samba process looping on some build servers

[Matthieu Patou]

Hello Tridge, Kamen,


so on some hosts (sernet-solaris8, builder72, sernet-sunX) we have a 
strange phenomena where process go in timeout but the time between the 
last test and the end is a couple of hours and sometime much more (ie 
8/9 hours).

I managed to capture this backtrace with dbx on sunX it was on the 
vampire_dc and according to the log it's when the dc is trying to 
vampire the DC "dc":

   [1] __time(0x4cf17f01, 0x7fb7b468, 0x0, 0x80808080, 0xff0000, 
0x80808080), at 0x7df3c898
   [2] _time(0x0, 0x7fb7b468, 0x7fb7b418, 0x27294b, 0xcaf638, 0x12c), at 
0x7df2a220
   [4] ldb_set_timeout_from_prev_req(ldb = 0x41df8, oldreq = (nil), 
newreq = 0xd98528), line 645 in "ldb.c"
   [5] ldb_build_search_req_ex(ret_req = 0xffbfb260, ldb = 0x41df8, 
mem_ctx = 0x114c30, base = (nil), scope = LDB_SCOPE_SUBTREE, tree = 0x2a8
b40, attrs = 0xffbfb2f8, controls = (nil), context = 0x6c61a8, callback 
= 0x7fb3cf50 = &ldb_search_default_callback(struct ldb_request *req,
  struct ldb_reply *ares), parent = (nil)), line 1078 in "ldb.c"
7dbd18 "objectGUID=f0bf42b7-0e13-42f0-bf6b-b4b93701450b", attrs = 
0xffbfb2f8, controls = (nil), context = 0x6c61a8, callback = 0x7fb3cf50 =
&ldb_search_default_callback(struct ldb_request *req, struct ldb_reply 
*ares), parent = (nil)), line 1116 in "ldb.c"
   [7] dsdb_module_search(module = 0x723f0, mem_ctx = 0x196ce8, _res = 
0xffbfb2fc, basedn = (nil), scope = LDB_SCOPE_SUBTREE, attrs = 0xffbfb
2f8, dsdb_flags = 1049605, format = 0x7d41ba58 "objectGUID=%s", ...), 
line 147 in "util.c"
   [8] dsdb_module_dn_by_guid(module = 0x723f0, mem_ctx = 0x161748, guid 
= 0xffbfb3f0, dn = 0x161748), line 200 in "util.c"
   [9] replmd_process_linked_attribute(module = 0x723f0, la_entry = 
0xd76210), line 3973 in "repl_meta_data.c"
   [10] replmd_prepare_commit(module = 0x723f0), line 4176 in 
"repl_meta_data.c"
   [11] ldb_transaction_prepare_commit(ldb = 0x41df8), line 409 in "ldb.c"
   [12] dsdb_replicated_objects_commit(ldb = 0x41df8, objects = 
0xf6c4f8, notify_uSN = 0xeba758), line 339 in "replicated_objects.c"
   [13] dreplsrv_op_pull_source_apply_changes_trigger(req = 0x8ee5d8, r 
= 0x476ec0, ctr_level = 6U, ctr1 = (nil), ctr6 = 0x89d1f8), line 602
in "drepl_out_helpers.c"
   [14] dreplsrv_op_pull_source_get_changes_done(subreq = (nil)), line 
528 in "drepl_out_helpers.c"
   [15] _tevent_req_notify_callback(req = 0xf6c4f8, location = 
0x7ede8fd8 "default/librpc/gen_ndr/ndr_drsuapi_c.c:751"), line 95 in 
"tevent_r
eq.c"
   [16] tevent_req_finish(req = 0xf6c4f8, state = TEVENT_REQ_DONE, 
location = 0x7ede8fd8 "default/librpc/gen_ndr/ndr_drsuapi_c.c:751"), line
104 in "tevent_req.c"
   [17] _tevent_req_done(req = 0xf6c4f8, location = 0x7ede8fd8 
"default/librpc/gen_ndr/ndr_drsuapi_c.c:751"), line 110 in "tevent_req.c"
   [18] dcerpc_drsuapi_DsGetNCChanges_r_done(subreq = 0x762310), line 
751 in "ndr_drsuapi_c.c"
   [19] _tevent_req_notify_callback(req = 0x762310, location = 
0x7f3dc030 "../../librpc/rpc/binding_handle.c:431"), line 95 in 
"tevent_req.c"
   [20] tevent_req_finish(req = 0x762310, state = TEVENT_REQ_DONE, 
location = 0x7f3dc030 "../../librpc/rpc/binding_handle.c:431"), line 104 i
n "tevent_req.c"
   [21] _tevent_req_done(req = 0x762310, location = 0x7f3dc030 
"../../librpc/rpc/binding_handle.c:431"), line 110 in "tevent_req.c"
   [22] dcerpc_binding_handle_call_done(subreq = (nil)), line 431 in 
"binding_handle.c"
   [23] _tevent_req_notify_callback(req = 0xd64ce8, location = 
0x7f3dbc68 "../../librpc/rpc/binding_handle.c:162"), line 95 in 
"tevent_req.c"
   [24] tevent_req_finish(req = 0xd64ce8, state = TEVENT_REQ_DONE, 
location = 0x7f3dbc68 "../../librpc/rpc/binding_handle.c:162"), line 104 i
n "tevent_req.c"
   [25] _tevent_req_done(req = 0xd64ce8, location = 0x7f3dbc68 
"../../librpc/rpc/binding_handle.c:162"), line 110 in "tevent_req.c"
   [26] dcerpc_binding_handle_raw_call_done(subreq = (nil)), line 162 in 
"binding_handle.c"
   [27] _tevent_req_notify_callback(req = 0x3bee70, location = 
0x7f3d7d60 "../librpc/rpc/dcerpc.c:221"), line 95 in "tevent_req.c"
   [28] tevent_req_finish(req = 0x3bee70, state = TEVENT_REQ_DONE, 
location = 0x7f3d7d60 "../librpc/rpc/dcerpc.c:221"), line 104 in "tevent_r
eq.c"
   [29] _tevent_req_done(req = 0x3bee70, location = 0x7f3d7d60 
"../librpc/rpc/dcerpc.c:221"), line 110 in "tevent_req.c"
   [30] dcerpc_bh_raw_call_done(subreq = 0x44ba68), line 221 in "dcerpc.c"
   [31] dcerpc_request_recv_data(c = 0x64ec28, raw_packet = 0xffbfc208, 
pkt = 0xffbfc0b8), line 1315 in "dcerpc.c"
   [32] dcerpc_recv_data(conn = 0x64ec28, blob = 0xffbfc208, status = 
RECORD), line 980 in "dcerpc.c"
   [33] sock_process_recv(private_data = 0x64ec28, blob = RECORD), line 
122 in "dcerpc_sock.c"
   [34] packet_recv(pc = 0x57a790), line 416 in "packet.c"
   [35] sock_io_handler(ev = 0x3d2f0, fde = 0x8fa440, flags = 1U, 
private_data = 0x64ec28), line 146 in "dcerpc_sock.c"
   [36] std_event_loop_select(std_ev = 0x3d368, tvalp = 0xffbfc4f0), 
line 509 in "tevent_standard.c"
   [37] std_event_loop_once(ev = 0x3d2f0, location = 0x22d50 
"../smbd/server.c:480"), line 548 in "tevent_standard.c"
   [38] _tevent_loop_once(ev = 0x3d2f0, location = 0x22d50 
"../smbd/server.c:480"), line 493 in "tevent.c"
   [39] tevent_common_loop_wait(ev = 0x3d2f0, location = 0x22d50 
"../smbd/server.c:480"), line 594 in "tevent.c"
   [40] _tevent_loop_wait(ev = 0x3d2f0, location = 0x22d50 
"../smbd/server.c:480"), line 613 in "tevent.c"
   [41] binary_smbd_main(binary_name = 0x22700 "samba", argc = 6, argv = 
0xffbfc92c), line 480 in "server.c"
   [42] main(argc = 6, argv = 0xffbfc92c), line 491 in "server.c"


Even if I'm don't know if it's exactly the same symptoms we have it with 
different tests here:

http://build.samba.org/build.cgi?function=View+Build;checksum=f2ee51ae0a9a60953da98d69d2e8674ad33065f2;host=builder72;revision=d237698850f00325346335b53cfe31657026b940;tree=samba_4_0_test;compiler=cc

Matthieu.

-- 
Matthieu Patou
Samba Team        http://samba.org
Private repo      http://git.samba.org/?p=mat/samba.git;a=summary