error 4
Oct 8 16:50:18 gwdom smbd[16972]: [2009/10/08 16:50:18, 0] rpc_server/srv_dce_funnel.c:ncalrpc_read_from_funnel(296)
Oct 8 16:50:18 gwdom smbd[16972]: ncalrpc_read_from_funnel: no data returned, exiting
Oct 8 16:50:18 gwdom smbd[16972]: [2009/10/08 16:50:18, 0] smbd/server.c:exit_server_common(750)
Oct 8 16:50:18 gwdom smbd[16972]: ================================================== =============
Oct 8 16:50:18 gwdom smbd[16972]: [2009/10/08 16:50:18, 0] smbd/server.c:exit_server_common(752)
Oct 8 16:50:18 gwdom smbd[16972]: Abnormal server exit: funnel closed on read
Oct 8 16:50:18 gwdom smbd[16972]: [2009/10/08 16:50:18, 0] smbd/server.c:exit_server_common(753)
Oct 8 16:50:18 gwdom smbd[16972]: ================================================== =============
Oct 8 16:50:18 gwdom smbd[16972]: [2009/10/08 16:50:18, 0] lib/util.c:log_stack_trace(1737)
Oct 8 16:50:18 gwdom smbd[16972]: BACKTRACE: 12 stack frames:
Oct 8 16:50:18 gwdom smbd[16972]: #0 /usr/sbin/smbd(log_stack_trace+0x1c) [0x5555557739bc]
Oct 8 16:50:18 gwdom smbd[16972]: #1 /usr/sbin/smbd [0x5555558266d5]
Oct 8 16:50:18 gwdom smbd[16972]: #2 /usr/sbin/smbd [0x555555826881]
Oct 8 16:50:18 gwdom smbd[16972]: #3 /usr/sbin/smbd(ncalrpc_read_from_funnel+0x181) [0x5555556cd581]
Oct 8 16:50:18 gwdom smbd[16972]: #4 /usr/sbin/smbd [0x5555555d5b9a]
Oct 8 16:50:18 gwdom smbd[16972]: #5 /usr/sbin/smbd [0x5555555d5f42]
Oct 8 16:50:18 gwdom smbd[16972]: #6 /usr/sbin/smbd(reply_trans+0x700) [0x5555555d6dc0]
Oct 8 16:50:18 gwdom smbd[16972]: #7 /usr/sbin/smbd [0x555555628a04]
Oct 8 16:50:18 gwdom smbd[16972]: #8 /usr/sbin/smbd(smbd_process+0x7b1) [0x555555629971]
Oct 8 16:50:18 gwdom smbd[16972]: #9 /usr/sbin/smbd(main+0xa20) [0x555555828270]
Oct 8 16:50:18 gwdom smbd[16972]: #10 /lib64/libc.so.6(__libc_start_main+0xf4) [0x2b5991110164]
Oct 8 16:50:18 gwdom smbd[16972]: #11 /usr/sbin/smbd [0x5555555bc689]
Oct 8 16:50:18 gwdom smbd[16972]: [2009/10/08 16:50:18, 0] smbd/server.c:exit_server_common(757)
Oct 8 16:50:18 gwdom smbd[16972]: Last message was SMBtrans
Oct 8 16:50:18 gwdom smbd[16972]: [2009/10/08 16:50:18, 0] lib/fault.c:dump_core(181)
Oct 8 16:50:18 gwdom smbd[16972]: dumping core in /var/log/samba/cores/smbd

It is strange, really. At first, our idea was that some computer probably emits malformed packets and the problem could be rectified by isolating the miscreant. Unfortunately, things were not that simple. Even when we configured the firewall to drop absolutely all packets to the DSfW server, the daemon still kept crashing (starts, runs approx. few minutes and then dies), with no external stimuli whatsoever.

How could we resolve this situation? What kind of misconfiguration might cause such problems?