[Bug 5478] rsync: writefd_unbuffered failed to write 4092 bytes [sender]: Broken pipe (32)

[samba-bugs at samba.org]

https://bugzilla.samba.org/show_bug.cgi?id=5478

--- Comment #12 from Eric Shubert <ejs at shubes.net> 2011-09-22 18:35:06 UTC ---
(In reply to comment #11)
> When the process hangs, it always times out waiting for the
> first of the two smaller chunks (after 7 reads of 4092 bytes).
> 
Correction. It times out on the select function following the successful read
of the first of the two smaller chunks:
13018 10:38:47.708267 read(0, "\374\17\0\7", 4) = 4
13018 10:38:47.708522 select(1, [0], [], NULL, {60, 0}) = 1 (in [0], left {60,
0})
13018 10:38:47.708757 read(0,
"\6B#\377a\357\223\360\237\251J\2477q\216+\252s\\\363X\223\264\244pC\241y\350\321\240R"...,
4092) = 4092
13018 10:38:47.709034 select(1, [0], [], NULL, {60, 0}) = 1 (in [0], left {60,
0})
13018 10:38:47.709264 read(0, "\374\17\0\7", 4) = 4
13018 10:38:47.709619 select(1, [0], [], NULL, {60, 0}) = 1 (in [0], left {60,
0})
13018 10:38:47.709862 read(0,
"`\337\24f\220&\0379\223\"y\340\10\247\240\3702\333\360\311e/\362$\333\215\216r\24sw\361"...,
4092) = 3699
13018 10:38:47.710137 select(1, [0], [], NULL, {60, 0} <unfinished ...>
13017 10:39:22.695271 <... select resumed> ) = 0 (Timeout)
13017 10:39:22.695484 select(4, [3], [], NULL, {60, 0} <unfinished ...>
13018 10:39:47.701076 <... select resumed> ) = 0 (Timeout)
13018 10:39:47.701266 select(1, [0], [], NULL, {60, 0} <unfinished ...>
13017 10:40:22.696550 <... select resumed> ) = 0 (Timeout)

This is what happens in the normal flow:
13018 10:38:47.498236 read(0, "\374\17\0\7", 4) = 4
13018 10:38:47.498460 select(1, [0], [], NULL, {60, 0}) = 1 (in [0], left {60,
0})
13018 10:38:47.498710 read(0,
"\266\222\6\301\257q\224\0)\302\363R\273\rJ\264g\235\"0\230\311sq\277\243\260\277\322\32\313\213"...,
4092) = 4092
13018 10:38:47.499589 select(1, [0], [], NULL, {60, 0}) = 1 (in [0], left {60,
0})
13018 10:38:47.499852 read(0, "\374\17\0\7", 4) = 4
13018 10:38:47.500054 select(1, [0], [], NULL, {60, 0}) = 1 (in [0], left {60,
0})
13018 10:38:47.500278 read(0,
"z\33\0\276\304\177o\246\274\225gH\30\271|\317\365(h\234\314)\223\306\260?\370\274\355\272@\21"...,
4092) = 3699
13018 10:38:47.500599 select(1, [0], [], NULL, {60, 0}) = 1 (in [0], left {59,
810000})
13018 10:38:47.698206 read(0,
"\274&x\365\1\247\252Y\177\7e\17\227\325\207BL\177{\354_\317\2655\352~\20\303&\240\226\303"...,
393) = 393
1

I'm presently attempting to capture a corresponding strace on the source side
to see what's going on there.

-- 
Configure bugmail: https://bugzilla.samba.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the QA contact for the bug.